Strategic oil reserve sale: WCape High Court judgment

Disposal of R10m barrels of oil set aside, but CEF and SFF must pay Glencore $106m in compensation plus interest




Case No: 4305/18

In the matter between














Coram: Rogers J

Heard: 14-16 September, 23 October 2020; supplementary submissions on 10 November 2020

Delivered: 20 November 2020 (by email to the parties and same-day release to SAFLII)



Rogers J


[1] This case is about the sale, in late 2015/early 2016, of South Africa’s strategic stock of 10 million barrels of crude oil for about $281 million. The seller was the second applicant (‘SFF’), which is a wholly-owned subsidiary of the first applicant (‘CEF). They are public entities listed in Schedule 2 of the Public Finance Management Act 1 of 1999 (‘PFMA’). They seek the review and setting aside of the decisions to sell the oil and of the ensuing transactions.

[2] The oil is 5 million barrels of Basrah Light (‘Basrah’, an Iraq oil) and 5 million barrels of Bonny Light (‘Bonny’, a Nigerian oil). Basrah is less valuable than Bonny because it is heavier and has more sulphur. SFF has six underground storage tanks at Saldanha Bay with a combined capacity of 44,4 million barrels. The strategic stock was stored in Tank 2 (Basrah) and Tank 6 (Bonny).

[3] SFF sold 3 million barrels of the Bonny to the first respondent (‘Venus’). Venus immediately on-sold it to the second respondent (‘Glencore’).

[4] SFF sold 2 million barrels of the Bonny and 2 million barrels of the Basrah to the third respondent (‘Taleveras’). The fourth and fifth respondents (‘CTSA’ and ‘Natixis’) financed Taleveras’ acquisition of the oil. This was done by way of an on-sale of the oil by Taleveras to CTSA coupled with a repurchase obligation. (For convenience, I shall refer simply to CTSA, save where separate reference to Natixis is needed.)

[5] Finally, SFF sold 3 million barrels of the Basrah to the sixth respondent (‘Vesquin’). Vesquin is a subsidiary of the seventh respondent (‘Vitol Energy’) which is in turn controlled by the eighth respondent (‘Vitol SA’). Vesquin kept the oil. (Save where it is necessary to distinguish between them, I refer to the sixth to eighth respondents collectively as ‘Vitol’.)

[6] The sale agreements (SPAs) were coupled with storage agreements (SGAs) in terms whereof SFF would continue to store the oil in Tanks 2 and 6. The oil was still in these tanks when the applicants repudiated the transactions and gave notice that they were bringing review proceedings. Recently SFF pumped the Basrah from Tank 2 to Tank 5. Subject to any relief the applicants may be granted, the current owners of the oil are (or were until recently) Glencore (3 million barrels of Bonny), CTSA (2 million barrels of Bonny and 2 million barrels of Basrah) and Vitol (3 million barrels of Basrah). I say ‘until recently’, because Vitol cancelled its contracts with SFF in June 2020, the practical effect of which is that it has abandoned its oil with a view to claiming damages.

[7] The ninth respondent is the Minister of Energy (‘MoE’). The tenth respondent is the Minister of Finance (‘MoF’). They do not oppose the application and have not participated in the proceedings. The Organisation Undoing Tax Abuse (‘OUTA’) was admitted as an amicus curiae without opposition.

[8] Venus has not participated in the proceedings. Glencore, the party to whom Venus on-sold the 3 million barrels of Bonny, opposed the application and filed affidavits. During the course of the hearing, the applicants and Glencore came to terms reflected in a draft order. The result is that counsel for the applicants and Glencore did not address me on the merits of Glencore’s opposition or on its contentions about just and equitable relief.

[9] Taleveras does not oppose the review but filed ‘explanatory’ affidavits, the thrust of which is that it has accepted an offer, allegedly made by the applicants in their affidavits, to refund the purchase price and storage fees plus interest. CTSA opposes the review and contends that if restitution is to be made the money should go to it, not to Taleveras.

[10] Vitol, like CTSA, opposes the review. In June 2020 Vitol cancelled its contracts with SFF on the basis that it would pursue a claim for damages. It thus no longer lays claim to ownership of the Basrah which it bought. In this respect it differs from CTSA, which continues to assert ownership of the 4 million barrels it bought from Taleveras. Glencore’s formal position on the papers is the same as CTSA’s, namely that it is the owner of the 3 million barrels of Bonny which it bought from Venus.

[11] Various supplementary affidavits were delivered. At the start of the hearing the parties agreed that all such affidavits could be received and form part of the record.

[12] In their replying affidavits the applicants sought to amend their notice of motion. There was no objection, and the amendment was granted. The applicants’ counsel clarified that the omission, from the latest notice of motion,[1] of a reference to just and equitable relief was an oversight. Accordingly, the latest notice of motion must be read together with paras 9-11 of the previous notice of motion.[2]

[13] A point to make at the outset is this. Large volumes of crude oil are traded globally every day. Currently global demand is about 100 million barrels a day.[3] During 2015 South Africa imported 140 million barrels.[4] Although a global crisis could have left South Africa short of oil (hence the need for strategic stock), no such shock has occurred since the impugned disposals. Success for the applicants does not mean that the country will retain a strategic stock of oil which would otherwise be forever lost; it just means that SFF gets to keep the oil it sold rather than having to replenish it at potentially higher prices.

Oil terminology

[14] Crude oil is normally sold with reference to a benchmark price. In the present case, Brent (more specifically Dated Brent) was the benchmark. Brent represents a basket of North Sea oils. It is the main benchmark against which crude oil is priced. Brent prices are published daily by S&P Global Platts. Published prices on any given date differ according to delivery date. Dated Brent is the FOB price for prompt delivery (10-30 days hence). Prices for more distant delivery (forward prices) may be higher or lower than Dated Brent. These prices stretch several years into the future. Future-delivery oil is actively traded on the Intercontinental Exchange (‘ICE’) .

[15] A typical pricing formula would refer to the average of Dated Brent over a specified number of days following some relevant future event, eg bill of lading or in-tank transfer (‘ITT’). (I refer to this as the pricing window.) Where the subject oil is inferior or superior to Brent, the parties would agree a discount or premium to Brent.

[16] On any given day, the prices published for a range of future Brent delivery dates can be represented as a line (curve) on a graph, starting with Dated Brent on the left. The price (or forward) curve for future delivery may slope upwards (indicating that today’s prices for future delivery are higher than today’s prices for prompt delivery) or downwards (the opposite). Each new trading day has its own forward curve, because Dated Brent and its forward prices on a later date may turn out differently from those that were predicted by the forward curve on an earlier date.[5]

[17] When the forward curve is sloping upwards, the market is said to be in contango. When it is sloping downwards, the market is said to be in backwardation. In a contango market, traders may wish to buy prompt-delivery oil (at lower prices), store it, and sell it when the market reaches the higher prices predicted by the price curve. This is called a contango strategy or contango play. For the same reason, operators with storage space can expect to maximise their storage fees in a contango market, since this is when contango traders are looking for storage.

[18] During 2015 and until August 2017 the oil market was in contango. In the present case, Glencore, Taleveras and Vitol were engaged in a contango strategy. Because the forward price of oil in December 2015/January 2016 was higher than the current price, they bought the oil at relatively low prices, expecting that the market value would go up. If this happened, a point might be reached when the market switched to backwardation, so that current prices were higher than forward prices. Once the market was in backwardation, it would no longer make sense to store the oil. The traders would thus uplift and sell the oil, anticipating to do so in a way that covered the original purchase price and the storage fees and still leave them with a profit (their ‘contango premium’).

[19] However, because the traders could not be sure that the market would turn out as anticipated, it was prudent to hedge the risk that the SGAs might come to an end without the oil reaching or exceeding the forward prices prevailing when they bought the oil. Although CTSA was not a contango trader, it had a similar incentive to hedge its risk. At the risk of over-simplification, hedging instruments involve parties and counterparties who are willing to exchange the risk of any difference between the forward price at the time the hedge is concluded (this is fixed on day one) and the actual price when the forward date is reached (this is unknown on day one, and is thus called the ‘float price’).

[20] A popular hedging tool is the ICE Futures Brent contract, of which large volumes are traded daily. To hedge over a period of several years, a trader would typically sell a hedge with a closing date several months hence, and then ‘roll it over’ by buying back the hedge and selling a new one, so that by the time the trader is ready to uplift and sell its oil the most recent hedge is still being traded. Once the sale price of the physical oil has been fixed, the trader will close its hedging position by buying back the last hedge. Hedging allows the trader to ‘lock in’ its contango premium.

[21] In the present case, Glencore, CTSA and Vitol hedged their risk when buying the oil. This they did by selling hedges in an equal and opposite position to the oil itself. The effect of these hedges was that if the float prices were below the forward prices when the hedges came to an end, the hedging counterparties would have to pay the difference to the relevant respondents (this was the risk against which the respondents were hedging).

[22] However, it so turned out that in the latter part of 2017 and in 2018 the float prices were higher than the forward prices. This meant that when closing out their hedges, the relevant respondents would have to pay to the hedging counterparties the difference between the forward price and the float price. In the ordinary course, this would not be a problem, because the relevant respondents would simultaneously get access to the oil and sell it at the higher float prices. That is how hedging works: depending on the relationship between the float price and the forward price, you either lose on the hedge but make on the oil, or you lose on the oil but make on the hedge.

[23] However, the ordinary course of events was thrown into turmoil when the applicants repudiated the transactions. The market switched from contango to backwardation in August 2017. It was at around this time that the relevant respondents wanted to uplift and sell their oil. But it was also at around this time that the applicants asserted the invalidity of the transactions and stated their intention to bring review proceedings. So Glencore, CTSA and Vitol ended up closing their hedges at a loss, without having access to the oil from which to recoup the loss.

Mr Gamede’s amicus/intervention application

[24] On 25 September 2020, about a week after argument was completed, Mr Sibusiso Gamede, who was SFF’s Acting Chief Executive Officer (‘CEO’) at the relevant time and was central to the impugned transactions, applied to be admitted as an amicus curiae. The applicants and CNX opposed, the other respondents abided. Gamede wanted to file an affidavit on the merits. He did not give the gist of his version, but one could infer from his application that it would have been self-exculpatory.

[25] In the light of the opposing papers, Gamede in reply abandoned his request for admission as an amicus and instead asked the court to allow him to intervene as a party because the judgment might adversely affect his reputation. His counsel acknowledged in his written submissions that the amicus application had been hopeless. Counsel for the applicants and CTSA submitted that reputational harm was not a justification for intervention (see National Director of Public Prosecutions v Zuma 2009 (2) SA 277 (SCA) paras 84-87, recently followed in this division in B Xulu & Partners Incorporated & another v Department of Agriculture, Forestry and Fisheries & another [2020] ZAWCHC 99 para 57.) In oral argument Gamede’s counsel conceded that there was no way around these authorities.

[26] I accordingly dismissed Gamede’s application with costs, including the costs of two counsel where employed. In regard to costs, I assumed in Gamede’s favour that the principles laid down in Biowatch Trust v Registrar Genetic Resources & others [2009] ZACC 14; 2009 (6) SA 232 (CC) applied. The general principle that an unsuccessful private party in constitutional litigation should not be penalised in costs is not unqualified. In para 24 of Biowatch Sachs J said the following:

‘If an application is frivolous or vexatious, or in any other way manifestly inappropriate, the applicant should not expect that the worthiness of its cause will immunise it against an adverse costs order. Nevertheless, for the reasons given above, courts should not lightly turn their backs on the general approach . . .’

(See also Limpopo Legal Solutions and Another v Eskom Holdings Soc Limited [2017] ZACC 34; 2017 (12) BCLR 1497 (CC) paras 33 and 41; S S v V V S [2018] ZACC 5; 2018 (6) BCLR 671 (CC).)

[27] Gamede’s amicus application and his belated intervention application were ill-considered and manifestly inappropriate. I do not, here, criticise his counsel, who was briefed at a very late stage and had insufficient time to research the matter.

[28] To the extent that my findings reflect adversely on Gamede, they have been reached without regard to the evidence he wanted to adduce. If Gamede feels that a public statement setting out his side of the story is necessary to protect his reputation, my judgment will be no bar to his doing so. I should mention, however, that after an initial interview on 15 November 2018 with forensic investigators appointed by the applicants, Gamede did not cooperate further with the investigation. The applicants have placed the interview transcript before the court. If Gamede had granted further interviews, those transcripts would also have been placed before me. Furthermore, the case has attracted wide publicity since it was launched in March 2018. If Gamede believed that he had relevant evidence to offer, he could have tendered it to the litigants in good time.

Ms Joemat-Pettersson’s intervention application

[29] Ms Tina Joemat-Pettersson was the MoE at the time of the impugned transactions. On 17 September 2020, the day after argument in the main case was completed, her attorneys wrote to the Chief Registrar to say that their client had been adversely affected by the proceedings and would be applying to intervene. More than a month later, on 20 October 2020 she served an application for intervention. I issued directions in respect of her intervention application on 27 October 2020, specifically requiring that the heads of argument address the authorities which had caused Gamede’s counsel to throw in the towel. On 29 October 2020 she withdrew her application.

The issues

[30] The main issues in the case are delay, and just and equitable relief. Delay features in two ways: first, as a defence on the merits of the review; and second, as a factor relevant to just and equitable relief. CTSA and Vitol contend that the delay has been such that the court should not reach the merits. If that contention fails, they argue that delay is an important factor in favour of granting them more ample relief than the applicants tender.

[31] If I find that delay should not stop me reaching the merits, it is common cause that the impugned decisions and agreements are invalid and that a declaration to this effect should be made. On just and equitable relief, the following four questions will then arise:

(a) The first is whether I should set aside the contracts, something which would ordinarily, but not invariably, follow from a declaration of invalidity.

(b) The second is whether, if the contracts are set aside, accrued contractual rights should be preserved. In the context of this case, there is little practical difference between (a) and (b), since all relevant rights have accrued. If accrued rights are preserved, Venus, Taleveras and Vitol acquired good title to the oil, and Venus and Taleveras were able to pass good title to Glencore and CTSA. Contractual claims for damages against SFF would also be preserved.

(c) The third question arises if the transactions are set aside and accrued rights are not preserved. In that event, the position will be that SFF remained the owner of the oil. It is then common cause that SFF should restore the purchase price and storage fees plus interest. On restitution, there are issues about the period over which, and the rate at which, interest should be calculated. There is also the question whether SFF should pay the restitution to Taleveras or to CTSA.

(d) The fourth question is whether, in addition to restitution, CTSA and Vitol should be compensated for the losses they have suffered. If the impugned transactions are set aside without preservation of accrued rights, neither CTSA nor Vitol seek recompense for the profits they would have made had the transactions stood. They do contend, though, that they should be compensated for their actual losses. The biggest components of these losses are hedging losses. Other items of loss include the cost of insuring the oil, inspections costs and the cost of establishing letters of credit when buying the oil.

[32] Since nobody applied for a referral to oral evidence, fact-finding must take place in accordance with the Plascon-Evans rule.

[33] The further structure of this judgment is as follows:

(a) In paras 34-37 I provide further information about the two applicants.

(b) In paras 38-88 I give the background leading up to the conclusion of the impugned contracts.

(c) I then deal with the various sets of transactions: SFF/Venus/Glencore at paras 89-106; SFF/Taleveras/CTSA at paras 107-119; and SFF/Vitol at paras 120-129.

(d) Next I deal, in paras 130-169, with the events up to the launching of the review application on 12 March 2018 and with the way in which Gamede’s actions came under scrutiny.

(e) Thereafter I trace the procedural history of the review (paras 170-188). This includes a survey of why it was that the applicants’ final supplementary founding papers were only delivered at the end of February 2020.

(f) I then turn to the grounds of review, which I traverse in paras 189-285. I do so at that stage in the judgment because the grounds and merits of the review are relevant to the question whether delay should be overlooked or condoned. I find that the impugned decisions and transactions are susceptible to review on many grounds, and that such grounds are clear.

(g) I address, next, the legal principles applying to delay (paras 286-298). I deal with the distinction between delay in legality reviews and PAJA reviews. This part of the judgment also deals with the so-called Gijima principle, which holds that even where there has been an unreasonable delay which should not otherwise be overlooked, an impugned decision may be so patently invalid that the court must so declare it in line with the injunction contained in s 172(1)(a) of the Constitution. I conclude that the Gijima principle applies not only to legality review but also to PAJA review.

(h) In paras 299-328 I discuss the delay which occurred in this case. I find that in relation to the most important impugned decisions, namely the conclusion of the contracts, time began to run by not later than the meeting of SFF’s board on 5 February 2016. I deal with the delay up to the launching of the review in March 2018 and with the post-launch delay to the end of February 2020 and beyond. I find the delay to be unreasonable, egregious and unexplained or unsatisfactorily explained.

(i) In paras 329-333 I explain why, nevertheless, the Gijima principle compels me to overlook the delay and declare the impugned decisions and transactions invalid. This means that the focus turns to just and equitable relief.

(j) I accordingly address, next, the legal principles relating to just and equitable relief (paras 334-366). I emphasise the wide powers the court is given in s 172(1)(b) of the Constitution and in s 8(1) of PAJA. I note the distinctions between affected parties who are complicit and affected parties who are innocent or even proactive in seeking to act correctly. I deal with the subject of compensation, making the point that the parties seeking to vindicate constitutional rights in this case are the applicants. The relief they claim is a setting aside of all the decisions and transactions. Compensation in the present case is not a remedy claimed by a party alleging that its constitutional rights have been violated; compensation here would operate to ameliorate the otherwise harsh effects of the relief claimed by the applicants.

(k) I then turn to deal with the factors relevant to an assessment of just and equitable relief in the present case. Having already dealt with delay (which is one such factor) in earlier paragraphs, I address, next, the commercial prejudice suffered by the various respondents, focusing mainly on Glencore, CTSA and Vitol (paras 367-398).

(l) A further relevant factor is alleged misconduct by the respondents, which I address in paras 399-478. I find there to have been such misconduct in the case of Taleveras and Venus. Earlier in the judgment, I find that Taleveras paid bribes to Gamede. I also explain why there could have been no legitimate basis for Venus to have been selected as a buyer of oil and that there must have been some impropriety, even if the details have not been uncovered. On the other hand, I conclude that Glencore, CTSA and Vitol are innocent parties.

(m) Also relevant is misconduct by the applicants, which I address in paras 479-487. Apart from egregious delay, I point out that even though Gamede might have been the primary actor in the improper transactions, there was a pervasive lack of oversight and intervention by SFF’s senior management and the boards of SFF and CEF.

(n) This then leads me, in paras 488-501, to a broad assessment of just and equitable relief, focusing on Taleveras/CTSA and Vitol. I do not deal with Glencore at this point, because of the agreement it has reached with the applicants. My conclusion is that, insofar as CTSA and Vitol are concerned, their contracts should either stand or they should be set aside on the basis of compensation for the out-of-pocket expenses incurred by these two parties. I find, however, that compensation would only be a just and equitable remedy if I can be satisfied that CTSA and Vitol’s contractual rights to damages (which they would have if the contracts were not set aside) would exceed the proposed compensation, since otherwise the public interest might be better served by allowing the contracts to stand. I call this the ‘contractual qualification’.

(o) Having expressed these broad views, I look more closely at the appropriate relief pertaining to Vitol (paras 502-517) and Taleveras/CTSA (paras 518-563), and I also explain briefly why I am willing to endorse the draft order agreed between the applicants and Glencore (paras 564-567). I explain why the contractual qualification is satisfied in relation to Vitol, CTSA and Glencore. In regard to Taleveras/CTSA, I conclude that the purchase price and storage fees received by SFF should be repaid not to Taleveras but to CTSA, either as restitution or as compensation. I also explain why, provisionally, I will not allow CTSA to recover from SFF the net portion of the purchase price it paid to Taleveras ($22,568,426). This is on the basis that such amount represents Taleveras’ profit. Given the misconduct of which Taleveras was guilty, I need to construct my order in such a way that SFF does not end up paying for Taleveras to keep its profit. I do this by requiring CTSA to excuss Taleveras before looking to SFF.

(p) I then deal (paras 568-585) with claims for the time value of money (interest), and conclude that this should be additionally awarded, up to the date of judgment, as part of the compensation payable to Vitol and CTSA. In Vitol’s case, I am not satisfied that I should award interest at the Vitol group’s blended cost of borrowing, and accordingly confine Vitol to interest at the deposit rates earned by SFF on its dollar accounts. On the other hand, I am satisfied that CTSA incurred borrowing costs at the rates it claims. In regard to interest on the awarded compensation as from the date of judgment, I explain why I feel entitled to depart from the rate laid down in the Prescribed Rate of Interest Act 55 of 1975.

(q) I conclude in paras 586-588 with costs and my order.

The applicants

[34] CEF was incorporated as a company in terms of the Companies Act 61 of 1973. In terms of s 1D of the Central Energy Fund Act 38 of 1977 (‘CEF Act’), CEF’s sole shareholder is the State, and SFF’s sole shareholder is CEF. The MoE represents the State as CEF’s shareholder. In terms of s 1(4), the MoE appoints CEF’s board members. Section 1E provides that CEF’s board chairperson is its, and the SFF’s, accounting officer with the duties set out in that section.

[35] In terms of s 49(2)(a) of the PFMA, SFF’s ‘accounting authority’ is its board. SFF’s board thus has the fiduciary duties and general responsibilities set out in ss 50 and 51 of the PFMA while the CEF board chairperson, as SFF’s ‘accounting officer’, has the duties set out in s 1E of the CEF Act.

[36] The first-stated object in s 2 of the National Energy Act 34 2008 (‘Energy Act’) is to ensure uninterrupted supply of energy to the Republic. A related object is to

‘provide for optimal supply, transformation, transportation, storage and demand of energy that are planned, organised and implemented in accordance with a balanced consideration of security of supply, economics, consumer protection and a sustainable development’.

[37] Section 17(1) of the Energy Act states that, for purposes of ensuring security of supply, the MoE may, in a prescribed manner, direct any state-owned entity ‘to acquire, maintain, monitor and manage national strategic energy feedstocks and carriers’. In terms of s 17(2), the entity must perform the directed function ‘in accordance with the relevant published security of supply strategies or policies’. In respect of crude oil, SFF is the entity to which the MoE has issued such directions.

Factual history leading up to disposal and storage agreements

[38] The holding of strategic fuel stock was important to the apartheid government because of the international oil embargo. The embargo was lifted in 1994. Retaining a strategic stock of crude oil has nonetheless remained important to guarantee oil for our refineries in case of disruption in international supply.

[39] Fuel stock ‘rotation’ is sale of existing crude oil coupled with arrangements to replace it with other oil. Over the period 2003-2008 PetroSA, which at that time managed South Africa’s strategic oil stocks on SFF’s behalf, engaged in five rotations. In 2010 SFF took over the direct management of the oil.

[40] The MoE from 25 May 2014 to 31 March 2017 was Ms Tina Joemat-Pettersson. Gamede, an attorney, was SFF’s General Counsel from April to September 2015, and also served as Special Adviser to the MoE. Following the resignation of SFF’s CEO, Mr Bheki Gila, on 12 August 2015, Gamede was appointed Acting CEO on 4 September 2015 while continuing as the MoE’s Special Adviser.

[41] Gamede played a central role in the impugned transactions. The SFF’s board chair, effectively from January 2016, was Mr Riaz Jawoodeen. Gamede’s management team included Mr Mfanafuthi Nkutha as Chief Operations Officer (‘COO’); Mr Sivuyile Ngqongwa as Chief Financial Officer (‘CFO’); Mr Mzwakhe Ndlela as General Manager: Operations; Mr Lucky Mayaphi as General Manager: Safety, Health Environment and Quality and Risk; Ms Marion de Wet as General Manager: Commercial; and Ms Daphne Chili as General Counsel.

[42] On 7 August 2014 the MoE issued a directive (‘First Directive’) laying down an ‘interim regime’ for the management of the strategic stocks. In terms of the First Directive, SFF was to hold a strategic crude oil stock of 10,3 million barrels equivalent to 21 days of national import requirements.[6] The trading of strategic stock would be permitted under a regime to be determined by the Department of Energy (‘DoE’) and approved by the MoE. One of the financial sources to acquire, operate and manage the strategic stock was rental income from the leasing by SFF of surplus storage capacity. The MoE required SFF to submit an ‘optimisation plan’ within six months.

[43] On 1 November 2014 SFF’s CEO (Gila) submitted what he described as the first of SFF’s optimisation initiatives. The document explained to the MoE the terms ‘backwardation’ and ‘contango’. Gila said that when the market is in backwardation, spot transactions are favoured, with the result that there is not enough oil to fulfil forward contracts. This tends to drive up the price. Less oil is stored. Gila stated that since September 2014 the oil market had tipped towards contango, and oil prices were declining. This made SFF’s storage capacity more valuable. SFF thus proposed to the MoE leasing Tank 2, as well as the Basrah stored therein, for six months with an option to renew for a further six months. (This proposal was not taken further.)

[44] On 3 August 2015 the MoE issued a further notice (‘Withdrawal Notice) withdrawing the First Directive. The First Directive, she explained, was meant to have served as an interim regime until cabinet approved a strategic stocks policy and SFF legislation, but there seemed to be little intention to finalise these matters. She stated that in terms of the First Directive, SFF was to have presented a detailed optimisation proposal, which had not been done. She required SFF to submit a detailed report on this and other matters raised in the First Directive. One of these was the trading of strategic stocks under a regime to be determined by the DoE and approved by the MoE.

[45] On 15 August 2015, Nkutha and Ngqongwa made a presentation to Parliament’s Portfolio Committee on Energy. Although Vitol’s deponent says that this presentation highlighted SFF’s intention to move away from holding crude oil in favour of finished product, I do not so read it. What the presentation did highlight was that SFF’s business model, particularly the earning of revenue from storage, was vulnerable to the whims of the oil market. The authors stated that periods favourable to storage, viz contango periods, were few and far between. A contango phase had followed the global financial crisis of 2008. With the recovery of the oil price thereafter, SFF’s revenues have tumbled, until the price ‘nose dived’ again in mid-2014 (contango), leading to an improvement in SFF’s revenues as from January 2015. The presentation contained nothing material relating to strategic stocks policy.

[46] On 4 September 2015 Vitol, having cleared the wording with Gamede, wrote to the MoE with various proposals. One of these was that Vitol lease strategic stocks of Bonny or Basrah, pledging equivalent oil which it stored at Saldanha on a 1:1 basis. (This proposal did not come to fruition.)

[47] On 15 September 2015 Vitol wrote two more letters to the MoE. In one letter, Vitol told the MoE that South Africa was a core area for Vitol’s investment plans. It already had long-term SGAs with SFF. In order to strengthen its commitment to South Africa, Vitol sought permission to engage SFF with a view to concluding a long-term strategic collaboration agreement.

[48] In the other letter, Vitol sought permission to partner with SFF on various areas of mutual interest. The first was to generate greater and more consistent revenue for SFF by using its Saldanha Bay storage facilities, and the strategic stocks held there, ‘in a fully secured and risk free structure’. The lease-and-pledge proposal was repeated. Vital also proposed that SFF’s current grades of strategic stock be swapped ‘for more relevant grades’ preferred by the local refining industry. The author proposed, finally, that SFF and Vitol establish a task team to write a position paper on strategic stock holdings.

[49] On 6 October 2015 Gamede wrote to the MoE, asking her to retract the Withdrawal Notice and reinstate the First Directive. He told her that if SFF was to fulfil its mandate of energy security, it had to be given the responsibility to acquire, maintain, monitor and manage fuel stocks as provided for in s 17 of the Energy Act. He proposed a business model which included trading in commercial stock and the rotation of strategic stock. He proposed three conditions for rotation transactions: (a) that crude oil prices were on the rise; (b) that there was a positive margin for SFF, ie that the selling price should be more than the purchase price; and (c) that a risk assessment be done before any transaction was concluded. He ended by saying that in order for his proposals to be implemented successfully, SFF would establish a trading department headed by a general manager who would provide the MoE with monthly reports. (According to the applicants, Gamede was by now hatching a plan to dispose of the strategic stocks. He wanted the Withdrawal Notice to be retracted to clear the way.)

[50] In a letter of 8 October 2015 (‘Second Directive’) the MoE acceded to Gamede’s request. She stated that the following conditions would apply: (a) Any rotation of strategic stock would require ministerial approval, preceded by detailed due diligence and supported by a comprehensive motivation. (b) The integrity of strategic stock levels had to be ensured at all times. (c) A trading division had to be established and staffed with skilled personnel and resources to undertake the trading activities. (d) The SFF was to provide monthly reports to the MoE and DoE on all activities undertaken in terms of the Second Directive.

[51] Two days later the MoE, through her special adviser, replied to Vitol’s letters of 15 September, expressing delight at its interest and advising that Gamede was mandated to engage with Vitol.

[52] SFF’s board considered the Second Directive at a meeting on 13 October 2015. Gamede was present as an invitee. The board noted that it would need a project plan and an action plan for implementing the directive. The board noted that SFF did not have funds to implement some of the activities and would need clarification from the MoE about funding the strategic stocks and infrastructure. Gamede was asked to draft a clarificatory letter for review by the board. The board noted, further, that SFF had to develop a policy for the rotation of stock and that the policy needed to be approved by the board whereafter it could be submitted to the MoE.

[53] On the same day, and despite the terms of the Second Directive and the board’s decisions, Gamede issued identical requests for proposals (‘RFPs’) to five entities. Vitol was one. The others were Golden Nest International (‘GNI’), Skydeck, Total and Mecuria. The RFP, to which the Second Directive was attached, stated that, following the authorisation contained therein, SFF ‘would like to invite your company to submit a proposal for rotation of strategic stock’. That was all. The letter did not contain specifications with which proposals had to comply, did not say how proposals would be evaluated, did not specify documents to be submitted, and did not give a time-limit. There is nothing in the record revealing why Gamede chose these five entities.

[54] Although the RFPs were issued on 13 October 2015, GNI had already submitted a proposal on 8 October 2015, presumably pursuant to informal engagement with Gamede. On 14 October 2015, and again despite the board’s decisions, Gamede wrote to the MoE to say that SFF had assessed GNI’s proposal and found it to be sound. He sought approval to explore that proposal. There is nothing to show that anyone other than Gamede was involved in ‘assessing’ GNI’s proposal. (In the event, GNI was not involved in the impugned transactions.)

[55] Gamede chaired SFF’s exco meeting on 16 October 2015. The meeting decided that SFF’s operational (strategic) plan needed to be ‘operationalised’. The Second Directive, particularly trading and the rotation of strategic stock, had to be incorporated into the plan. Although not recorded in the minutes, the applicants say that Gamede, with the undisclosed intention of favouring Vitol, proposed that in implementing the Second Directive SFF should engage only with those parties who already had storage agreements with SFF. In response, Ngqongwa said that SFF would have to comply with its procurement and disposal processes, and that a competitive and open process should be followed. Gamede did not disclose that he had already issued five RFPs or that he had already recommended GNI’s proposal to the MoE.

[56] In the second half of October 2015 Vitol and Skydeck submitted responses to the RFPs. Gamede also issued three more RFPs, identical to the earlier ones, to Enviroshore, Zittatu and Taleveras. Again, why these three entities were favoured does not appear. Taleveras submitted a proposal in early November, Zittatu in late November.

[57] The exco met again on 2 November 2015. Included in the agenda was a draft stock rotation policy written by Mayaphi. The draft defined rotation of strategic stock as meaning that a predetermined quantity would be moved in and out of the storage terminal over a specified period. The quality of the stock would be maintained or improved. The purpose of rotation was to generate revenue through trading the physical stock, ‘viz a rotation mechanism as well as applying derivative instruments’. Rotation would be done in a way that would not compromise South Africa’s security of supply.

[58] Mayaphi proposed a rotation of two million barrels per rotation cycle. Replacement stock was to be returned to SFF within three to six months. Potential customers for rotation would need to meet the following criteria: (a) that the customer had a matching quantity of stock in storage at Saldanha Bay, with API[7] and sulphur content[8] being key quality parameters; (b) a rigorous due diligence on financial standing; and (c) an acceptable bank guarantee or letter of credit from a first-class international bank. Potential customers would present proposals specifying the quantity and period of rotation. The proposal would ‘be taken through the necessary approval channels’ before product was released, meaning evaluation by exco members for commercial viability, followed by a recommendation to the MoE for consideration and approval.

[59] The exco minutes on this item simply record that ‘a draft or finalised document needs to be compiled for exco members to go through prior to being submitted to board’. According to the applicants, Gamede did not disclose that RFPs had already been issued and that some responses had been received.

[60] On the same day as exco’s meeting, the MoE replied to Gamede’s letter of 14 October 2015 about GNI. She authorised Gamede to engage with GNI on two conditions: (a) that the agreement between SFF and GNI should specify that ownership of the oil was to remain with SFF coupled with suitable security; (b) that GNI should replace the stock with crude of the same quantity and quality every three months.

[61] On 4 November 2015 Taleveras submitted its proposal. By this stage, therefore, Vitol and Taleveras, being two of the three entities to whom the strategic stock was eventually sold, had submitted proposals.

[62] On 11 November 2015, by which time Gamede had engaged in further correspondence with Vitol and Taleveras, he wrote to the MoE to request permission to sell the entire strategic stock of 10,3 million barrels on the basis that each barrel would be sold at the prevailing market price and the proceeds used to buy 10,3 million barrels of crude in the open market.

[63] He told her that most of the strategic crude had been sitting in storage for many years and was ‘losing its relevance with the changing market conditions’. For environmental reasons, lighter sulphur was in demand. This had caused SFF to review the quality of its current strategic stock. If the MoE granted the approval, SFF could access fresh stock in line with changing market requirements. The current stock would be sold at market prices ‘and will be replenished when the market prices of favourable for SFF to acquire such barrels’. This would ensure that SFF ‘creates value that will yield a positive net margin on the selling and buying initiatives of crude oil’.

[64] The applicants say that this letter was materially misleading. There were no legitimate quality concerns. Furthermore, because the oil price was depressed, it was unlikely that SFF would be able to replenish stock at lower prices. In other words, the oil price was more likely to rise than fall further.

[65] Gamede’s request to the MoE had not been considered or approved by the board, and there is no evidence that other members of exco knew of it. Gamede’s proposal was not in line with Mayaphi’s draft rotation policy.

[66] The next day, 12 November 2015, the MoE granted the requested approval (‘First Approval Notice’). She stated that the plan had to be executed in a way that addressed South Africa’s need for strategic stock while at the same time serving as a catalyst to ensure SFF’s financial self-sustainability. There is no evidence that she interrogated Gamede’s claims before issuing her letter.

[67] On 22 November 2015, Gamede issued a more extensive RFP to Vitol. (More accurately, he emailed it to Vitol’s Marc Ducrest with a message: ‘My brother please look at this letter and see if it is in order before I sign and send it to you.’ The RFP contained, as background, the same motivation for rotation which Gamede had given the MoE on 11 November 2015. The RFP went on to say that rotation, sale and purchase of strategic stock would be subject to terms and conditions ‘as will be advised in the term sheet and general form of agreement’. Vitol was invited to submit a proposal to participate. The RFP specified various documents that had to be submitted. The response had to be submitted in a sealed envelope within two days of receipt of the letter. (It is unclear why another RFP was sent to Vitol.)

[68] At SFF’s board meeting on 23 November 2015, it was reported (presumably by Gamede, who was present as an invitee) that SFF’s strategic plan would be submitted at the board’s January 2016 meeting. The board pack included Gamede’s letter to the MoE of 11 November 2015 and the MoE’s approval of 12 November 2015. The board noted (a) that it had not been made aware of Gamede’s request before it was sent;(b) that it was unfortunate that the MoE’s approval referred to a disposal of stock, rather than stock rotation. The board noted that the First Approval Notice should not be read in isolation but in the context of the Second Directive. The board was told (presumably by Gamede) that a stock rotation policy would be submitted to it in January 2016.

[69] The board resolved

(a) that every sale of strategic stock should be back-to-back with a purchase, in the sense that SFF should first identify the purchase opportunity before making the sale;

(b) that the favourable difference between the sale and purchase prices should be enough to cover all incidental costs while still leaving SFF with a margin;

(c) that purchased oil should be of a quality suitable for use in the country’s refineries and in line with its clean fuels policy;

(d) that any purchase or sale had to be preapproved by the board;

(e) that SFF’s Supply Chain Management (‘SCM’) department had to ensure that purchasers, sellers and agents were on the supply database;

(f) that there had to be strict adherence to SFF’s procurement and SCM policies.

With reference to the last point, it was reported (presumably by Gamede) that SFF would get a better price if it followed a negotiation process rather than a tender process, subject always to adherence to SFF’s procurement processes.

[70] In regard to the trading division’s need for specialist skills, it was reported (again by Gamede, one assumes) that SFF already had the skills, as Mayaphi and Nkutha had prior experience as traders. The detailed trading division proposal to be submitted in January 2016 would include a recommendation that Mayaphi be redeployed to head the new division.

[71] It appears from the minutes of this meeting that the clarificatory letter which was to have been sent to the MoE, as discussed at the previous board meeting, had not yet been finalised.

[72] During November, perhaps in light of the board’s resolutions of 23 November, Ngqongwa gave Gamede and Mayaphi a summary of SFF’s procurement procedure. The document deals with the acquisitions of goods and services rather than disposals. In back-to-back transactions such as the board was insisting on, there would be a disposal and an acquisition. According to Ngqongwa, he told them that rotation transactions would need approval from the board, the MoE and National Treasury.

[73] In the latter part of November 2015, the third of the firms to which the strategic stock was sold, Venus, made its appearance. Peculiarly, the first document in the record about this company is an email which Venus’ Mr Lasun Oladeji sent Gamede on 22 November 2015. It seems that Gamede and Oladeji had had previous communication about the form of an RFP to be issued to Venus. On 22 November Oladeji sent an amended draft RFP, which Gamede issued two days later to Venus and to Mbobgeni Investment.

[74] The RFPs issued to Venus and Mbongeni Investment on 24 November 2015 were similar to the expanded RFP sent to Vitol on 20 November 2015, save that these RFPs related specifically to 5 million barrels of Bonny. Proposals had to be submitted by 27 November 2015. Importantly, the following required documents specified in the expanded RFP sent to Vitol were omitted in the list of documents to be submitted by Venus and Mbongeni Investment: (a) audited accounts and documents on bank credit line for the last two years; (b) a tax clearance certificate; and (c) applicable petroleum licenses confirming the handling and exportation of crude oil.

[75] Venus apparently submitted a proposal, but the applicants have been unable to find it. I should mention, here, that unlike Vitol and Taleveras, Venus was not an oil trader. It immediately on-sold its oil to Glencore. Venus had no track record before or after this transaction.

[76] By the end of November 2015, Gamede was in possession of proposals from Vitol, Taleveras, Venus and several other entities. These proposals did not specify a method for determining the price or tender a quantified discount or premium to Dated Brent.

[77] On 30 November Gamede wrote to Venus, stating that its proposal had been found to be sound and acceptable but that, given its size, management had to submit the proposal and its recommendation to SFF’s board and to the MoE for consideration and approval. On the same day, Gamede wrote to the MoE, recommending the Venus proposal. The applicants assume that identical letters recommending the Vitol and Taleveras proposals were sent to the MoE, though they have not been found.

[78] In the letter recommending Venus, Gamede said that in compliance with the First Approval Notice, SFF had received a number of proposals ‘from different local BEE companies’. He sought permission to pursue the Venus transaction on the following conditions: (a) that the transaction would require board approval; (b) that a sale and purchase agreement would be concluded with Venus; and (c) that Venus would provide SFF with a letter of credit from a reputable institution, which SFF would verify with the institution.

[79] On the assumption that identical letters in respect of Vitol and Taleveras were sent to the MoE, Gamede did not tell the MoE anything about the proposals. If Gamede was conveying to the MoE that the three recommended proposals were all from local BEE companies, that was not true. The applicants have no evidence, documentary or otherwise, of how Gamede evaluated the proposals. The applicants say that no other executives were involved.

[80] The exco met on 3 December 2015. Mayaphi presented an updated version of his draft stock rotation policy. The draft said that stock rotation would be conducted via the trading division to be established, details of which would be outlined in a separate document. The recommended rotation quantity was increased from 20% to 30% per cycle. Whereas previously a rotation cycle of three to six months had been proposed, Mayaphi now suggested that the frequency of product return would be agreed ‘on a case by case basis, subject to prevailing and future market conditions’. The new draft contained the same approval processes as before.

[81] The exco decided to set up a team, comprising four members of management together with the board chair, to review the draft policy with a view to presenting a final document to the board. According to the applicants, Gamede did not disclose to his executives that he had already received proposals or that he had recommended them to the MoE.

[82] It seems that Gamede must have written another letter to the MoE on 6 December 2015, requesting approval for ‘rotational sales and purchases of strategic oil reserves’ with Venus, Vitol, Taleveras and GNI/Enviroshore, because on 7 December the MoE, in replying to Gamede, referred to such a letter. In her reply (‘Second Approval Notice’) she granted approval, based on Gamede’s assurance ‘that you have perused the proposals … and that you are satisfied with the sound propositions made’. Approval was subject to the conditions stipulated in Gamede’s letter. It may safely be assumed that these included the conditions contained in Gamede’s letter of 30 November, including SFF board approval.

[83] It is clear from the Second Approval Notice that the MoE did not ask for and read the proposals. It is a fair assumption, also, that Gamede’s letter of 6 December 2015 was, like the earlier letter of 30 December 2015, devoid of detail about the proposals. The MoE did not ask how competing proposals had been evaluated or whether SFF’s procurement and disposal processes had been followed.

[84] Her Second Directive had specified various conditions for rotation transactions:

(a) The first was that ministerial approval needed to be preceded by a detailed due diligence and supported by a comprehensive motivation. (She did not receive this. Due diligence was not done and no comprehensive motivation existed.)

(b) The second was that the strategic stock levels had to be assured at all times. (She was given no information on that score.)

(c) The third was that a trading division had to be established and suitably staffed. She did not ask whether it had been established. (It had not.)

Her Second Approval Notice thus ignored the conditions contained in her Second Directive.

[85] With the MoE’s Second Approval Notice in hand, Gamede wrote to Vitol, Taleveras and Venus on 8 December 2015 stating that SFF, having received authority from the MoE, had approved their respective proposals. Vitol was offered 3 million barrels of Basrah, Taleveras 2 million barrels of Basrah and 2 million barrels of Bonny, and Venus 3 million barrels of Bonny. (At this stage the question of price was still up in the air.)

[86] Save for these quantities, the letters were identical. The successful recipients were told that the stated quantity of crude was being offered ‘for rotation, sale and purchase purposes’ under the following conditions:

(a) the conclusion of a sale and purchase agreement;

(b) a letter of credit from a first-class international bank;

(c) a storage agreement for the offered volumes at $0.13/bbl/m;

(d) product rotation every six months from commencement date;

(e) an option in favour of SFF to buy the product in the tank as and when deemed necessary at a price discounted by $2.

The letter concluded by stating that the offer would become legally binding upon the recipient’s acceptance and signature.

[87] SFF had not in truth approved these offers. The board had resolved that such transactions needed board pre-approval, and Gamede himself had previously specified this as a requirement, though he did not repeat it in the letters of 8 December 2015. The applicants say that the executive team was also unaware of these offers.

[88] Vitol, Taleveras and Venus did not submit written acceptances of the offered terms. Instead there was a process of negotiation resulting in the conclusion of contracts (in each case an SPA and related SGA). Vitol states that it did not even receive the letter of 8 December 2015 and its attachment until 13 January 2016.

The Venus contracts

[89] Since the Venus proposal has not been found, one cannot compare its terms with the contract subsequently concluded. The SFF/Venus SPA and SGA were signed on 15 December 2015. The SPA, which was for 3 million barrels of Bonny, included the following terms:

(a) The date of delivery was 1 January 2016, and the place of delivery was Tank 6. Ownership of the oil would pass to Venus on delivery date.

(b) The price was average Dated Brent for the pricing window of 26-30 December 2016, minus $4/bbl. The price was payable within seven days from the completion of the window period.

(c) Venus acknowledged that the crude oil was part of South Africa strategic reserves. If, at any time before the oil was uplifted from Tank 6, there was a shortage or emergency threatening security of supply and the stock was needed for strategic reasons, SFF would have the option to buy back the crude ‘at a price to be agreed between the parties’ (clause 10). (The discount of $2 which Gamede had specified in his letter of 8 December 2015, and the requirement for six-monthly rotation, did not feature.)

[90] The SPA was subject to various suspensive conditions, including:

(a) that SFF concluded a five-year SGA with Venus for 3 million barrels;

(b) that SFF, Venus and Glencore concluded a tripartite agreement recognising Glencore’s title to the oil and its responsibility to pay the storage fees;

(c) the conclusion of an oil rotation and buyback option in favour of SFF on specified terms.

[91] The SGA had a five-year term running from 1 January 2016 to 31 December 2020. The fixed monthly storage fee was $0.11/bbl escalating at 6% annually (not $0.13, as Gamede had previously specified).

[92] On 29 December 2015 SFF and Venus concluded a first amendment to the SPA and SGA.[9] Among other changes, the delivery date, and the date for passing of title, were brought forward to 30 December 2015 but the pricing window was changed to 4-8 January 2016. Clause 10 of the SPA, giving SFF the right to repurchase the oil in case of emergency, was deleted.

[93] On 31 December SFF and Venus concluded a second amendment by which the SGA was further amended. The commencement date was now 30 December 2015. Venus was given the right to terminate the SGA at the end of each of the first four years by giving notice on or before 30 September of the year in question. By way of a new clause 6.4, SFF had the option, in case of a shortage of energy in South Africa, to buy any oil stored by Venus in Tank 6 at the prevailing market price on terms to be agreed. (This took the place of the deleted clause 10 of the SPA.)

[94] On 1 January 2016 SFF issued a tank warrant to Venus. It is not in the record, but if it was in the same terms as other warrants, it was made conditional on payment of the purchase price.

[95] On 8 January 2016 Venus and Glencore concluded their first SPA. The place and date of delivery matched the amended SFF/Venus SPA. The price payable by Glencore to Venus was average Dated Brent for 4-8 January 2016, plus a premium of $2. (It follows that Venus would make a margin of $6/bbl ($18 million in total), given that it was paying SFF the same average Dated Brent less a discount of $4. At the time, Glencore did not know the price agreed between SFF and Venus.) The agreement contained provisions to ensure that Venus would, upon payment by Glencore, effect payment to SFF. The SPA was subject to various suspensive conditions similar to those to which the SFF/Venus SPA was subject.

[96] In terms of clause 4.7 of the Venus/Glencore SPA read with annexure 2 thereto, Glencore was to pay Venus monthly storage fees of $0.25/bbl. The SFF/Venus SGA, by contrast, only required Venus to pay monthly storage fees of $0.11/bbl escalating by 6% annually. The fact that Glencore, a large global resources trader, was willing to pay a monthly storage fee which was more than double the fee SFF charged Venus is not something which attracted attention in the case.[10]

[97] The tripartite agreement was concluded on 12 January 2016. It recorded that although Glencore was not a party to the SGA, the oil belonged to Glencore and that SFF would issue a tank warrant to Glencore in respect thereof. Glencore would have the same access to the oil as Venus had under the SGA. SFF was not to release the oil without Glencore’s written confirmation, and only Glencore was entitled to give instructions to SFF regarding the oil. Venus was not to exercise its right of early termination without Glencore’s prior written consent.

[98] The tripartite agreement included a buyback option in terms whereof SFF, upon the termination of the SGA, had a first right to buy the stored crude at the prevailing market price on terms to be agreed between SFF and Glencore. If SFF elected not to buy the stored crude, Glencore was to remove it. Following such removal, SFF was entitled to request 3 million barrels of a new quality of crude to be delivered to Tank 6, which Glencore would sell to SFF, again at the prevailing market price on terms to be agreed. In addition to these rights, in case of an energy shortage in South Africa SFF had the option to buy the stored crude at the prevailing market price on terms to be agreed. (Again, the discount to market of $2, specified in Gamede’s letter of 8 December 2015, found no expression.)

[99] On 22 January 2016 SFF cancelled Venus’s tank warrant and issued a new warrant to Glencore, confirming the latter’s ownership. The warrant was conditional on payment of the purchase price by 25 January 2016. Glencore was not satisfied with the warrant, which led to its cancelling the tripartite agreement.

[100] Things got back on track when, on 5 February 2016, Venus and Glencore concluded a new SPA. The pricing window was 8-10 February 2016. The premium remained $2 and the storage fees remained at $0.25. On 8 February 2016 SFF and Venus followed suit, concluding a third amendment to their SPA by which their pricing window changed to 8-10 February 2016, the discount remaining at $4. The next day SFF, Venus and Glencore concluded a new tripartite agreement in terms similar to the previous one. On 9 February 2016 SFF issued a new tank warrant to Glencore.

[101] By this stage there were rumblings within SFF, and Gamede’s free reign was under threat. I shall return to this subject later, but for present purposes I note that on 10 February 2016 Gamede wrote to Venus stating that SFF had to investigate various issues and that he was thus putting the transaction on hold. Venus on 15 February 2016 insisted that SFF comply with its obligations. On 16 February Gamede, one of whose concerns was evidently the selling of Bonny at a $4 discount, responded, indicating that he had in mind amending the SFF/Venus, his suggestion being

‘that we don’t refer to a discount in the contract but craft it in such a way that we look at price windows [that] take a window when the crude was selling at 28 and we sell it to you at Brent plus 3 which gives us 31 and in this way no one will allege any inappropriate conduct and reckless trading.’

[102] On 19 February 2016 SFF and Venus concluded a fourth amendment to the SPA. The pricing window was now specified as being the five-day average starting from the ‘Transfer Date’ plus a premium of $1.66. The ‘Transfer Date’ was defined to mean ‘the date nominated by [Venus] and accepted by [SFF] that fixes the date when all risks and all liabilities with respect to the crude oil shall pass to [Venus] from [SFF]’. This did not yet give effect to Gamede’s improper proposal of 16 February 2016, something which only happened by way of the fifth amendment discussed below.

[103] On 29 February 2016 Glencore and Venus amended their SPA by increasing their premium to $3 (the window of 8-10 February 2016 was unchanged).

[104] On the same day SFF and Venus concluded a fifth amendment to their SPA:

(a) Clause 1.3 purported to set out the market conditions prevailing when the SPA was concluded and to summarise the pricing terms of the original agreement. The summary is patently false (bizarrely, the summary appears to based on the SPA between SFF and Taleveras[11]), and the content of the clause in its entirety is bogus. Its sole purpose was to justify a pricing window of 18-20 January 2016, when Brent oil was at an 11-year low.[12]

(b) Clause 1.3.5 self-righteously declared that Bonny was always sold at a premium with no discount. Based on the average Brent prices over the period 18-20 January 2016, the parties now agreed that the oil would be sold at $27.075 plus a premium of $3, giving a price of $30.075. On the date this fifth amendment was concluded (29 February 2016), Brent closed at $35.97. If the parties had retained the pricing window of 8-10 February 2016 plus a premium of $3, the price would have been around $33.60/bbl (see next paragraph). But the careful selection of the window 18-20 January 2016, coupled with the pricing formula in the amended Venus/Glencore SPA, ensured that Venus would still make a handsome profit.

[105] On 3 March 2016 SFF invoiced Venus $90,225,000 ($30.075 × 3 million). Venus invoiced Glencore $100,761,000, from which one can infer that the pricing formula in the amended Venus/Glencore SPA yielded a price of $33.587 (ie average Dated Brent for 8-10 February 2016 of $30.587 plus a premium of $3). These invoices were settled on 4 March 2016.

[106] Venus thus made a profit of $10,536,000 ($3.512/bbl) by immediately on-sold the oil to Glencore. Venus disappears from the scene, except that each month it invoiced Glencore for storage fees of $0.25/bbl, pocketing the difference between this amount and the storage fees of $0.11/bbl which SFF was charging Venus. The practical result was that over the period January 2016 to September 2017 Venus paid SFF storage fees totalling $7,458,371,[13] while Glencore paid Venus storage fees totalling $17,200,075. In this way, Venus made an additional profit of $9,741,704 million over about 20 months.

The Taleveras contracts

[107] Taleveras’ proposal of 4 November 2015 was that it would ‘rotate’ part of the strategic stock in Tanks 2 and 6. The initial term would be three years, starting on 15 December 2015, renewable for another three years. Taleveras would buy the oil at an agreed value, ‘bearing in mind the quality depreciation level’. Taleveras would have a right of first refusal to supply any replacement crude that was exported. Taleveras would pay SFF monthly storage fees for the duration of the contract. This was all pretty vague.

[108] The SFF/Taleveras SPAs were concluded on 28 December 2015. For unexplained reasons, the SGAs were concluded earlier, on 15 December. The SPAs are governed by English law and subject to the jurisdiction of the English courts while the SGAs are governed by South African law and subject to the jurisdiction of this court.

[109] There were separate SPAs for 2 million barrels of Bonny and 2 million barrels of Basrah. Save for the pricing clauses, the terms were similar to those of the SFF/Venus SPA, including clause 10 (SFF’s right to buy the stored product in case of shortage of supply). The delivery date was an unspecified date in January 2016. The pricing clauses stipulated that the final price would be agreed after an independent inspector’s certificate of quantity and quality. The final price would include a discount as agreed, and was to be calculated with reference to the average of Dated Brent for three consecutive days following the inspector’s certificate. (On the face of it, the pricing clauses rendered the SPAs void for vagueness, at least as judged by our law.)

[110] Unlike the SFF/Venus contracts, the two SFF/Taleveras SPAs were concluded simultaneously with a back-to-back purchase agreement (‘BPA’) in terms whereof Taleveras sold to SFF 4 million barrels stored in Tank 1. The delivery date and pricing in the BPA were the same as in the SPAs. In clause 10, SFF granted Taleveras ‘the right to utilise the oil for commercial benefits to benefit from the contango’, in consideration for which opportunity Taleveras was to pay SFF storage fees ‘for 4 million barrels purchased by SFF and stored in Tank 1 as strategic stock’.

[111] The SGAs ran from 1 January 2016 to 31 December 2020. The terms were the same as the Venus SGA, save that the fixed monthly fee was $0.13/bbl (the rate specified in Gamede’s letter of 8 December 2015). By way of an amendment of 13 January 2016 the fee was reduced to $0.12/bbl. On 16 January 2016 SFF issued tank warrants to Taleveras in respect of Tanks 2 and 6. The inspector’s certificates were issued on 17 January 2016, so that the pricing window in the SPAs was 18-20 January.

[112] The SPAs were varied by a first amendment signed on 29 January 2016. The price of the oil was now fixed at $30 for Bonny and $26 for Basrah. There is nothing to indicate that these prices were determined with reference to the pricing window of 18-20 January. On the same date Taleveras and SFF concluded an SGA relating to Tank 1.

[113] By early February 2016 Taleveras had arranged to finance its purchase of the oil with CTSA/Natixis. On 4 February 2016 SFF, Taleveras and CTSA executed a side letter which recorded that CTSA would be buying the oil from Taleveras in terms of a master sale and purchase agreement (‘MPA’). SFF approved the transfer of title. SFF and Taleveras acknowledged CTSA’s title. CTSA was not, however, a party to the SGA, and Taleveras remained liable for storage fees. On the same day, SFF issued tank warrants to CTSA and issued invoices to Taleveras for $60 million (2 million barrels of Bonny at $30/bbl) and $52 million (2 million barrels of Basrah at $26 million).

[114] The MPA, governed by English law, was concluded on 8 February 2016. It made provision for a facility of $165 million over 27 months, i.e. until early April 2018. This facility was determined with reference to Taleveras’ acquisition of the 4 million barrels from SFF. The MPA stipulated that Taleveras could avail itself of the facility by delivering sale and repurchase confirmations, which Taleveras did on 10 February 2016 by way of confirmations styled as Amendment 1 to the MPA.

[115] The MPA as read with the confirmations had the following practical effect:

(a) CTSA bought and became owner of the 4 million barrels on the ‘Settlement Date’, namely 8 February 2016.

(b) Taleveras simultaneously undertook to buy back the oil on the ‘Repurchase Date’, namely 5 April 2018 (the expiry of the 27-month facility).

(b) The total ‘Sale Price’ for the oil bought by CTSA was $179,920,000, calculated at a ‘Forward Price’ of $44.98 for 4 million barrels. The ‘Forward Price’ was based on the forward curve, as at February 2016, for Brent delivery in March 2018. (So whereas Brent closed at $32.88 on 8 February 2016, the average ‘Forward Price’ at that time for delivery in March 2018 was $44.98.)

(c) The net amount actually payable by CTSA was $134,168,426, arrived at by deducting the ‘Total Haircut’ and the ‘Transaction Cost Payment’ from the ‘Sale Price’.

(i) The ‘Total Haircut’ was $15,597,600, comprising (aa) $5,397,600, being 3% of the ‘Sale Amount’; and (bb) $10,2 million, being 4 million × $2.55, the latter being the defined ‘Differential’ (the net discount/premium between Dated Brent and Bonny and between Dated Brent and Basrah, determined in accordance with a specified formula).

(ii) The ‘Transaction Cost Payment’ was $30,153,974, comprising (aa) $10,776,811, being the ‘Treasury Fee’ and ‘Treasury Premium’; (bb) $2,816,800, being the ‘Hedging Premium’; (cc) $800,000, being the ‘Offtake Premium’; (dd) $1,314,579 as ‘Flat Transaction Fees’; (ee) $485,784 as insurance costs; (ff) $12,960,000 in respect of the storage fees for which Taleveras would be liable to SFF but which CTSA would pay; and (gg) $1 million, being the defined ‘Provision’.

(d) Of the net amount of $134,168,426, $112 million funded the price which Taleveras had to pay SFF, leaving Taleveras with $22,168,426. (The settlement of these payments took place on 25 February 2016. In March 2016 CTSA released $400,000 of the provision of $1 million, which meant that it had paid out a net amount of $134,568,426 and that Taleveras was left with a net amount of $22,568,426.)

(e) The price at which Taleveras was required to buy back the oil on 5 April 2018 was the ‘Repurchase Amount’, defined as the ‘Repurchase Price’ (i) less any costs and expenses in relation to the transaction paid by CTSA or any member of its group, such costs to be reimbursed by Taleveras on the Repurchase Date on presentation of valid invoices; (ii) plus the balance of the Provision’ (if any, after deducting the foregoing costs).

(f) The ‘Repurchase Price’ was defined as the prevailing market price for the oil on 5 April 2018. (This would be the actual spot price on 5 April 2018. This could turn out to be more or less than the ‘Forward Price’. In the event, the spot price on 5 April 2018 was well above the ‘Forward Price’ – Brent closed at $66.54 on 5 April 2018.)

[116] SFF and Taleveras concluded a second amendment of the SPA on 22 February 2016. It contained the recordal and terms which later served as Gamede’s bizarre model for the fifth amendment to the SFF/Venus SPA. The only practical effect was to increase the price for the Bonny from $30 (as per the first amendment) to $30.075. Its main purpose seems to have been to dress up this price as representing a $3 premium above Dated Brent. Although the pricing window of 18-20 January 2016 accorded with the original Taleveras SPAs concluded on 28 December 2015 (ie the three-day period following inspection and certification), Taleveras did not have financing for the transaction at that time, and the original SPAs were in any event hopelessly vague as to price. So much had changed by February 2016 that the pricing window initially stipulated seems to have lost any sensible connection with the transaction as finally agreed and implemented.

[117] In the event, when SFF again invoiced Taleveras for the oil on 25 February 2016, the Bonny was charged at the old price of $30, and that is the price which SFF received the same day. If the second amendment was intended truly to take effect, there was a short-payment for the Bonny of $150,000.

[118] On 24 June 2016 SFF and Taleveras concluded a novation agreement by which the BPA (in terms whereof Taleveras sold 4 million barrels of oil to SFF) was rescinded and replaced with a new contract. Evidently no effect had in the meanwhile been given to the BPA. The novation agreement made provision for Taleveras to sell finished petroleum products to SFF in lots of 15,000-30,000 tons. This novation was itself terminated a few days later, thus reinstating the BPA.

[119] To conclude this section, I should mention other transactions relating to CTSA/Natixis’ financing of Taleveras:

(a) In accordance with Natixis’ requirements, Charmondel Holdings Ltd (‘Charmondel’), Taleveras’ parent company, issued a guarantee for Taleveras’ obligations to CTSA.

(b) Natixis issued an irrevocable documentary credit in SFF’s favour for the purchase price due by Taleveras to SFF.

(c) In order to hedge its exposure to the actual oil price on the MPA’s ‘Repurchase Date’ of 5 April 2018, CTSA concluded a commodity swap with Natixis GMC with an expiry date of 6 April 2018. The swap rate replicated the MPA’s forward price of $44.98, while the float rate replicated the repurchase price formula. (This protected CTSA against the risk that on the 5 April 2018 the spot price of the oil was materially less than $44.98.) Natixis GMC in turn concluded a back-to-back trade with an external counterparty at exactly the same swap rate and with the same expiry date.

(e) SFF’s monthly invoices to Taleveras for storage fees were passed on to CTSA which effected payment, a state of affairs which continued until January 2018.

The Vitol contracts

[120] Vitol’s proposal of 16 October 2015 was that it be granted the option to rotate a maximum quantity of 5 million barrels stored in Tanks 2 and 6, with delivery by SFF as from 1 November 2015. It would provide a first-class international bank guarantee for the value of the rotated crude. Vitol would restore the exact same quantity and quality of oil within 36 months, subject to any strategic emergency. Vital would cover all costs associated with the rotation, and pay SFF a fixed fee to be agreed for each cycle. This was repeated, along with various other proposals (not now relevant), in Vitol’s letter of 29 October 2015. Vitol made a further proposal on 1 December 2015, but it is not in the record and its content is not disclosed.

[121] On 8 December 2015 Gamede offered Vitol 3 million barrels of Basrah for rotation. On 13 January 2016 Vitol accepted the offer subject to further terms which had in the meanwhile been agreed. Vitol’s letter identified Vesquin as the buyer.

[122] The SFF/Vesquin SPA and SGA were concluded on 20 January 2016. The SPA was based on a Vitol draft and thus differed from the Venus and Taleveras SPAs. The delivery date was not later than 31 January 2015. The price was average Dated Brent, the pricing window being the five days following delivery, less a quality discount of $8. Payment would be made within 30 days of invoice.

[123] In terms of clause 5.2, SFF had the right to borrow any sold product still in storage at the relevant time, or its equivalent if no longer in storage, but only in the event of an emergency and after exhausting any other strategic stocks. In that event, SFF had to replace the borrowed stock at its own cost within 30 days.

[124] In terms of clause 9 of the SPA, Vesquin/Vitol was to sell back to SFF 3 million barrels of oil conforming with the quality specifications contained in the SGA. This resale would take place on the termination of the SGA or at an earlier date if Vesquin/Vitol gave 30 days’ notice thereof. (This placed the timing of the resale in the hands of Vesquin/Vitol rather than SFF.) The price was Dated Brent, with a window period of five days after delivery in case of an ITT (there was a different window if the transaction was CFR[14]), less a quality discount of $8.25.

[125] The SGA was in standard form. It had a three-year period expiring on 31 January 2019. The monthly fee, as with Venus, was $0.11. On 17 May 2016 the SGA was amended in respects not now material.

[126] A dispute then developed between Gamede and Vitol, because Gamede wanted the pricing window to be an earlier period, when Dated Brent was higher. Vitol’s view was that it had been willing to conclude the transaction earlier but SFF had delayed. It was not Vitol’s fault that the SPA was only concluded on 20 January 2016, when Brent was at an 11-year low. (The applicants’ expert, Mr Driscoll, supports Vitol on this point. A retrospective pricing window is, he says, ‘objectionable and unacceptable within the oil trading community’.[15])

[127] On 22 January 2016 SFF and Vesquin concluded an amendment of the SPA. The delivery date became 22 January 2016, with the pricing window to be 25-29 January 2016. Vitol/Vesquin was thus able to resist the use of a more expensive retrospective pricing window. They did, however, appease SFF by reducing the discount to $5.50. In line with this change, the discount in the resale purchase price was reduced to $5.75. Vitol/Vesquin’s sacrifice was to some extent offset by the fact that the SGA was simultaneously amended to reduce the monthly fee from $11 to $10.

[128] Average Dated Brent for the amended pricing window was $31.702, yielding a price of $26.202 after deducting the discount. It is unclear why it took SFF until 29 February 2016 to invoice Vesquin $78,606,000, based on this price. Vesquin made payment on 11 March 2016. Regarding itself as the owner of 3 million barrels of Basrah in Tank 2, Vitol paid monthly storage fees until February 2018.

[129] Vitol also concluded hedging transactions. These were of two kinds. First, on each of the five days in the pricing window 25-29 January 2016 (the ‘pricing in’ period), Vitol sold 600,000 barrels of short-term ICE[16] Brent Futures (totalling 3 million barrels over five days), which it bought back as soon as pricing window expired. Second, it then immediately sold 3 million barrels of longer-dated ICE Brent Futures with settlement dates corresponding with the expiry date of the SGA (31 January 2019). Although Vitol could elect to uplift the oil earlier, it would naturally do so only if it was commercially favourable. However, by 31 January 2019 Vitol had to uplift the oil. If at that time the market price was below what it had paid SFF at January 2016 prices, Vitol would sustain a loss.

Further events up to launching of application

[130] During the time the contracts were being concluded, and thereafter, Gamede’s activities started to come under scrutiny. Mayaphi, who seems to have known more about what was going on then he has cared to acknowledge in his confirmatory affidavits for the applicants, wrote a memorandum to Gamede on 13 January 2016 expressing concerns about implementing strategic stock sales. He did not refer to any specific transactions. He said that in terms of the CEF Act, the proceeds from oil disposals would need to be paid into, and the purchasing of oil funded from, the Equalisation Fund referred to in that Act, in the latter case with the concurrence of the MoF.

[131] He also considered that the disposals fell within the ambit of s 54(2)(d) of the PFMA as being disposals by a public entity of a ‘significant asset’. Before concluding such a contract, the entity’s accounting authority (here, SFF’s board) must promptly and in writing inform the relevant treasury (here, the National Treasury) of the transaction, and submit relevant particulars of the transaction to its executive authority (here, the MoE) for approval.

[132] Gamede did not respond to Mayaphi’s memorandum. National Treasury was not notified of the disposals. Sale proceeds were not paid into the Equalisation Fund. SFF’s board was not asked to seek approvals from the MoE.

[133] The minutes of the exco’s first meeting of the new year, held on 19 January 2016, show that the strategic stock policy was not yet finalised. Gamede and Mayaphi submitted a memorandum on the proposed trading division in order to give effect to the Second Directive and First Approval Notice. The meeting agreed to recommend to the board an organisational restructuring in terms whereof Mayaphi would become General Manager: Trading Division, into which would be incorporated the existing Commercial Division.

[134] Although, remarkably, the exco minutes do not reflect this, the applicants say that Gamede told the meeting that he had already sold the 10,3 million barrels of strategic stock. The other executives were ‘shocked’. Ngqongwa immediately asked for the commercial terms of the sales but got no response. The allegation that none of the other executives knew of the disposals has not gone unchallenged. Vitol states that at least Mayaphi and De Wet knew about them. I shall need to return to this question.

[135] The first SFF board meeting of the year was held on 27 January 2016. This was the first meeting chaired by Jawoodeen. Present by invitation were the senior executives. They reported to the board that the rotation and trading policy would be incorporated into the corporate plan and presented at the board’s next meeting. It was also reported that the clarificatory letter about funding had been sent to the MoE (this letter is not in the record). The proposed restructuring, including Mayaphi’s appointment as General Manager: Trading Division, was approved.

[136] In a staggering conspiracy of silence, Gamede, Mayaphi, Ngqongwa and Nkutha did not tell the board that the strategic stock had been sold. Given what the applicants have said about Gamede’s sway, one or more of the others may have been cowed into submission, but this is hardly an excuse.

[137] The board met again on 5 February 2016, and again the senior executives were present. The directors were now told about the contracts, though precisely what they were told is unclear. According to Ngqongwa, board packs were only handed out at the beginning of the meeting and there were not enough copies for all attendees. One of the directors, Mr Vilakazi, complained that he had not received the pack and had to rely solely on the presentation made at the meeting. Jawoodeen states that the contracts were not included in the board pack and that the board did not call for them. He claims to have expected the agreements ‘to include terms that are standard in the oil industry’.[17] Ngqongwa states that the exco members saw the contracts at the meeting venue (he does not say that the directors saw the contracts), and they began analysing them afterwards.[18]

[138] According to the minutes, the directors received a trading report but the applicants have not produced it. What the minutes record as having been reported to the board is: that SFF had received approval from the MoE to sell and repurchase the 10,3 million barrels of strategic stock; that ‘projected revenues would be based on the various discounts offered as per the quality of the material’; that Vitol, Venus and Taleveras ‘were three companies considered for the stock rotation purposes’; and that these three companies ‘were expected to present the financial guarantee to SFF to secure the given quantities by 31 January 2016’. Following this report, the board agreed that ‘the Vitol and Taleveras transactions be approved’. Venus would be given 30 days to perform, failing which its contract would be terminated.

[139] In the context of Venus, the minutes refer to an SPA and an SGA, so I think one can infer that the board was aware that similar contracts had been concluded with Vitol and Taleveras. The applicants say that the board made no enquiry about the disposal process, the content of bids, evaluation and adjudication criteria or compliance with regulatory provisions.

[140] On 8 February 2016, Nkutha, having to some extent found his voice, submitted a memorandum to Gamede with copies to Mayaphi and Ngqongwa about the Bonny SPAs. He described the ‘deal economics’ of the contracts as ‘sub-optimal’, since the Bonny was being sold at a discount to Dated Brent of $4 (this discount was explicit in the Venus SPA as it then read, and was implicit in the Taleveras pricing), whereas benchmarks suggested a premium of $1.70. He also said that a market price for an ITT of Bonny in Saldanha Bay should take into account that ordinarily a buyer wanting to store Bonny at Saldanha Bay would bear the freight cost of getting it from Nigeria to South Africa.

[141] He also made the obvious point that SFF’s buyback right exposed it to ‘flat price risk’, ie changes in the spot price of oil. For every increase of $1.00 in the market price of Bonny in comparison with the disposal price, SFF would suffer a loss of $5 million if it wished to replenish the oil.

[142] He thus recommended that the terms of the Bonny transactions be renegotiated. Although Nkutha’s memorandum was marked strictly confidential and not for distribution, Gamede sent it to Venus and Taleveras. His formal position in this correspondence was that the memorandum raised serious issues and that if they were not carefully considered there could be a perception that the transactions were reckless and irregular. He had thus decided to review the issues and refer them to the board, and in the meanwhile to place the transactions on hold. He also notified the board chair of the action he was taking.

[143] Gamede’s later conduct shows that this was not his real intention. On the contrary, the way he ‘solved’ the problem of discounted Bonny pricing was to conclude SPA amendments with Venus and Taleveras which dressed up the price as being at a premium to average Dated Brent in a window period selected for no other reason than that it represented Dated Brent’s all-time low. SFF’s pricing risk under the repurchase terms was left unaddressed.

[144] The side-letter agreement and tank warrants required SFF to copy CTSA on all notices and other communications sent to Taleveras. Gamede failed to do so, as a result of which CTSA remained in ignorance of the potential difficulty. This remained the state of play on 25 February 2016 when CTSA caused $112 million to be paid to SFF for the Taleveras oil. Gamede also failed to notify CTSA of the amendment to the Bonny SPA on 22 February 2016. If SFF had complied with its notification obligations, CTSA would in all likelihood not have implemented the MRA or paid SFF, at least not until the concerns raised in Nkutha’s memorandum had been resolved to CTSA’s satisfaction. This very fact may explain why neither Gamede nor Taleveras informed CTSA of developments.

[145] The applicants allege that CEF Treasury, which managed SFF’s treasury function, became aware of the sale of the oil reserves on 26 February 2016, upon notification from Absa that SFF had received $112 million. However, according to the applicants’ replying papers, the treasury official in question was Ngqongwa. He had learnt of the disposals, at least in general terms, by not later than 19 January 2016.

[146] On 26 February 2016 Gamede sent the MoE copies of contracts concluded with Taleveras and Venus. There was no accompanying explanation.

[147] In April 2016 Gamede wrote to the MoE, requesting condonation for non-compliance with regulatory requirements. The applicants say that Ngqongwa drafted the request. Gamede told the MoE that SFF, in its ‘haste’ to ‘enhance’ its ability to meet its mandate of providing the country with security of supply, had ‘missed a few regulatory processes’, for which SFF was ‘deeply remorseful’. Condonation was needed regarding (a) non-compliance with s 54(2)(d) of the PFMA; (b) non-compliance with delegations of authority subsisting between CEF and SFF, in particular the need for CEF board approval in case of any ‘material changes in [SFF’s] strategic direction or material deviations in [SFF’s] business plans’; (c) the non-payment of the disposal proceeds into the Equalisation Fund.

[148] The implication that these three requirements had been overlooked was untrue. Mayaphi had made two of these points in his memorandum of 13 January 2016, at a time when the SFF/Vitol contracts had not yet been concluded and when various amendments to the Venus and Taleveras contracts lay in the future. Furthermore, SFF had not received any disposal proceeds as at 13 January 2016. The implication that ‘haste’ had been needed was also untrue. If there was reason to sell the strategic stock at all, it was not urgent.

[149] In her budget speech delivered on 11 May 2016, the MoE referred to her Second Directive, stating that its implementation had resulted in increased revenue for SFF while still maintaining stocks within SFF’s storage tanks for security of supply. This was in place ‘through long term lease and contractual arrangements with the buyers’. Rotation had allowed SFF ‘to replace the unsuitable stock’ it had been storing. It is thus clear that the MoE had knowledge of the disposals, though what she knew of the terms is unclear. Her reference to ‘unsuitable stock’ seems to have been based on what Gamede had told her.

[150] In the latter part of May 2016 the disposal of the strategic stock began to attract media attention. National Treasury, having been alerted by these reports, set up a meeting with officials of the DoE, SFF and CEF on 1 June 2016. SFF undertook to supply National Treasury with supporting documentation.

[151] The media reports prompted CEF to issue a statement on 7 June 2016. CEF justified the disposals on grounds similar to those Gamede had given to the MoE. ‘Rotation’ was said to be advantageous, because it allowed SFF to generate substantial revenue from storage. If SFF’s oil had stayed in the tanks, it would have lost volume at about 1% annually. The disposals had been based on ‘transparent market related price formulae’. (The applicants, while alleging the media statement to be factually incorrect, do not say that it were not issued with CEF’s authority.)

[152] The media reports came to CTSA’s attention. This was its first inkling that the transactions might be questionable. They contacted Taleveras, and received the reply that Gamede would be travelling to Geneva to clarify matters. On 10 June 2016 Gamede, Mayaphi and a third SFF official met CTSA/Natixis’ representatives in Geneva. Gamede told them that the media reports were fuelled by political rivalry and that CTSA’s ownership was not at risk.

[153] Prompted perhaps by the condonation request and media attention, the MoE called on SFF’s board chair to give feedback on compliance with her Second Directive. Jawoodeen replied on 23 June 2016, attaching three Gamede reports which contained the following assertions:

(a) Gamede stated that the country’s biggest risk was not a shortage of crude but a shortage of refined products.

(b) SFF’s crude oil, or at least the Basrah, had to be rotated as part of the move towards clean fuel. South African refineries were no longer using Basrah. (This justification did not apply to the Bonny. The applicants have obtained evidence from two local refineries contradicting this statement in relation to Basrah. The applicants’ expert, Mr Ara Barsamian, a chemical engineer and crude oil blending expert, says that there is no such thing as a ‘bad’ crude oil – blending to achieve desired qualities is a common practice.[19])

(c) The crude had degraded over the years. (Mr Barsamian shows this claim to be false. Bonny has an almost infinite degradation-free life in underground storage with low sludging propensity.[20] The Basrah had been mishandled because crude oils of unknown quality were placed on top of it. While sludging was possible, SFF’s Basrah quality report of December 2015 provided no evidence of deterioration or propensity to sludge. A full assay, which SFF never undertook, would have been needed to assess the Basrah’s quality.[21] Assays were done very recently, based on samples drawn on 14 July 2020, and these confirmed no deterioration for the Bonny or Basrah.[22] The respondents themselves were clearly not worried about degradation. In the execution of their contango strategy, they were content to let the oil remain in the tanks for several years until it was opportune to sell.)

(d) The crude was suffering a volume loss of 10% annually. (Mr Barsamian again refuted this claim. Crude oil stored underground, he says, does not reach high enough temperatures to allow for volume loss through vaporisation. The CEF media statement said 1% and Foster thought 0.35%.[23] In another memorandum, Gamede himself said 1%.)

(e) The pricing formula for the Bonny disposals had been based on market conditions with reference to an appropriate three-day pricing window and a premium of $3. The 18-20 January 2016 window had been chosen because the quality inspection certificate was issued on 17 January 2016. (I have already explained the disreputable reasons for the selection of that pricing window. The certificate of 17 January, which applied to Tank 2 and Tank 6, ie to the Basrah as well as the Bonny,[24] had no rational connection to the appropriate pricing window.)

(f) The disposals, effected in a contango market, enabled SFF to earn storage fees for its full 45 million-barrel capacity, while at the same time having access to crude oil if there were an emergency. (Such access was on terms which exposed SFF to the price risk of the oil at the time of the emergency.)

(g) Gamede stated that SFF’s repurchasing of oil ‘will occur when the prices are low, as this is the reflection of surplus production in the crude oil market and allows SFF to make a further margin’. (This might be true for elective replenishment but not for emergencies. In an emergency, prices would more probably be high because of disrupted international supply. The strategic stock was held to cover emergencies.)

(h) Gamede described an extensive process of investigation and deliberation supposedly involving the full SFF management team. (The evidence does not support these allegations.)

(i) Management chose a process of negotiation without tendering, which CEF’s procurement policy and National Treasury regulations permitted when competitive tendering was not suitable. SFF had always used the negotiation process in its storage contracts. (The evidence in this case shows that oil disposals are often done by competitive tendering. A tender process was not unsuitable.)

(j) He claimed that management took into account various legal and policy instruments, including the PFMA and SFF’s procurement policy. SFF’s considered view was that s 54(2)(d) of the PFMA was not applicable to ‘operational assets’. (This seems not to have been his belief when seeking condonation from the MoE in April 2016.)

(k) He said that he had asked the CEF Internal Audit Department to audit the disposals. The main findings were that the process had been in line with the Second Directive and that there was no violation of the PFMA.

[154] After engagement between with the MoF, the MoE instructed CEF to conduct a legal review of all contracts concluded by SFF since 2014 relating to the sale or storage of crude oil reserves. Attorneys Allen & Overy (‘A&O’) were mandated to do so. The applicants say that this instruction was initially met with resistance and a lack of organisation.

[155] Gamede and Jawoodeen resigned at the end of June 2016. On 5 July Nkutha wrote to Vitol, stating that Gamede’s resignation would have no material impact on SFF’s daily operations: ‘The internal policies and processes designed to safeguard your crude oil currently stored in our Saldanha Bay terminal are still effective and are unaffected by the changes announced last week.’ (This may have been a standard letter to all customers for whom SFF was storing oil. Vitol was not told that a legal review of SFF’s contracts was afoot.)

[156] In late August 2016 National Treasury, having reviewed the information supplied by SFF, concluded that the transactions were inconsistent with s 1A(3A) of the CEF Act and s 54(2)(d) PFMA. The MoF notified the MoE of these findings with recommendations for remedial action.

[157] Despite these developments, CEF and SFF continued to defend the transactions in the second half of 2016. In CEF’s annual group report for the year ended 31 March 2016, the following statement was made in connection with SFF’s operations:[25]

‘10 million barrels of crude oil stock was sold based on a transparent market-related price formula. The sold stockpile still remains in tank at the SFF Saldanha terminal with SFF having the first right to buy the crude oil and supply the market in the event of a crisis. The Rotation of the Strategic Stock is financially and economically advantageous to government, as SFF will generate more than US$15 million on the storage or R180 million per annum over a period of five years. The alternative economic value proposition of the stock in tank, is that it could lose about 1% of the stock per annum.’

[158] A&O completed their investigation in December 2016. Based on A&O’s findings, CEF concluded that the contracts for the disposal of the strategic oil were liable to be set aside on review. CEF sought senior counsel’s opinion. Counsel was briefed in December 2016. He furnished his opinion on 10 February 2017.

[159] In the meanwhile, CTSA representatives were in South Africa in January 2017 to conduct a due diligence of the maintenance conditions at the Saldanha Bay tanks. Afterwards they met at SFF’s Cape Town offices with Ngqongwa, Mayaphi and SFF’s Financial Manager, Cynthia Beukes. CTSA again sought clarity on the status of the transactions. They were assured that CTSA held valid title to the 4 million barrels. They were told, however, that any request to uplift the oil would have to be submitted for approval to CEF and the MoE. Because of Gamede’s unauthorised conduct in an unrelated matter, there was heightened governance, and there was a review of all storage contracts. CTSA was not told that judicial review proceedings were in contemplation or that the oil would not be released on request.

[160] On 8 February 2017 Vitol notified SFF that it intended to export some of its crude in Tanks 2 and 3, nominating export slots during March-May 2017. (The oil in Tank 3 was not part of the impugned transactions.) Vitol alleges that SFF initially refused on the basis that the oil was ‘inaccessible’. When Vitol gave another notice on 19 April 2017, SFF again prevaricated.

[161] On 6 and 9 June 2017 SFF notified Taleveras, Vito and Venus that the MoE had instructed CEF to commission a legal review of the circumstances surrounding the sale of the strategic oil. The review had been concluded ‘and certain findings and recommendations are currently being considered by SFF, [CEF] and the [MoE]’. In the meanwhile, CEF and the MoE had instructed SFF that the oil should not be removed. (This letter did not disclose that the applicants had already decided to institute court proceedings.)

[162] On 22 June 2017 CEF mandated KPMG to undertake a financial analysis of the transactions. KPMG’s brief included an instruction that 1,2 million of the 5 million barrels of Basrah in Tank 2 was ‘unpumpable’, meaning that SFF would have had to buy this crude in the market to fulfil the Taleveras and Vitol transactions. KPMG issued its report in July 2017. CEF sought a second senior counsel’s opinion in the light of the financial analysis, which counsel furnished on 27 July 2017.

[163] In September 2017, by which time the applicants had taken no further steps, the market shifted from contango to backwardation. From Glencore’s perspective, this was an ideal time to exercise its right to terminate the Venus SGA. Notices to this effect in respect of Tank 2 and Tank 6 were delivered on 14 September 2017. (SFF’s letter to Venus of 9 June 2017 had not come to Glencore’s attention.)

[164] On 12 September 2017 SFF’s acting CEO, Mr Thabane Zulu, wrote to CTSA in light of previous meetings, stating that the boards of CEF and SFF were optimistic that solutions could be reached beneficial to CTSA and SFF. This optimism was dashed when, on 26 September 2017, SFF addressed identical letters to Taleveras, Vitol and Venus, stating that the transactions were invalid and that it was thus in the process of preparing a review application. In the meanwhile, no strategic stock could be uplifted. (It is unclear what prompted the timing of these letters. The conclusion of invalidity had been reached by A&O in December 2016, and there is no indication that counsel in their opinions of February and July 2017 disagreed.)

[165] Unaware of the above letter, Glencore wrote to SFF on 26 September 2017, repeating its intention to terminate the Venus SGA as at 31 December 2017. Glencore asked whether SFF wished to exercise its first right to buy the oil, failing which Glencore would make immediate arrangements to sell it to a third party. The next day Venus forwarded SFF’s letter of 26 September 2017 to Glencore.

[166] In the days following SFF’s letters of 26 September 2017, Vitol, Glencore and CTSA wrote to SFF, rejecting the outcome of the legal review and asserting their ownership of the oil.

[167] On 6 October 2017 Vitol notified SFF that it intended withdrawing its Basrah from Tank 2, and nominated loading dates in the first half of November 2017. Vitol told SFF that its conduct in refusing to allow Vitol to uplift the oil was a breach of the agreements, in respect of which it reserved its rights.

[168] On 12 October 2017 CEF engaged PwC to undertake a further financial analysis, its mandate being broadly similar to KPMG’s. The applicants say that this was done because KPMG was attracting unfavourable public attention due to alleged unethical conduct in relation to another public entity’s affairs. PwC issued its report on 7 November 2017.

[169] A&O withdrew as the applicants’ attorneys in November 2017 due to a conflict of interest. Their place was taken by Cliffe Dekker Hofmeyr (‘CDH’). Without-prejudice meetings held during November and December 2017 did not bear fruit. By this stage Werksmans were on record for Glencore and Norton Rose Fulbright (‘NRF’) for CTSA and Natixis. After the institution of the review, Herbert Smith Freehills (‘HSF’) came on record for Vitol.

Procedural history of review application

[170] The applicants launched their application on 12 March 2018. Part A of the notice of motion sought the review and setting aside of the relevant decisions and transactions. Part B, concerned with just and equitable relief, was to be decided later. No specific relief was sought in that regard. The applicants (somewhat oddly for a self-review) tendered to file their record within 15 days, and called on the MoE to do likewise. The founding affidavit mentioned, but did not attach, the ‘legal review’ (ie A&O’s review), senior counsel’s opinions and the KPMG and PwC reports.

[171] The MoE did not file a record, claiming to have nothing to add to the documents attached to the founding papers. The applicants themselves filed their record on 14 May 2018, some days late. On 29 May 2018 the applicants delivered an amended notice of motion and a short supplementary affidavit.

[172] CTSA and Glencore served notices requiring the applicants to disclose various documents mentioned in the founding papers, including the legal review, opinions and accounting reports. When the applicants failed to make disclosure to their satisfaction, CTSA and Glencore launched applications to compel. The applications were argued before Saldanha J on 1 November 2018. He delivered judgment on 29 January 2019, ordering disclosure in some but not all respects. He refused to order disclosure of the legal review, legal opinions and accounting reports.

[173] CTSA and Glencore applied for leave to appeal, which Saldanha J granted on 30 April 2019. The disclosure appeal was argued in the Supreme Court of Appeal (‘SCA’) on 26 November 2019.[26] The SCA delivered judgment on 13 December 2019, holding that the ‘legal review’ mentioned in the founding papers was not a reference to a ‘document’; that counsel’s opinions were legally privileged; but that the applicants had to disclose the KPMG and PwC reports. The applicants complied with this order.

[174] In the meanwhile, the applicants, despite having already launched the review application, decided to commission a forensic investigation into the impugned transactions. On 7 May 2018 CEF issued a tender for the provision of this service. On 27 July 2018 Gobodo Forensic and Investigative Accounting (Pty) Ltd (‘Gobodo’) was appointed. Gobodo issued preliminary reports in September and October 2018. Gobodo interviewed Gamede on 15 November 2018 and Mr Joemat-Pettersson on 10 April 2019. It issued its report on 30 April 2019. The report was considered and approved by CEF’s board on 25 June 2019.

[175] The respondents only learnt of the Gobodo report when it was mentioned in the applicants’ supplementary founding papers delivered on 28 February 2020. To explain the filing of these supplementary papers, I must retrace my steps. Because the parties could not agree on the conduct of the case, in particular the bifurcation of merits and remedy, the matter became the subject of judicial case-management by the Judge President in August 2018. On 28 August 2018 he ordered the consolidation of Parts A and B, and enrolled the consolidated case for hearing on 26-28 March 2019. The applicants were to file any further supplementary founding papers by 15 October 2018.

[176] By early October 2018 it emerged that the applicants were not going to comply with the timetable. CDH intimated, though, that significant developments as a result of a further forensic investigation would be revealed in due course. The parties appeared before the Judge President on 17 October 2018. The respondents’ lawyers wanted the application struck from the roll because of delay. The Judge President directed the applicants to report back to him and the respondents on 7 November 2018 about progress in finalising their investigations.

[177] At the meeting of 7 November 2018, CDH reported that the applicants’ forensic investigation was advanced but interviews with certain key figures were outstanding. The respondents placed on record that they were suffering severe financial losses because of delay. The Judge President ordered the applicants to file their supplementary papers by 21 November 2018, even if the forensic investigation was incomplete.

[178] On 21 November 2018 the applicants filed their second supplementary founding papers, but these were not affidavits on the merits, which is what the Judge President’s order had envisaged. These papers simply explained that the forensic investigation was incomplete and that it would not be ‘prudent’ to reveal any of the investigation’s details until it was complete. Gobodo’s identity as the firm conducting the investigation was not disclosed.

[179] On 12 March 2019 the Judge President vacated the dates 26-28 March 2019 and re-enrolled the case for hearing on 4-6 February 2020.

[180] Gobodo delivered its final report on 30 April 2019. It appears that the respondents were not notified of this fact. The applicants did not file supplementary affidavits at that time.

[181] On 31 May 2019 Glencore served an application to bar the applicants from filing further supplementary affidavits (‘the bar application’). (Glencore had launched a similar application in December 2018, but had withdrawn it when the applicants threatened to end parallel settlement discussions unless the application was withdrawn.) In their opposing papers in the bar application, the applicants stated that they would file their supplementary papers within 20 days from finalisation or withdrawal of the disclosure appeal. This was the first intimation that the applicants were linking the filing of their supplementary papers to the disclosure appeal.

[182] In June 2019 the Judge President transferred judicial case-management to Nuku J. On 26 June 2019 the latter issued timetables on two scenarios, viz the withdrawal of, or persistence with, the disclosure appeal. If CTSA and Glencore persisted with the disclosure appeal, there was a timetable for further affidavits in the bar application and in the applicants’ counter-application to extend the time for filing their supplementary papers. CTSA and Glencore persisted with the appeal, and the bar application and related counter-application were set down for hearing on 7 November 2019. On 5 November 2019 the applicants withdrew the counter-application.

[183] The bar application served before Nuku J on 7 November 2019, but instead, and at his urging, the parties agreed on a procedural way forward which was incorporated in an order. In terms thereof, Glencore withdrew its bar application (costs to stand over), on the basis that the applicants would file their supplementary papers by 7 February 2020, with a timetable for further papers. The main case would be heard on 14-17 September 2020. The order provided that if the SCA were to order disclosure of further documents, the applicants would be given 15 days from such order to file yet further supplementary papers dealing with those documents. As previously mentioned, the SCA on 13 December 2019 ordered production of the KPMG and PwC reports.

[184] In early February 2020 it emerged that the applicants’ legal team thought that their supplementary founding papers were only due on 28 February and that there had been a mistake in the order. This led to another meeting before Nuku J, who directed the applicants to file by 28 February 2020. (The respondents did not accept that there had been any error in the date of 7 February.)

[185] Thus it was that on 28 February 2020 the applicants filed their third, and most substantive, set of supplementary founding papers. The main supplementary affidavit ran to 130 pages and its attachments to another 423 pages. The KPMG and PwC reports were attached and discussed. In total, the founding papers cover pp 1-1361 of the record.

[186] The Gobodo report was mentioned but not attached to the supplementary papers. CTSA’s attorneys asked for it, offering confidentiality undertakings. CDH replied that they would only provide the report under court-approved confidentiality. It took two case-management meetings to iron out this issue. The report was furnished on 30 April 2020. A redacted version was placed in the court file. With its annexures, the report runs from pp 1362 to 2710 of the record.

[187] In accordance with Nuku J’s directions, Vitol, Glencore and CTSA delivered their opposing papers on 22 May 2020. Taleveras filed ‘explanatory’ papers on 26 June 2020. Taleveras’ position was that the applicants had, in their founding papers, tendered restitution, an offer accepted by Taleveras on 5 June 2020. This paved the way for the contest between CTSA and Taleveras as to which of them should receive the money if restitution were ordered. The respondents’ papers cover pp 2711-5005 of the record.

[188] The applicants filed their replying papers on 22 July 2020. Given the volume of the respondents’ affidavits, it is unsurprising that the replying papers run from pages 5005-5613. The respondents filed supplementary affidavits in response to new matter contained in the replying papers. These span pp 5614-5998. OUTA’s admission application and miscellaneous additional papers make up the modest balance of pp 5999-6196.

Grounds of review

[189] Since prospects of success are a factor in condoning or overlooking delay, I now deal with the merits of the review. Although the respondents concede that the impugned decisions and transactions are reviewable on some of the alleged grounds, it is desirable, in my view, to assess all of them. As OUTA’s counsel submitted, it is in the interests of transparency and accountability that the public should know what went wrong.

[190] It is also in accordance with the corrective principle that lessons be learnt. Apart from standards expected from executive officials, the case concerns failures of oversight by the SFF’s board, by CEF and by the MoE. A full survey may bring home to those responsible for appointments the need for greater care in selecting directors and senior executives of state-owned enterprises.

[191] The impugned acts are:[27]

(a) the MoE’s First Approval Notice of 12 November 2015;[28]

(b) SFF’s decisions, taken in late November 2015, to award volumes of strategic oil to Venus, Taleveras and Vitol;[29]

(c) the MoE’s Second Approval Notice of 7 December 2015;[30]

(d) SFF’s conclusion of SPAs and SGAs with Venus, Taleveras and Vitol;[31]

(e) SFF’s conclusion of a tripartite agreement with Venus and Glencore; [32]

(f) SFF’s conclusion of the side letter agreement with Taleveras and CTSA;[33]

(g) SFF’s board decision of 5 February 2016 to ratify the awarding of contracts to Taleveras and Vitol;[34]

(h) generally, the decision of SFF and/or the MoE to dispose of 10 million barrels of crude without following a procurement system that was fair, equitable, transparent, competitive and cost-effective.[35] 

[192] All the review relief claimed by CEF is governed by the Promotion of Administrative Justice Act 3 of 2000 (‘PAJA’). The review relief claimed by SFF is also governed by PAJA insofar as it is directed at decisions of the MoE. The review relief claimed by SFF in relation to its own conduct is self-review, governed by the legality principle (State Information Technology Agency SOC Limited v Gijima Holdings (Pty) Limited [2017] ZACC 40; 2018 (2) SA 23 (CC); Altech Radio Holdings (Pty) Ltd & others v City of Tshwane Metropolitan Municipality [2020] ZASCA 122 para 17). Since CEF claims the same relief as SFF, there is no point in exploring the extent, if any, to which the legality review in the present case would be more restrictive than the PAJA review or the question whether the exclusion of PAJA in cases of self-review applies where the public body, as here, purports to act in the public interest.

[193] I would simply observe, in relation to the second of these questions, that the recent judgment to which I was referred in argument, Compcare Wellness Medical Scheme v Registrar of Medical Schemes & others [2020] ZASCA 91, deals (in paras 13-20) with the case where one public body, acting in the public interest, seeks the review of another public body’s decision. A review of that kind was held to be governed by PAJA. The review relief which SFF seeks in regard to its own actions does not fall within the scope of Compcare.

The First Approval Notice (12 November 2015)

[194] The First Approval Notice was based on Gamede’s representations in his letter of 11 November 2015. These were seriously flawed. First, Gamede’s statement that most of the reserves were losing their relevance did not apply to the Bonny and was questionable in relation to the Basrah.

[195] Second, it was not a sensible time to sell. The market was in contango. Although the market could have dropped, and did drop, below its levels of 10 November 2015, the likelihood was that in the not too distant future the oil price would start an upward trajectory. Unless SFF replenished its stock promptly, there was a fair chance that it would have to do so at higher prices than the disposal prices. The applicants’ expert, Mr Driscoll, put the matter this way:[36]

‘In a backwardated market where supplies are tight and prompt values are elevated, storage is less attractive. This would present a favourable time for SFF to execute a sale buyback trade … [I]t would sell its strategic stock into a strong market and repurchase the forward stock at lower forward prices. In a contango market where supplies are surplus, demand weak and prompt values depressed, the incentives to sell strategic stock and repurchase forward are suppressed; however, SFF’s storage capacity would command more value especially from companies that are compelled to store inventory or traders looking to launch a contango play. If SFF holds excess storage capacity, a contango structure presents favourable opportunities to either lease storage at a premium or build up strategic stocks…

… It is not hindsight to state that SFF liquidated its strategic stockpile at a time when companies would normally store and defer inventory. Storage owners with spare capacity and [strategic oil stocks] would perceive a contango structure as an optimal signal to build up, not liquidate, inventory’

[196] Relevant to the foregoing point is that when the MoE issued her First Approval Notice, her Second Directive was in place. Her Second Directive did not permit outright disposals coupled with the mere hope of being able to repurchase stock when the price was favourable. Her Second Directive permitted rotation, not disposal. She had issued the Second Directive on the basis of Gamede’s statement that rotation would occur when oil prices were on the rise and that it would be conducted by a new trading division whose personnel would see to it that selling prices exceeded purchase prices. It is not quite clear to me how this stated objective would be met if rotation occurred when prices were on the rise. The only way in which it could be achieved was if the trading division concluded purchase contracts at lower prices followed shortly afterwards by sales at higher prices. Her Second Directive required that any rotation transaction be preceded by a detailed due diligence supported by a detailed motivation to the MoE.

[197] The MoE disregarded the Second Directive when issuing the First Approval Notice. What Gamede proposed on 11 November 2015 was not a rotation. His letter did not contain a comprehensive motivation. The MoE did not ask whether there had been detailed due diligence (on the evidence, it was lacking). She did not ask whether oil prices were on the rise (they were not, though the forward curve was upwards). She did not ask whether the trading division existed (it did not). She did not ask when replacement oil would be bought or how SFF would ensure that it was bought at lower prices than the disposal prices. Since the strategic stock was held as a reserve in case of emergency, one would have expected her to ask what would happen if the emergency struck at a time when oil prices were high. The MoE simply did not apply her mind to these matters.

[198] This is enough to show that a review of the First Approval Notice would succeed. To this I must add that the MoE was entitled to an honest motivation from Gamede. For reasons I shall presently explain, it can be concluded that Gamede was not genuinely concerned with the best interests of SFF and the country. His motives were improper.

The awards in general (late November 2015)

[199] By the end of November 2015, Gamede had identified Venus, Taleveras and Vitol as the parties to buy the oil, and had determined the quantity and grades of oil that each would be offered. If the First Approval Notice was unlawful, the awards of late November 2015 would fall with it. There are, however, other grounds for impugning the awards.

[200] There was a debate, in argument, about the procedure to be observed in the disposal of assets by a public body. Vitol’s counsel challenged the applicants’ contention that s 217 of the Constitution applied. Procurement, he submitted, is the opposite of disposal. The same was true, he argued, of s 51(1)(a)(iii) of the PFMA, which requires a public entity to have and maintain ‘an appropriate procurement and provisioning system which is fair, equitable, transparent, competitive and cost-effective’.

[201] He supported this distinction by pointing to the differing treatment of procurement and disposals in paras 16A.6 and 16A.7 of the Treasury Regulations. He acknowledged that para 16A was not actually applicable to the applicants as Schedule 2 entities,[37] but he made reference thereto in support of his general proposition. In para 16A, disposals are regulated with a lighter touch than procurement. In particular, para 16A.7(1) permits disposals to take place by way of price quotations, competitive bids or auction, whichever is the most advantageous to the State.

[202] Section 217 is headed ‘Procurement’. Section 217(1) refers to the case where a public body ‘contracts for goods or services’, which in its natural meaning conveys contracting to acquire good or services. Section 217(2) provides that subsection (1) does not prevent organs of state from implementing a ‘procurement policy’ of the kind described in subsection (2). The national legislation contemplated in s 217(3) is the Preferential Procurement Policy Framework Act 5 of 2000. That Act, and the regulations promulgated thereunder, are concerned with acquisition, not disposal (the best bid on price is the lowest one, not the highest).

[203] Dodson AJ surveyed the conflicting provincial judgments on this point in Airports Company South Africa Limited v Airport Bookshops (Pty) Ltd t/a Exclusive Books 2016 (1) SA 473 (GJ) paras 56-63. Two of the cases he cited were judgments in this division: SA Metal & Machinery Co (Pty) Ltd v The City of Cape Town 2011 (1) SA 348 (WCC) and CShell 271 (Pty) Ltd v Oudtshoorn Municipality [2012] 3 All SA 527 (WCC). In former case Binns-Ward J took it for granted that s 217(1) applied to disposals as well as acquisitions (para 5 and fn 2). In the latter case Henney J took the opposite view (para 36). Neither judgment was reasoned on this aspect.

[204] OUTA’s counsel submitted that Airports Company South Africa SOC Ltd v Imperial Group Ltd & others [2020] ZASCA 2; 2020 (4) SA 17 (SCA) was binding authority that s 217(1) applied to disposals as well as acquisitions. I disagree. The judgments in that case point in the opposite direction. The SCA held that although, in form, ACSA was granting concessions for commercial space in return for payment, it was in substance doing so in order to ensure that users of the airport would have access to car rental services. Section 217 required there to be a procurement, an acquisition, but the acquisition did not need to be for the public body itself and it did not need to involve the expenditure of money by the public body (see paras 22-23 per Molemela JA and para 63 per Ponnan JA). If ‘contracting for goods and services’ did not mean an acquisition, and could apply to a pure disposal, the learned Judges of Appeal would not have needed to reason as they did.

[205] I thus conclude that s 217(1) does not, in its own terms, apply to a public body’s disposal of assets. Nevertheless, a public body must adhere to the principles of administrative justice and legality when exercising its powers. The values embodied in s 217(1) should not be disregarded in determining whether a disposal decision complies with the public body’s constitutional duties.

[206] I see no reason why disposal decisions should not, in general (ie absent special circumstances), be subject to a requirement that the public body follow a fair, equitable, transparent and competitive process. The public interest is generally served if disposals take place at the highest price. That will usually be achieved by ensuring that all parties who can realistically bid have an opportunity to do so, and that the process is competitive.

[207] Para 16A.7(1) of the Treasury Regulations, although not directly applicable, is consistent with this in requiring the use of the method most advantageous to the State. Here, private negotiation was not more advantageous than competitive bids. Furthermore, Gamede selected Vitol, Venus and Taleveras before getting price quotations from them. None of the responses to the RFPs committed the bidders to a pricing window or to a particular discount or premium to Dated Brent.

[208] CEF’s group procurement policy, which applied in the present case, is consistent with this approach. From clause 2 one sees that the policy, which applies to ‘purchases, disposals and/or leasing of goods and services’, seeks to uphold ‘the constitutional principles of fairness, equitability, transparency, cost-effectiveness and competitiveness’. Clause 13, dealing with ‘procurement methods and thresholds’, explicitly applies to disposals as well as acquisitions. For all transactions exceeding R500,000, ‘open competitive tendering is the main preferred procurement method’ followed by ‘close competitive tendering as the second preferred procurement method’. Clause 13.4 permits negotiation without prior tendering ‘where open or close competitive tendering is not suitable’. The tendering process must be managed by a Procurement Committee whose proceedings have to be minuted. The Procurement Department is the ‘custodian of the procurement process’, with responsibility for ensuring compliance with legislation and SCM best practice.

[209] There is evidence that, in the international oil trade, competitive bidding is common, and that it is the usual way in which sales and rotations of crude are undertaken by national oil companies, although direct negotiation is not uncommon.[38] There is no reason why a competitive process could not have been done here.

[210] Gamede’s insistence on private negotiation, rather than competitive bidding, lacked rationality. Private negotiation creates opportunity for corruption (Allpay Consolidated Investment Holdings (Pty) Ltd & others v Chief Executive Officer of the South African Social Security Agency & others (1) [2013] ZACC 51; 2014 (1) SA 604 (CC) para 27). As I explain later, there is reason to conclude that Gamede preferred it for this very reason. Private negotiation also resulted in apparently arbitrary differences in contractual terms between the SPAs and SGAs concluded by SFF with Venus, Taleveras and Vitol respectively. This affected, among other things, provisions relating to the governing law and limitations on liability, a matter to which I shall return later in this judgment.

[211] A great deal has been said in the papers about the prices at which the oil was sold, but when one stands back and looks at the bigger picture, the Venus and Taleveras transactions tell their own story:

(a) Venus, an entity with no oil history (and which, I may add, had no existing storage relationship with SFF), bought 3 million barrels of Bonny which it immediately on-sold to Glencore, turning a profit of $10,536,000. If Glencore had had the chance to bid directly to SFF, there is no reason to think it would not have offered SFF the price it agreed to pay Venus.

(b) SFF gave Venus a storage rate of $0.11/bbl per month. Glencore was prepared to pay Venus $0.25/bbl. If Glencore had had the chance to bid directly to SFF, there is no reason to think it would not have offered SFF the price it agreed to pay Venus.

(c) Taleveras was able to get financing from Natixis on a forward price of $179,920,000. After deduction of substantial transaction fees for CNX and of the anticipated storage costs, a net amount of $134,568,426 was left over for Taleveras, leaving it with a profit of $22,568,426 after deducting what it owed SFF. Even if Taleveras had spent some of this money to hedge its obligation to repurchase the oil after 27 months, it would have been left with a large profit.

[212] The process followed by Gamede did not comply with SFF’s constitutional duties to observe administrative justice and legality, and was an unauthorised and unjustified deviation from CEF’s procurement policy.

[213] Even if SFF did not have to follow a process which was fair, equitable, transparent and competitive, the disposal awards were irrational. There is no evidence explaining why particular parties were selected to receive RFPs; why different parties received different RFPs with different dates for responses; and what the evaluation criteria were. Gamede, who on the evidence was acting alone in choosing the successful bidders, did not document his process.

[214] The awards were also in conflict with the SFF board decisions of 13 October 2015 and 23 November 2015. At the former meeting the board had required a project plan and an action plan for implementing the Second Directive, setting out the short, medium and long term intentions, the resources needed and the approvals required from the board. The board had also decided that a clarificatory letter on funding should be sent to the MoE. The board’s intention must have been that there was to be no rotation decisions until the board had approved the plans and until the MoE had provided the necessary clarification.

[215] At the meeting of 23 November 2015 the board was clearly unhappy with Gamede’s letter of 11 November 2015 and the resultant First Approval Notice. The board resolved that every sale should be back-to-back with a purchase at a favourable margin to SFF and that every purchase or sale had to be pre-approved by the board. The board was assured that it would receive the strategic plan in January 2016.

[216] In making the awards, Gamede did not intend there to be back-to-back purchases. He did not seek the board’s pre-approval. The strategic plan had not been placed before the board. The clarificatory letter had not been sent to the MoE.

The award to Taleveras – ulterior purpose

[217] In its explanatory affidavit, Taleveras has given background to the conclusion of its purchase of 4 million barrels of strategic stock. There was a commercial relationship with SFF going back to an SGA concluded in October 2014. It is unnecessary to go into details. It is enough to say that there were accusations and counter-accusations of breach and repudiation in relation to this and later contracts. Taleveras considered that it had substantial claims for damages against SFF.

[218] According to Taleveras, the parties pursued negotiations to reach a commercial arrangement that would enable Taleveras to recover some of its losses. These negotiations were the ‘springboard’ for the conclusion of the impugned Taleveras contracts, which ‘spared SFF facing large claim for damages and a very expensive round of litigation’.

[219] In its replying papers, the applicants – who were unaware of this explanation for the Taleveras contracts – allege that the conclusion of the Taleveras SPAs and SGAs in December 2015 was, on Taleveras’ own version, vitiated by improper proper. Gamede was not entitled to favour Taleveras for contracts in December 2015 in order to compensate the latter for alleged losses arising out of earlier contracts. Taleveras had not proved an entitlement to contractual damages. Based on the correspondence, SFF was disputing that it was liable for any losses.

[220] I agree with the applicants’ contentions. SFF’s decision to sell some of the strategic oil to Taleveras had to be justified independently of the unresolved commercial disputes between the parties.

The award to Taleveras - bribery[39]

[221] Gobodo reported that over the period January 2015 to April 2016 large amounts of cash were paid into Gamede’s bank account at ATMs. Often multiple deposits were made at a single ATM over short spaces of time. While these deposits raise eyebrows, and while some of them might be linked to significant events in the disposal of the strategic stock, the applicants’ counsel did not argue that I could find that any of them were bribes. In fairness to Mr Gamede, it must be noted that there were also significant deposits in March and April 2015, which bear no relation to the impugned transactions. Furthermore, at least some of the cash deposits were made at ATMs located in the vicinity of casinos, so it is possible that Gamede was a gambler who was depositing his winnings.

[222] However, there are four other payments to Gamede which can on the papers be found to have been bribes:

(a) On 24 November 2015 R684,605 was deposited into Gamede’s dormant practice account (‘first Lengard payment’).

(b) On 30 November 2015 R707,000 was deposited into Gamede’s dormant practice account (‘first Mulaudzi payment’).

(b) On 19 January 2016 R601,000 was deposited into Gamede’s dormant practice account (‘second Mulaudzi payment’).

(b) On 3 February 2016 R670,000 was deposited into Gamede’s dormant practice account (‘second Lengard payment’).

[223] Taleveras’ controller is Mr Igho Sanomi. Other companies in the Taleveras group include Charmondel, Lengard Projects Ltd (‘LPL’) and Taleveras Oil (SA) (Pty) Ltd (‘TOSA’). In negotiations which preceded the Taleveras contracts, Taleveras proposed LPL as a company which would provide a letter of credit. LPL is managed out of Taleveras’ offices in Dubai. Godfrey Mulaudzi was formerly South Africa’s Deputy High Commissioner in Nigeria. He was a director of TOSA from June 2015 until July 2016.

[224] In regard to the first Lengard payment, on 23 November 2015 Gamede sent his bank an invoice in relation to an incoming payment he was expecting. The invoice, INV-0001, is dated 1 November 2015 and was issued to ‘Langard Projects Ltd’ of Dubai. Gamede purported to charge for legal services rendered in June and July 2015 (60 days at $25,000) and September and October 2015 (another 60 days at $25,000). I am satisfied that ‘Langard’ is a spelling error by Gamede, one he repeated in respect of the second Lengard payment. He intended to issue these invoices to the Taleveras entity, LPL.

[225] The first Lengard payment was made the next day, 24 November 2015. The amount received into Gamede’s practice account was R684,605. If this was payment for an invoice totalling $50,000, the payment occurred at an exchange rate of $13.69. According to publicly available information, the lowest quoted dollar/rand exchange rates on 23 and 24 November 2015 were $14.00 and $13.97 respectively.

[226] Gamede received the first Mulaudzi payment of R707,000 about a week later, on 30 November 2015. The bank reference for the deposit was ‘Godfrey Mulaudzi’. In its affidavit, Taleveras, having consulted with Mulaudzi (who gave a confirmatory affidavit), offered the following explanation. Mulaudzi, in addition to being a director of TOSA, provided consultancy services to a Nigerian company, Omuza Global Services Nigeria Ltd (‘OGS’). OGS wanted legal advice on regulatory matters concerning the financial, mining and property sectors in various SADC countries. To this end, on 1 June 2015 Mulaudzi concluded a written agreement with ‘Gamede Attorneys t/a Gamede Legal Consultants’ for the provision of such advice. Gamede was entitled to charge at an hourly rate of R7500.

[227] Allegedly pursuant to this agreement, on 30 September 2015 Gamede issued Invoice 001 to Mulaudzi for R750,000 in respect of ‘Analysis of the Investment Regulatory Framework in SADC’. In the schedule of hours attached to the invoice, the work was said to relate to the ‘event date’ ‘July-September 2016’. Gamede charged the following hours: 5 hours in June for taking instructions; 30 hours in each of June, August and September for conducting research; and 5 hours in September for writing and presenting his report to Mulaudzi.

[228] Mulaudzi’s explanation for the payment of R707,000 rather than R750,000 is that he and Gamede intended that the latter’s remuneration would be the rand equivalent of $50,000. (If that is so, it is not readily apparent why the mandate agreement of 1 June 2015 did not say so.) The amount Mulaudzi billed OGS was $50,000, and the latter’s payment to Mulaudzi came to R708,948. On the same day he paid Gamede R707,000 in full and final settlement.

[229] Gamede received the second Mulaudzi payment, in the amount of R601,000, on 19 January 2016. Taleveras’ explanation is that this was, once again, payment for legal services rendered to Mulaudzi in terms of the mandate of 1 June 2015. Mulaudzi supplied Taleveras with Gamede’s invoice dated 15 December 2015 in the amount of R750,000. This invoice, in the name of Gamede’s defunct firm, was Invoice 002, and related to ‘Analysis of the Legal Regime for protection of foreign investments in Botswana’. The attached timesheet specified an ‘event date’ of ‘September-December 2015’ but listed alleged attendances in ‘February’ and ‘March’, as follows – 50 hours of research in ‘February’, 45 hours of research in ‘March’, and 5 hours in ‘March’ for ‘writing the report and presenting it to client’. Taleveras offers no explanation for why only R601,000 was paid in respect of the second invoice, save to allege that according to Mulaudzi the sum was paid ‘in full and final settlement’ of the invoice.

[230] In regard to the second Lengard payment, on 3 February 2016 Gamede, who was evidently expecting an incoming payment, sent his bank an invoice dated 30 January 2016, marked ‘Invoice No 2’, in respect of legal services supposedly rendered to ‘Langard Projects Ltd’ of Dubai. His charge was $50,000 for legal services spanning 60 days (the invoice does not state when the services were rendered). On the following day Gamede received R670,000 into his account. Since the exchange rate was about $15.86 on 4 February 2016, one would have expected an invoice of $50,000 to lead to a payment of R793,000.

[231] During argument, I asked Taleveras’ counsel whether they would be able to produce the reports which Gamede prepared for Mulaudzi. Counsel replied that they had asked Mulaudzi who claimed not to have them.

[232] That Gamede’s invoices to LPL and Mulaudzi are bogus is clear beyond doubt. At the time he supposedly performed the legal services, he was no longer in legal practice. He was employed full-time as SFF’s Acting CEO at a monthly salary of R120,000. Even if he did some moonlighting, he could not possibly have spent the days and hours alleged. On the assumption that the second Lengard invoice related to time spent in November and December 2015, Gamede was claiming to have spent the whole of June-July and September-December 2015 in rendering legal services to LPL (30 days per month). In two of those months (June and September) he also claims to have devoted 30 hours and 35 hours respectively to Mulaudzi (the equivalent of 4-5 days per month), while in August (when he supposedly took a break from working on the LPL mandate) he devoted another 30 hours to Mulaudzi. And in the second Mulaudzi invoice he charged for another 100 hours of work over two months (about 12-13 days of work).

[233] There are other tell-tale signs of fraud:

(a) In his first invoice to Mulaudzi, Gamede identified the ‘event date’ as ‘July-September 2016’. Since he supposedly issued the invoice in 30 September 2015, one would have to treat ‘2016’ as a typographical error. That type of mistake sometimes occurs early in a new year; it would not happen in September. Despite the fact that the period ‘July-September’ is identified, Gamede charged for 35 hours spent in June but did not charge for any time in July.

(b) In his second invoice to Mulaudzi, Gamede identified the ‘event date’ as ‘September-December 2015’ but charged for hours spent in ‘February’ and ‘March’. On Mulaudzi’s version, this could not refer to February and March 2015, since Gamede only received his mandate in June 2015. However, since the invoice was issued on 15 December 2015, and purported to relate to work already done, it could not refer to February and March 2016 either.

(c) Then there is the fact that Gamede issued two invoices with the invoice number ‘1’ and two invoices with the invoice number ‘2’.

(d) Finally, there is the suspicious circumstance that none of the payments accord with the invoiced amounts, and in three instances there is a significant difference.

[234] Although TOSA, the company of which Mulaudzi was a director, was not directly involved in the negotiations with SFF, Mulaudzi featured in dispatches. He was the person to whom Gamede on 30 October 2015 emailed the Taleveras RFP. Although the RFP was subsequently re-directed to Sanomi, Mulaudzi (among others) was copied in internal Taleveras communications relating to the transaction (see the emails of 2 November 2015 and 15 January 2016). And he signed most of the contracts between SFF and Taleveras as a witness.[40]

[235] In my view, Taleveras’ denials of corruption are so far-fetched and untenable that they can be dismissed on the papers. The four payments I have identified were bribes. The evidence does not allow me to say who within the Taleveras group, apart from Mulaudzi, was privy to the corruption, but the bribes must have been paid with a view to advancing Taleveras’ interests.

[236] Corruption has been condemned by the Constitutional Court as inconsistent with the rule of law and the Constitution’s fundamental values. If unchecked, it poses a serious threat to our democratic order: it imperils the State’s capacity to fulfil its obligations, stunts economic growth and puts the stability and security of society at risk (South African Association of Personal Injury Lawyers v Heath & others 2001 (1) SA 833 (CC) para 4; Glenister v President of the Republic of South Africa & others 2011 (3) SA 347 (CC) para 166).

[237] What makes the bribes, from Gamede’s perspective, all the more brazen, is that on 14 September 2015, about two months before receiving the first Mulaudzi payment and awarding strategic stock to Taleveras, Gamede wrote to the MoE to explain why SFF’s previous relationship with Taleveras had gone sour. His explanation included the following statement: ‘The Taleveras saga was an unfortunate incident due to the dishonesty of Taleveras in its dealings with African governments.’

The Second Approval Notice (7 December 2015)

[238] If preceding decisions fail, the Second Approval Notice would fall with them. Once again, however, there are independent attacks on the Second Approval Notice.

[239] In the Second Approval Notice, the MoE referred to a Gamede letter dated 6 December 2015. So poorly was the process handled that if such a letter ever existed, neither SFF nor the DoE/current MoE have been able to find it. The applicants make the reasonable inference that if such a letter existed, it was along similar lines to Gamede’s letter to the MoE dated 30 November 2015, in which he motivated Venus’ selection.

[240] Gamede’s representation to the MoE that SFF had assessed the proposals and found them to be sound and acceptable was untrue or at least misleading. The statement that ‘SFF’ had undertaken the assessment conveyed that it was a decision in which SFF’s senior management, if not its board, was involved. On the evidence, only Gamede was involved.

[241] Because of the improper way in which the process was documented, Venus’ proposal of 30 November 2015 has not been found. The Taleveras and Vitol proposals did not offer a price formula, and there is no reason to believe that Venus’ proposal was any different. Price was important to ensure that SFF’s interests were safeguarded. One would have expected SFF to want to know (a) how the bidder proposed to determine the pricing window; (b) importantly, the discount/premium to Dated Brent which the bidder was offering for Basrah/Bonny.

[242] Gamede did not tell the MoE about non-compliance with the board decisions of 13 October and 23 November 2015. He was under a duty to tell her. She would not have issued the Second Approval Notice if he had conveyed their import. Gamede did not tell the MoE that Taleveras had been favoured as a way avoiding costly litigation. If Taleveras’ version is true, Gamede’s non-disclosure was material.

[243] The contango market conditions which made rotational transactions at the time of the First Approval Notice irrational persisted when the MoE issued the Second Approval Notice. Ms Bossley says in her report that in December 2015 there was a widely held belief that prices would fall further in 2016 but that there were also signs that the contango market was threatening to tip into backwardation. The significance of this for SFF, she says, was that storage space is more valuable in contango than in backwardation, so that if one wants to generate storage revenue one should not wait for the market to be in backwardation. It is only with the benefit of hindsight, she observes, that one knows (a) that SFF sold its oil at or near the lowest point in the market; (b) that the market retreated in the first quarter of 2016 from the threatened backwardation into a firm state of contango.[41]

[244] Ms Bossley may be right about what knowledgeable parties were expecting in December 2015. However, there is no evidence that Gamede was knowledgeable or that he was thinking along these lines. Gamede was representing to the MoE that he would only sell when the market was rising. If, like Ms Bossley and knowledgeable experts, he expected prices to drop further in 2016, a sale of the stock in December 2015, however favourable for the earning of storage fees, was contrary to the market conditions which he had specified to the MoE as a prerequisite for rotational transactions. There is, in any event, no evidence that Gamede was intending to execute buybacks in the near future in expectation of further price drops.

[245] Moreover, whatever the market might have been threatening, it did not in fact reach a state of backwardation. Prices for future delivery remained higher than for immediate delivery. While this might have been favourable for concluding storage agreements, storage revenue could not come at the expense of security of supply. Gamede was dealing with strategic stock. The process required careful management if SFF was not to be exposed to the price of oil in a rising market. This was the point Mr Driscoll made in the passage I quoted earlier.

[246] Mr Driscoll states, further, that during 2015 there was a protracted downward price correction, further exacerbated on 4 December 2015 (about two weeks before the first SPA) when OPEC members failed to reach agreement on production cuts.[42] These are matters of which those managing our country’s strategic reserves could be expected to have been aware. Mr Driscoll observes that SFF (or Gamede, more accurately) ‘shrugged off or missed key market structure signals … and elected to sell its entire inventory of strategic stock in adverse market conditions’. Mr Driscoll found it ‘difficult to fathom Mr Gamede’s rationale for pressing forward with his apparent determination to liquidate SFF’s entire stockpile during a time of extreme, adverse and deteriorating market conditions’. Gamede, he says, seems not to have conducted a risk assessment before concluding the SPAs; and having concluded them, he neglected to hedge SFF’s exposure.[43]

[247] I should add that although the market was continuing to fall in December, it remained in contango. In other words, on each date that the price fell, that date’s forward curve still sloped upwards, albeit from a lower starting point than on earlier dates.[44] The market was thus still commanding higher prices for future delivery than prompt delivery. And so, as Mr O’Driscoll states, if SFF had executed buybacks on the dates of the SPAs, ‘it would have paid a substantial premium for forward resupply and locked in a trading loss’.[45] If Gamede was hoping that oil prices would drop further, he was gambling, and his gamble failed, but there is no evidence that he was thinking along these lines.

[248] The MoE’s Second Approval Notice was thus vitiated by material factual errors induced by Gamede’s misrepresentations and non-disclosure. However, the MoE herself cannot be exonerated. She had to apply her own mind. As the executive authority responsible for CEF and SFF in terms of the PFMA, she was required to exercise the State’s control over these entities to ensure that they complied with the PFMA and the MoE’s financial policies (s 63(2) of the PFMA).

[249] She was not given the proposals. She explicitly granted approval on the basis that SFF/Gamede (not that she) thought the proposals were sound. In her Second Directive she had quite properly insisted that proposals to her should be in the form of a comprehensive motivation preceded by a detailed due diligence; that strategic stock levels were to be ensured at all times; and that there should be an appropriately staffed trading division. Nothing in the record indicates that she received a ‘comprehensive motivation’ or that she asked for particulars of the ‘detailed due diligence’. She did not ask whether the trading division had been set up. Importantly, she did not ask what steps were being taken to ensure that strategic stock levels were not impaired. We know as a fact that Gamede was taking no such steps. He was simply going to sell the strategic stock, perhaps hoping that on some future occasion it would be possible to replenish at lower prices.

The conclusion of the contracts in general

[250] If the preceding decisions of SFF and the MoE were invalid, the contracts concluded on the strength thereof must also fall. Once again, though, there are independent grounds for impeaching the contracts.

[251] In their first iterations, all the contracts, including the side agreements involving Glencore and CTSA, were concluded before the SFF board meeting of 5 February 2016. In purporting to bind SFF to those contracts, Gamede was acting without authority and in violation of the board resolution of 23 November 2015. He did not have the board’s pre-approval. He did not comply with the requirement that SFF should have back-to-back purchase contracts at a margin sufficient to cover all incidental costs while still leaving SFF with a margin.

[252] Whether the SPAs and SGAs concluded before 5 February 2016 were validly ratified at the board meeting of that date will depend on the review directed at the board’s decision. The amendments which Gamede purported to execute after 5 February 2016 were not submitted to the board and were thus unauthorised. These include the final pricing amendments made to the Taleveras and Venus SPAs on 22 and 29 February 2016.

[253] The conclusion of the contracts violated s 54(2)(d) of the PFMA. In terms thereof, SFF’s ‘accounting authority’, ie its board, was required, before the conclusion by SFF of any contract for the disposal of a ‘significant asset’, to notify the National Treasury of the transaction and to ‘provide relevant particulars of the transaction’ to the MoE for her approval. In my view, these are peremptory requirements, non-compliance with which renders a transaction of the kind in question invalid.

[254] Although treasury approval is not required, the requirement of advance notification allows the treasury to take the matter up with the executive authority if there are concerns. In the case of an innocent oversight or technical infringement, a court would not necessarily couple a declaration of invalidity with a setting aside of the transaction. However, one would undermine the important purposes served by s 54(2) if one dismissed its requirements as a mere reporting function without practical consequence if ignored.

[255] Some indication of the detail which National Treasury expects is seen in the guidance contained in its ‘Practice Note on Applications under Section 54’. In terms of clause 4, the application in the present case was required to be addressed to the MoE (as the executive authority) and MoF (as head of National Treasury). Clause 7 read with Annexure A contains guidelines on information to be supplied. These guidelines would have required SFF’s board to deal with the following matters: a description of the transactions and their objectives; the financial viability of the transactions (cash flow analyses, estimates of future revenue); details of how the transactions would be funded; risks pertaining to the transactions, and strategies for mitigating those risks; whether alternatives were considered; how success would be measured; human capital capacity; and necessary board approvals.

[256] Even if notification by Gamede, rather than SFF’s board, could be substantial compliance with the section, Gamede did not notify the National Treasury, and there is no evidence that in seeking the MoE’s approval he gave her ‘relevant particulars’ of the transactions in accordance with the Practice Note.

[257] The applicants have also alleged that the disposals contravened s 112(2) read with s 115 of the Companies Act 71 of 2008 in that the 10 billion barrels represented ‘all or the greater part’ of SFF’s assets. Sections 112(3)(b)(i) and 115(1) make clear that a disposal of the greater part of a company’s assets covers the case where this is achieved by a series of agreements. In terms of s 112(3), compliance would at a minimum have required a special resolution by CEF approving the specific transactions after notice from SFF accompanied by a summary of the precise terms of the transactions or series of transactions.

[258] Although s 112(2) of the Companies Act might not attract public law remedies in relation to an ordinary company, the position is different in my view where the company is a state-owned enterprise performing public functions. In relation to such a company, compliance with the Companies Act is as much part of its public law duties to act lawfully as compliance, for example, with the PFMA.

[259] Vitol denied that the 10 million barrels constituted the greater part of SFF’s assets.[46] Although Vitol’s denial was bald, it was responding to a bald allegation.[47] The applicants should have provided financial information from which the court could draw the legal conclusion that the 10 million barrels fell within the ambit of s 112(2). I thus cannot make a finding on this ground.

The Taleveras and Venus contracts – price of Bonny

[260] In the original Venus SPA, SFF sold the Bonny to Venus at a discount of $4 to Dated Brent. This discount was not justified, and was irrational. Gamede himself in effect admitted this in the amendments to the Taleveras and Venus SPAs concluded on 22 and 29 February 2016. He, and Taleveras and Venus, nevertheless evaded the implications of this by selecting a past pricing window which represented an 11-year low for Dated Brent and by deviously dressing up the price of the Bonny as being at a $3 premium to Dated Brent.

[261] The Taleveras SPA originally sold the Bonny at a price still to be agreed, which left plenty of scope for manoeuvre. In the amendment of 29 January 2016, the price for the Bonny was set at $30. On that date Brent was trading about $4 higher than this price. On 22 February 2016 the SPA was again amended, this time in the way that served as the model for the Venus amendment of 29 February 2016. As at 22 February 2016 Brent was trading at around $34.60, so the agreed price of $30.075 more or less preserved the $4 discount.

[262] In Taleveras’ case, the bribes paid to Gamede probably explain the original pricing and the later devious manipulation. In Venus’ case, there is no evidence of bribery, but the fact that this unknown entity features at all is highly suspicious. Its selection did not fit with Gamede’s supposed preference for dealing with existing storage clients. Gamede told Gobodo that Zittatu was disqualified because it was a new company without the required financial capacity. If he had granted further interviews, he would no doubt have been asked in what way Venus differed from Zittatu. I have drawn attention to the fact that certain important documents, specified in the expanded RFP issued to Vitol, were omitted from the RFP issued to Venus – audited accounts, a tax clearance certificate and the necessary petroleum licences to handle and export crude oil. The inference is irresistible that there was an improper motive for Gamede to have made an award to Venus and to have sold Bonny to it at an irrationally low price.

[263] Glencore’s expert, Ms Jago, explains why the price Glencore paid Venus for the Bonny was market-related. Given the fact that SFF’s price to Venus was $10,536,000 less than Venus’ price to Glencore, the fact that the latter sale was below market value is self-evident.

[264] CTSA’s expert, Ms Bossley, states that the premium ultimately paid by Taleveras in relation to the pricing window of 18-20 January 2016, viz $3, was actually quite high. She would have expected a premium in the range of $0.50 – $1.50.[48] She fairly observed that there were ‘atypical features’ in the Taleveras pricing clause: the fact that the initial pricing window related to the inspection and certificate of quality, given the arbitrary nature of the timing of the inspection (quite possibly selected, I may add, by Taleveras); that the SPA was then amended to specify a flat price; and that the final amendment reinstated a formula price, ‘allowing observers to infer a formula price differential from the fixed and flat price agreed’.[49]

[265] Perhaps the premium of $3 was high, though it is not without significance that Glencore, a very experienced international trader, initially agreed to pay a premium of $2 which it later increased to $3. Glencore’s expert, Ms Jago, expressed the view that for a pricing window of 8-10 February 2016, the market premium was $3.02,[50] while the applicants’ expert, Mr Driscoll, believed that $3 was below a true market premium.[51] Be that as it may, the true effect of the premium depended on the genuineness of the pricing window. Ms Bossley did not have all the evidence bearing on this question.

[266] Moreover, Gamede was not thinking along Ms Bossley’s lines. The recordals in the amendment of 22 February 2016 show that he was under the impression that he needed to demonstrate a premium of $3, having previously agreed (albeit in relation to Venus) to sell Bonny at a discount of $4. The clear impression one gets is that he was dead-set on doing the deals done but hopelessly out of his depth (even if he was acting honestly, which he was not).

[267] The fact that Taleveras bought at less than market value again seems to be borne out by the fact that it was able to realise an immediate amount of $22,568,426 in its transaction with CTSA.

The Taleveras contracts – the pricing of Basrah

[268] The original Taleveras SPA in respect of the Basrah left the price open to future negotiation. The price was eventually fixed on 29 January 2016 at $26. It will be recalled that on 22 January 2016 Vitol and SFF amended their SPA to fix a pricing window of 25-29 January 2016 and a discount to Dated Brent of $5.50. This pricing window yielded an average Dated Brent price of $31.702, resulting in a final price of $26.202.

[269] It is a fair inference that the Taleveras amendment of 29 January 2016 was concluded to bring the Taleveras price more or less in line with the Vitol price. However, if Gamede had insisted on exactly the same price ($26.202 rather than $26.00), SFF would have received an additional $404,000. Since this particular aspect of the pricing was not advanced as a ground of review, I merely mention it for the sake of completeness.

[270] It is noteworthy that whereas the Bonny SPA was amended to refer to the low pricing window of 18-20 January 2016, a similar change was not made to the Basrah SPA, which reinforces the view that the final pricing of the Bonny was a deliberate manipulation to create the impression of a premium. If a pricing window of 25-29 January 2016 (the implied amended pricing for Taleveras’ Basrah) had been applied to the Bonny, the price of the latter would have been considerably higher.

Storage fees

[271] I have previously mentioned that whereas Venus was paying SFF a monthly storage fee of $0.11/bbl, Glencore was paying Venus $0.25/bbl. The fact that an experienced international trader was willing to pay $0.25 for storage casts doubt on the propriety of the fee of $0.11.

[272] KPMG reported that a pricing study had been commissioned to determine whether the storage fees in the impugned transactions were market-related.[52] If the study was done, it has not been adduced in these proceedings. During his interview with Gobodo, while being questioned on another aspect, Gamede in passing stated that SFF charged $0.11 as a storage fee but that he knew in Singapore they were charging $0.90. He claims to have told his colleagues that ‘we can’t be charging 11 cents, because we are losing’.[53] This suggests that $0.11 may have been SFF’s usual rate at the time and that Gamede thought it to be uncommercial. In those circumstances, it is surprising that he concluded storage agreements with Venus, Taleveras and Vitol at these levels. Since his professed goal was to boost SFF’s revenue by way of storage fees in a contango market favourable to storage operators, this was hardly achieved by charging a very low rate at which SFF was losing money. Mr Driscoll opined that given the contango state of the market, storage would have commanded a significant premium. In his view, the storage fees did not reflect ‘the intrinsic value of the contango carry’.[54]

[273] I mention this for the sake of completeness. The applicants did not allege that a storage rate of $0.11 was not market-related or make it a ground for impugning the transactions.

The Vitol contracts

[274] I shall deal with Vitol’s conduct more fully in relation to a just and equitable remedy. At this point I simply record that I have found no peculiar improprieties regarding the conclusion of contracts with Vitol.

[275] The applicants complain that whereas the Second Approval Notice approved a disposal to Vitol, Gamede concluded the relevant contracts with Vesquin, the contention being that the MoE did not approve of a disposal to Vesquin. The latter is a BEE entity and subsidiary of seventh respondent, which is a South African company controlled by the eighth respondent, a Swiss company. Ducrest told Gamede that Vesquin was the entity Vitol would use to buy and store the oil (see his email to Gamede of 23 November 2015). Vesquin had already been identified as the Vitol entity to transact with SFF in the proposals of 4 and 15 September 2015. The applicants have not located or produced Gamede’s recommendation to the MoE in relation to Vitol, so we do not know whether Vesquin was mentioned. The Second Approval Notice did not refer to a specific entity, simply to ‘Vitol’.

[276] The terms of the SPA and SGA were such that SFF was not exposed to a credit risk by virtue of Vesquin’s nomination as the purchaser and lessee. In terms of clause 8 of the SPA, the buyer had to open a letter of credit in SFF’s favour prior to delivery for the full value of the product. In terms of clause 8.5 of the SGA, the lessee’s liability for storage fees had to be secured by an amendment of the existing parent-company guarantee in respect of Tank 3.

[277] In the circumstances, I consider the MoE’s Second Approval Notice to have been sufficient to cover a transaction with a nominated company within the Vitol group. Any infraction in this regard was de minimis.

[278] On this point the applicants’ counsel relied on the majority judgment in Areva NP Incorporated in France v Eskom Holdings Soc Ltd & another [2016] ZASCA 51; 2017 (6) SA 621 (CC). That case is distinguishable. It was concerned with locus standi. The company which sought to review Eskom’s awarding of a tender to Areva was not the company which had submitted the rival bid (A) but a company (B) belonging to the same group as A. The majority held that B lacked own-interest standing and had not claimed to be acting on behalf of or in the interest of A. In the present case we are not concerned with locus standi. The question is the true scope of the approval given by the MoE in the Second Approval Notice. On that question, I am not satisfied that the MoE’s reference to ‘Vitol’ was such as to confine approval to one specific identifiable company within the Vitol group.

The board decision of 5 February 2016

[279] As SFF’s ‘accounting authority’, SFF’s board had the duties set out in ss 50 and 51 of the PFMA, including the duty of utmost care to ensure reasonable protection of SFF’s assets; to act with fidelity, honesty, integrity and in SFF’s best interests in managing its financial affairs; to seek, within their sphere of influence, to prevent any prejudice to the financial interests of the State; to ensure that SFF had and maintained effective, efficient and transparent systems of financial and risk management and internal control; and to take effective and appropriate steps to prevent irregular, fruitless and wasteful expenditure. Substantially the same duties would in any event flow from the ordinary responsibilities of directors under s 76 of the Companies Act.

[280] If the board was told no more than what is reported in the minutes, the board had insufficient information to reach a proper decision. The difficulty is that the minutes refer to a ‘trading report’ which was ‘taken as read and duly noted’. The applicants have not produced the report. However, Jawoodeen tells us that the board did not have the contracts. Ngqongwa states that management was meant to prepare board packs a week in advance. In respect of the board meeting of 5 February 2016, board packs were only handed out at the meeting. There were not enough copies for all attendees.

[281] Ngqongwa does not say that the board packs included the contracts, but he does say that the exco members themselves saw the contracts on 5 February at the meeting venue. After Nkutha, Mayaphi and he were excused from the meeting, they went to the dining hall ‘to go through the agreements thoroughly’. While doing so, the board meeting finished. They felt uncomfortable about continuing their exercise and resolved to meet again on Sunday 7 February.

[282] Jawoodeen and Ngqongwa’s evidence in this regard should have formed part of the founding papers. Nevertheless, the respondents were afforded the opportunity of filing supplementary answering papers, and they did not apply to strike out the material. On the strength thereof, I consider the conclusion to be justified that the directors did not have sufficient information or sufficient time to form a proper view of the propriety of the contracts, bearing in mind their duties under the PFMA and Companies Act.

[283] In the case of the Taleveras transaction, if the board was not told what price Taleveras was paying for the Bonny, they did not have enough information to make a proper decision. If board was told the price, they could not rationally have approved it. And obviously the board did not know that Gamede had been bribed. For these reasons alone, the board’s approval of the Taleveras transactions cannot stand.

[284] It is unclear from the minutes precisely what the board decided in respect of Venus. The requirement that Venus should perform within 30 days points to an acceptance of the transaction, provided Venus performed. To the extent that the board by necessary implication approved the Venus transaction, such approval was irrational, either because the board did not know the price or because it approved an irrationally low price. Furthermore, the board could not have had any information to satisfy itself of Venus’ credentials as a buyer of the strategic stock. I have concluded that there must have been an improper motive for Gamede to have made an award to Venus, and self-evidently the truth was not disclosed to the board.

[285] Although the board would have needed more time in order to make an informed decision about the commercial propriety of the contracts, it did not need more time in order to know that the disposal process did not accord with the CEF group procurement policy; that the disposals should have occurred by way of a competitive process; and that the disposals violated s 54(2)(d) of the PFMA. The directors should at least have been aware that internal and legislative procurement issues were at stake, and should have taken time to get advice on compliance before approving the transactions.

Delay – legal principles

[286] The main principles on the question of legality and PAJA delay I can take from paras 44-71 of the majority judgment in Buffalo City Metropolitan Municipality v Asla Construction (Pty) Limited [2019] ZACC 15; 2019 (4) SA 331 (CC), supplemented where necessary with reference to other authorities.

[287] In both legality and PAJA reviews, time starts to run from the date the applicant becomes aware or reasonably ought to have become aware of the impugned action and the reasons for it (I call this the trigger date). Although Theron J in Buffalo City spoke (para 49) of knowledge of the action taken (ie the impugned action) as setting the clock ticking, she supported this with reference to City of Cape Town v Aurecon South Africa (Pty) Ltd [2017] ZACC 5; 2017 (4) SA 223 (CC) (‘Aurecon CC’) para 41, stating (in her fn 36) that there was no reason why the PAJA rule should not apply also to legality reviews. As appears from para 41 of Aurecon CC read with s 7(1)(b) of PAJA, the relevant knowledge is knowledge of the decision and the reasons for it. Aurecon CC emphasises, though, that it is not necessary, in order for time to run, that the applicant should have knowledge that the impugned action was tainted by irregularity.

Legality delay

[288] The legality delay rule consists of two parts. The first is whether the delay was unreasonable. This is a factual enquiry on which a value judgment must be made. If the court finds the delay not unreasonable, the court goes to the merits. If the court finds the delay unreasonable, the second question is whether the court should overlook the delay. Unlike PAJA’s s 9(1), this does not require an application for condonation.

[289] In deciding whether a delay was unreasonable, the court must have regard to the explanation for the delay, which must cover the whole period. Without an explanation, the delay is unreasonable. (In putting the matter in this way in para 52 of Buffalo City, Theron J obviously intended ‘delay’ to mean a period longer than that which would be required by a litigant acting promptly and facing no obstacles.)

[290] The approach in deciding whether to overlook an unreasonable delay is a flexible one, grounded in the proven facts and objectively available considerations, and taking into account various factors:

(a) The first is potential prejudice to affected parties, the possible consequences of setting aside the impugned decision, and the possible amelioration of prejudicial consequences by the granting of a just and equitable remedy.

(b) The second is the nature of the impugned decision. This requires a consideration of the merits of the legality challenge and of the extent and nature of the illegality.

(c) The third is the conduct of the applicant. State actors are subject to a higher duty to respect the law, including a duty to act promptly when rectifying illegality. However, even where the public functionary has not acted as a model litigant, unreasonable delay may be overlooked if the functionary acted in good faith or with the intent to ensure clean governance. There is a difference between muddle and malice.

(d) The fourth factor (the Gijima principle, after the decision in State Information Technology Agency SOC Limited v Gijima Holdings (Pty) Limited [2017] ZACC 40; 2018 (2) SA 23 (CC)) is that even where there is no basis to overlook an unreasonable delay, the court may be compelled, by virtue of s 172(1)(a) of the Constitution, to declare the impugned decision unlawful. In favour of the Gijima principle is that the court should be slow to allow procedural obstacles to prevent scrutiny of a challenge to the exercise of public power. In order, however, not to undermine the valuable rationale behind the rules on delay, the principle is a narrow one – it applies only where the unlawfulness of the impugned decision ‘is clear and not disputed’. (See also, on the Gijima principle, Notyawa v Makana Municipality & others [2019] ZACC 43; 2020 (2) BCLR 136 (CC) paras 48-52.)

[291] OUTA’s counsel referred me to Valor IT v Premier, North West Province & others [2020] ZASCA 62; [2020] 3 All SA 397 (SCA). That was also a legality self-review. There had been a very substantial delay, and the explanation for it was wanting. Plasket JA said (para 30) that whether an unreasonable delay should be condoned involved a ‘factual, multi-factor and context-sensitive’ enquiry in which a range of factors are considered and weighed. Prospects of success are relevant when deciding whether to overlook delay. Strong prospects may, up to a point, excuse an inadequate explanation (para 38). In view of the strong prospects of success in that case, the court granted condonation (para 39).

[292] Reading Buffalo City and Valor IT together, one can say that in some cases (such as Valor IT) good prospects of success will simply be a tilting factor in the ‘multi-factor and context-sensitive’ balancing exercise of deciding whether to overlook an unreasonable delay while in other cases (such as Buffalo City) the prospects will be so strong that the Gijima principle finds operation.

[293] In a post-hearing communication, I was alerted to the judgment in Altech Radio Holdings (Pty) Ltd & others v City of Tshwane Metropolitan Municipality [2020] ZASCA 122, another legality self-review. There the delay was not as long as in Valor IT or in the present case, but the SCA held that the court a quo had erred in overlooking it. Important considerations in the balancing exercise were that the City’s review lacked merit and that its vacillation had caused the contractor and its financiers to layout more than R610 million. There was an ongoing contract for the performance of work, and the contractor was continuing to perform its obligations throughout the delay period, at times being exhorted by the City to do so.

[294] Although not mentioned in Buffalo City, I consider that the explanation offered for the delay, even if it is not adequate to avoid a conclusion of unreasonableness, may nevertheless feature in deciding whether to overlook it.

PAJA delay

[295] The PAJA delay rule is contained in s 7(1) read with the power of condonation in s 9(1). A delay of less than 180 days could on the facts be found to be unreasonable. A delay of longer than 180 days is per se unreasonable, and the question becomes one of condonation.

[296] In terms of s 9(2), the touchstone for condonation is the interests of justice. In Aurecon South Africa (Pty) Ltd v City of Cape [2015] ZASCA 209; 2016 (2) SA 199 (SCA) (‘Aurecon SCA’) Maya P said (para 17) that whether it is in the interests of justice to condone an unreasonable delay depends on the facts of each case. Relevant factors generally include: (a) the nature of the relief sought; (b) the extent and cause of the delay; (c) its effect on the administration of justice and other litigants; (d) the reasonableness of the explanation for the delay, which must cover its whole period; (e) the importance of the issue to be raised; and (f) prospects of success.

[297] It will be observed that whereas in legality reviews the adequacy of the explanation is central to the question whether the delay was unreasonable, in PAJA cases the adequacy of the explanation is usually relevant to the question of condonation, since in most PAJA delay cases non-compliance with the 180-day period leads to the delay being per se unreasonable.

[298] Although the Gijima principle has been formulated in legality reviews, I do not agree with CTSA’s submission that it does not apply to PAJA reviews. While I accept the principle of subsidiarity, PAJA must be construed as far as reasonably possible in a way consistent with s 172(1)(a) of the Constitution (Bengwenyama Minerals (Pty) Ltd & others v Genorah Resources (Pty) Ltd & others [2010] ZACC 26; 2011 (4) SA 113 (CC) para 82). Just as in a legality review the Gijima principle may require a court to overlook an unreasonable delay, just so in a PAJA review it may require the court, in the interests of justice, to extend the period for review in terms of s 9(2) of PAJA. There is no rational basis to treat the two cases differently.

The delay in this case

The trigger date

[299] There was some debate about the trigger date. The matter is complicated because multiple decisions are impugned.

[300] Dealing first with SFF’s knowledge, I consider that Gamede’s knowledge should not be attributed to SFF, given my conclusion that he acted for purposes he knew to be improper (R v Kritzinger 1971 (2) SA 57 (A) at 59H-60F; NBS Bank Ltd v Cape Produce Company Pty Ltd & others [2002] 2 All SA 262 (A) para 34; Mostert NO v Old Mutual Life Assurance Co (SA) Ltd (2) [2001] 4 All SA 250 (A) para 65). However, the knowledge of his senior management team would in my opinion be sufficient, and the board’s knowledge would obviously suffice.

[301] By 23 November 2015 the board had knowledge of the First Approval Notice and of Gamede’s motivation to the MoE. Time in respect of SFF’s review of the First Approval Notice thus began to run by not later than 23 November 2015.

[302] On 19 January 2016 SFF’s exco learnt that Gamede had sold the strategic stock. Although Mayaphi may not have been as shocked as the others, I cannot find that the senior management team as a whole knew of the disposals before this date. Time thus began to run on 19 January 2016 in SFF’s review of Gamede’s award of volumes of oil to Venus, Taleveras and Vitol.

[303] There is no evidence that as at 19 January 2016 SFF’s exco was aware of the Second Approval Notice. Mayaphi made no mention of it in his memorandum of 13 January 2016. Furthermore, as at 19 January 2016 SFF did not have knowledge of the terms of the contracts. The applicants say that Gamede did not reply to Ngqongwa’s request for the commercial terms. However, the exco could by exercising reasonable diligence have obtained copies of the contracts, if necessary by asking the board to instruct Gamede to make them available.

[304] The board learnt of the awards and contracts on 5 February 2016. One does not know exactly how much the board knew, but they knew enough to instruct Gamede to furnish them with any further information they needed to assess the contracts. If the trading report did not mention the Second Approval Notice, an appropriate enquiry as to whether the MoE had approved the transactions would soon have elicited a copy. From a practical perspective, therefore, 5 February 2016 can be taken as the trigger date for the most important impugned decisions (the conclusion of the contracts).

[305] Since SFF should, from not later than 5 February 2016, have closely monitored Gamede’s further actions, time began to run in respect of Gamede’s conclusion of subsequent contracts and amendments from the dates they were concluded. In fact, if SFF’s board had adopted the position they have subsequently taken in the review application, everything should have been stopped in its tracks there and then. Since the amendments are ancillary matters and would not add more than two to three weeks to periods reckoned from 5 February 2016, I shall disregard them.

[306] The applicants have not sought to distinguish CEF’s knowledge from SFF’s. They say that CEF Treasury became aware of the sale of the oil reserves on 26 February 2016, Ngqongwa being the treasury official in question. Ngqongwa knew of the disposals by 19 January 2016. Furthermore, in my view SFF’s board was under a clear duty, on 5 February 2016, to notify CEF of the unexpected developments. The applicants have not disclosed when CEF’s board learnt of the disposals. The information was peculiarly within their knowledge.

[307] In the circumstances, the applicants cannot justify a later trigger date for CEF than for SFF. For the sake of completeness, however, I may add that CEF definitely had the relevant knowledge by the time it issued its defensive press release on 26 May 2016. Accordingly, if this were taken as CEF’s trigger date, time began to run for CEF four months later than for SFF.

The extent of the delay until institution of application

[308] In regard to PAJA, 180 days from 5 February 2016 takes one to early August 2016. The review was launched on 12 March 2018, about two years and one month from the trigger date and nearly 18 months after expiry of the 180 days. This is per se unreasonable, and condonation is needed. In regard to legality, a period of two years and one month from the trigger date would plainly be an unreasonable delay unless there were a satisfactory explanation covering the whole period.

The explanation for delay

[309] I thus turn to the explanation offered for the delay. In the PAJA review, this is one of several factors relevant to condonation. In the legality review, it is central to the question whether the delay was unreasonable.

[310] Between February and June 2016 the reaction of the applicants’ boards and senior management seems to have been to defend the disposals. While one or two members of senior management may have been tainted, it is not the applicants’ case that entire management teams or boards were tainted. Reconstitution of senior management and the boards was not a necessary precursor to investigation. (The present case is thus distinguishable from Passenger Rail Agency of South Africa v Swifambo Rail Agency (Pty) Ltd 2017 (6) SA 223 (GJ) and Eskom Holdings SOC Limited v McKinsey and Company Africa (Pty) Ltd & others [2019] ZAGPPHC 185, on which the applicants’ counsel placed reliance.) The failure of the boards to insist on a rigorous investigation during this initial five-month period is unexplained and unreasonable.

[311] In late June 2016, seemingly only after pressure from the MoF and MoE, the applicants decided to commission a legal review. The evidence does not disclose when they engaged A&O, but we do know that the latter’s review was only concluded in December 2016. Since the applicants have claimed privilege, we do not know what work A&O did. A legal review did not require six months. It was a matter of obtaining and examining relevant minutes, directives and contracts, and identifying applicable statutory provisions. Not more than two months was needed. Either the applicants delayed in instructing A&O or did not insist on the expeditious performance of the mandate.

[312] Senior counsel was briefed in December 2016 and furnished his opinion on 10 February 2017. Given the importance of the matter, it was prudent to seek counsel’s opinion. In light of the intervening festive period, there was no unreasonable delay in the furnishing of the opinion.

[313] The applicants have claimed privilege over counsel’s opinion, but I am sure that A&O and senior counsel impressed upon the applicants the need to institute proceedings promptly. But four and a half months were to pass before KPMG was engaged on 22 June 2017. Assuming that the review could reasonably be held back pending an accounting report, this delay has not been explained.

[314] I am not satisfied, however, that it was reasonable to delay the review in order to get an accounting report. The applicants do not say that by February 2017 they were in any doubt about the illegality of the transactions. The question as to what remedy should follow was not something which had to delay proceedings. After all, when the applicants eventually launched in March 2018, they left the question of an appropriate remedy open for later determination, and did not attach the accounting reports or even address the questions canvassed in them.

[315] However, on the assumption that an accounting report was needed, KPMG produced its report on 25 July 2017, about one month after its engagement. That was not unreasonable. Nor was there any delay in obtaining a second counsel’s opinion arising from the KPMG report.

[316] However, the applicants then delayed for two and a half months before briefing PwC on 12 October 2017 for a second accounting report, which the latter furnished on 7 November 2017. Although PwC discharged its mandate promptly, I do not think that the obtaining of a second report is a justification for further delay. The fact that certain personnel in KPMG were caught up in unsavoury press coverage did not mean that KPMG as a firm could not produce a competent and professional report. The applicants do not say that KPMG’s report was unsatisfactory.

[317] Although settlement discussions were held in the period November 2017-January 2018, the respondents’ evidence is that these were neither time-consuming nor productive. The applicants did not seek agreement that time would not run during the discussions. Review papers could have been prepared in parallel with talks.

[318] My conclusion is that the delay was substantial and most of it unexplained or not satisfactorily explained. For legality purposes, the delay was unreasonable. For PAJA purposes, the inadequacy of the explanation is a factor against condonation.

Delay post-institution of proceedings

[319] Although the application was launched on 12 March 2018, the respondents only finally knew the case they had to answer when the applicants served their supplementary founding papers on 28 February 2020, more than four years after the trigger date. The papers served in February 2020 were far more extensive than the initial papers. This further delay is also important (see Mkhwanazi v Minister of Agriculture and Forestry, KwaZulu 1990 (4) SA 763 (D) at 766F-G). Delay is concerned with prejudice to interested parties and prejudice in the administration of justice. If an applicant, when launching, goes off half-cock, the institution of the application does not achieve the purpose contemplated by delay rules. One can only say that an applicant has done what the delay rule envisages when it has presented its whole case, so that the respondents can progress the matter by answering it.

[320] In the present case, the notion that the applicants had to furnish a rule 53 record after launching the application is fanciful. The applicants should have been in a position to tender their full record when they launched. The MoE produced no further documents. Accordingly, the production of a record cannot justify delay beyond 12 March 2018.

[321] The first reason for post-launch delay is that the applicants decided to commission yet another forensic investigation, this time by Gobodo, starting with a tender invitation on 7 May 2018. If the applicants believed that further documents might be traced or recovered from computers (this is mainly what Gobodo added to the information already available), this further investigation should have been commissioned much earlier, and should have been completed before 12 March 2018. Instead, Gobodo was only appointed on 27 July 2018 and only submitted its final report on 30 April 2019.

[322] Although up to that time the applicants had intimated that their supplementary founding papers would be filed once the forensic investigation was complete, they thereafter adopted the stance that they would not do so until the disclosure appeal was abandoned or finalised. That appeal was resolved in a judgment delivered on 13 December 2019. It took the applicants two and a half more months to deliver the supplementary founding papers.

[323] The second reason for post-launch delay is the disclosure application and appeal. CTSA’s counsel submitted that this was not a legally justifiable basis for the applicants to delay filing their final supplementary founding affidavit, citing Potpale Investments (Pty) Ltd v Mkhize 2016 (5) SA 96 (KZP) paras 18-23, which has been followed in later cases. I accept that the rule 35(12) notices did not suspend the running of time. However, Glencore and CTSA’s stance (inconsistently with Potpale) was that they would not file their answering papers until the disclosure applications and appeals were determined. The practical reality is that the review would not have been ripe for hearing until the disclosure appeal was decided and until all parties thereafter had an opportunity to supplement their papers in the light of the further documents.

[324] If the Gobodo report had been commissioned at an earlier time and the information incorporated in the founding papers of March 2018, the filing of answering papers would still have been substantially delayed because of the disclosure applications and appeal. The same is true if the founding papers had been promptly amplified after Gobodo produced its report in April 2019. However, the respondents would have been able to substantially complete their answering papers while awaiting the outcome of the appeal. Following the SCA’s judgment of 13 December 2019, those papers could have been finalised and served by the end of January 2020. Instead, the respondents had to wait until the end of February 2020 to receive the applicants’ supplementary founding papers, which in turn meant that their answering papers were only filed on 22 May 2020. The bulk of the supplementary founding affidavits related to matters unconnected with the disclosure appeal. Accordingly, the post-launch delay in supplementing the founding papers can, at least to the extent of about four months, be held against the applicants.

[325] The supplementary founding papers served at the end of February 2020 did still not represent the applicants’ complete case. Only after several further case management meetings was the Gobodo report made available on 20 April 2020. There was also much new matter in the replying papers filed on 22 July 2020, hence the need for the respondents to file supplementary opposing papers.

[326] Forming part of the replying papers were the expert reports of Mr Ara Barsamian (a chemical engineer and crude oil blending expert) and Mr John Driscoll (an oil trading expert). Mr Barsamian’s report included a scathing critique of the Gamede memoranda which Jawoodeen sent to the MoE in late June 2016. Those memoranda had formed part of the original founding papers, and the expert analysis thereof should have been in the founding papers. Mr Driscoll’s report covered a wide range of issues that should have been in the founding papers.

[327] Although en passant these reports provided material, here and there, for reply, neither Mr Barsamian nor Mr Driscoll framed their reports as responses to the respondents’ affidavits. A theme which became prominent in the applicants’ argument, viz that an assay of the oil should have been part of a due diligence preceding disposal, was raised for the first time in reply on the strength of Mr Barsamian’s report.

[328] Also forming part of the replying papers were affidavits by Jawoodeen and Ngqongwa dealing with the lateness and inadequacy of the information placed before the board on 5 February 2016. Again, this should have been in the founding papers.

The Gijima principle in this case

[329] Despite gross delay which has not been satisfactorily explained, the Gijima principle requires me to declare the impugned decisions invalid if their illegality is ‘clear and not disputed’. Although the respondents do not concede all the grounds of review, they admit that on at least some grounds the decisions were invalid. In my view, the remaining grounds, even if not admitted, have been clearly established.

[330] It follows that the delay must be condoned (for PAJA purposes) or overlooked (for legality purposes) so as not to stand in the way of a declaration of illegality. However, the lengthy delay and unsatisfactory explanations will necessarily feature prominently in the formulation of just and equitable relief.

[331] Although the Gijima principle compels this result, there are other factors which justify condoning/overlooking the delay, at least for the limited purpose of a declaration of invalidity. The decisions related to the national interest in energy supply and to a quantity of oil of large value. The oil was sold for $280,831,000, equating to R3,317 billion in March 2018 and R4,36 billion in November 2020. The illegalities were serious and pervasive with wholesale disregard of corporate governance and transparency. In the case of Taleveras, the irregularities included bribery, while in Venus’ case there must have been an improper basis for its being favoured.

[332] Commercial prejudice to the respondents is the main counterweight to the foregoing considerations. Although delay can also cause litigation prejudice, it does not feature in this case. Because the Gijima principle finds application, the respondents’ prejudice is relevant to a just and equitable remedy rather than to condonation.

Just and equitable remedy – legal principles

[333] Following upon a declaration of invalidity, s 172(1)(b) empowers the court to

‘(b) may make any order that is just and equitable, including-

(i) an order limiting the retrospective effect of the declaration of invalidity; and

(ii) an order suspending the declaration of invalidity for any period and on any conditions, to allow the competent authority to correct the defect.

[334] This jurisdiction is also conferred by s 8(1) of PAJA, though the latter Act contains a different list of orders which the court may grant, among them being orders

‘(c) setting aside the administrative action and –

(i) remitting the matter for reconsideration by the administrator, with or without directions; or

(ii) in exceptional cases –

(aa) substituting or varying the administrative action or correcting a defect resulting from the administrative action; or

(bb) directing the administrator or any other party to the proceedings to pay compensation;

The width of the court’s equitable jurisdiction

[335] That s 172(1)(b) clothes the court with very wide powers is plain from its language. In Electoral Commission v Mhlope & others [2016] ZACC 15; 2016 (5) SA 1 (CC), the Chief Justice said (para 132):

‘Section 172(1)(b) clothes our courts with remedial powers so extensive that they ought to be able to craft an appropriate or just remedy even for exceptional, complex or apparently irresoluble situations. And the operative words in this section are “an order that is just and equitable”. This means that whatever considerations of justice and equity point to as the appropriate solution for a particular problem, may justifiably be used to remedy that problem. If justice and equity would best be served or advanced by that remedy, then it ought to prevail as a constitutionally sanctioned order contemplated in section 172(1)(b).’

The default position – the corrective principle

[336] The default position, where conduct is declared invalid, is the corrective principle, viz that the consequences of invalidity must be reversed. In the procurement context, this would entail setting aside the implicated contracts (Allpay Consolidated Investments Holdings (Pty) Ltd & others v Chief Executive Officer, South African Social Security Agency & others (2) [2014] ZASCA 12; 2014 (4) SA 179 (CC) paras 29-33).

[337] The applicants ask me to apply the corrective principle, and they regard restitution of the purchase price and storage fees as an inevitable consequence. At least in most cases, that seems to be right. In Allpay (2) Froneman J considered remedial action in the form of the corrective principle to be a logical consequence flowing from invalid and rescinded contracts and enrichment law generally (para 30). Restitution does not necessarily make the other party ‘whole’, because the latter may have spent money in anticipation of further performance, hence the possibility of loss flowing from the corrective principle.

[338] The restitution of the purchase prices and storage fees is arguably not ‘compensation’ for purposes of PAJA but a legal consequence of setting aside the impugned transactions. Although restitution may be an ex lege obligation, it can properly be made the subject of a just and equitable remedy for the avoidance of doubt.

Departing from the corrective principle

[339] In the present case, the applicants’ delay is one important factor in assessing whether and to what extent there should be a departure from the corrective principle. Delay was the primary reason for the Constitutional Court preserving accrued contractual rights in Gijima and Buffalo City.

[340] Another relevant factor is the prejudice which affected parties would suffer if the corrective principle were applied. Where the prejudice has been caused as a result of the applicant’s delay, prejudice and delay operate in tandem as factors potentially justifying a departure from the corrective principle. However, prejudice may also be a free-standing consideration. In the present case, some of the harm which the respondents stand to suffer if the corrective principle were applied is harm which became inevitable as soon as the impugned transactions were concluded, while other harm was caused or exacerbated by delay.

[341] Also relevant is the extent to which the respondents were guilty of blameworthy conduct. Most obviously this applies to any respondent which was complicit in SFF’s wrongdoing. It might also be relevant that a respondent turned a blind eye to possible malfeasance without making proper enquiry. If, on the other hand, a party was innocent, and even more so if the party was proactive in striving to behave properly, this would be a factor in favour of departing from the corrective principle.

[342] In Swifambo, for example, the court held that the corrective principle should be applied, having regard to the fact that Swifambo had not been an innocent tenderer but had engaged in illicit fronting. This decision was upheld on appeal: [2018] ZASCA 167; 2020 (1) SA 76 (SCA).

[343] In other cases the tenderer’s conduct might be such that it would not be just and equitable to require it to suffer a loss but that it would be just and equitable that it should likewise not derive a benefit. This was the Constitutional Court’s assessment of the tenderer’s position in Allpay (2) para 67 (and see also Black Sash Trust v Minister of Social Development & others (Freedom Under Law NPC intervening) [2017] ZACC 8; 2017 (3) SA 335 (CC) paras 40 and 50).

[344] In yet other cases, the party’s innocence coupled with other considerations may cause the court to allow the contract to stand, in which event the party is not only saved from being out of pocket but is allowed to keep its contractual profit. Gijima and Buffalo City are examples.

[345] The applicants’ conduct, including but not limited to delay, is also germane. Although egregious irregularities by a self-reviewing applicant may compel a court, in terms of the Gijima principle, to declare the impugned conduct invalid, such misconduct may also tilt the balance in favour of a just and equitable remedy which is more generous to an innocent respondent.


[346] CTSA and Vitol’s primary contention is that the corrective principle should not be applied in this case. If that contention succeeds, the court would declare the relevant conduct and transactions invalid but would not set them aside, alternatively would preserve accrued rights.

[347] CTSA and Vitol contend, in the alternative, that if the impugned transactions are set aside, they should, in addition to restitution, be awarded compensation for their losses. Section 8(1)(c) of PAJA explicitly makes provision for compensation. Although s 172(1)(b) of the Constitution does not, a compensation order is within the wide language of the section.

[348] The interpretation of s 8(1)(c) was discussed in Trustees of the Simcha Trust v De Jong & others [2015] ZASCA 45; 2015 (4) SA 229 (SCA), which was an appeal from my judgment in De Jong & others v The Trustees of the Simcha Trust & another 2014 (4) SA 73 (WCC). These judgments hold inter alia (a) that for purposes of PAJA, a just and equitable remedy in the form of compensation may be awarded to a respondent in review proceedings, not only to an applicant; (b) that the awarding of compensation is only available where the court has neither remitted the matter to the decision-maker nor issued a substituted decision. In the present case there is no question of a remittal or substituted decision. Accordingly, in principle this court’s just and equitable remedy could include an order for the applicants to pay the respondents compensation.

[349] Although ‘compensation’ could notionally extend to loss of profit, CTSA and Vitol do not seek compensation for lost profit if their primary case (that the contracts should stand) fails. They seek compensation for losses in the form of money fruitlessly spent (out-of-pocket expenses).

[350] An important consideration in the context of compensation is this. Although the irregularities here were a serious violation of the rule of law, ultimately the prejudice suffered by the applicants, and by the public interest they represent, is financial. It has always been within the power of SFF to replenish its stock of 10 million barrels, the only problem being one of price. The oil price started going up after January 2016, and has remained above the SPA selling prices except for about two months between March and May 2020 when Covid-19 caused a tumble. It is legitimate to ask whether it is just and equitable that financial loss should fall on the respondents or whether SFF should have to shoulder all or some of the burden.

Need for counter-claim?

[351] The applicants argue that the court cannot grant compensation as a constitutional/PAJA remedy in the absence of a counterclaim. In the alternative, the applicants submit that in assessing whether the respondents should be awarded compensation, the court should treat them effectively as claimants, with the consequence that the Plascon-Evans rule should operate against them in regard to disputes of fact.

[352] In my view, it is important to distinguish the role of compensation in the present case from the case where a party claims relief for a violation of its constitutional rights. In the latter situation, the party would be a claimant in relation to such relief. If for any reason the party already featured as a respondent, it would need to bring a counter-application to assert the violation of its rights and claim appropriate relief.

[353] In the present case, it is the applicants who claim a violation of the principles of legality and administrative justice. On their case, the party injured by these violations is SFF. The relief they claim is a setting aside of various decisions and contracts. Vitol and CTSA have not brought proceedings alleging that their rights have been infringed. On the contrary, their primary case is that the status quo should be preserved.

[354] As a fallback, Vitol and CTSA contend that if the court sets aside the various decisions and contracts, the harsh effects of such an order should be ameliorated by compensation. The prejudice which Vitol and CTSA stand to suffer is not prejudice arising from the fact that the various decisions and contracts were taken and concluded in violation of their constitutional rights (as I have said, they are quite content to allow those decisions and contracts to stand) but prejudice arising from an order which this court might make to remedy the injury done to SFF.

[355] In such a case, so it seems to me, the question of compensation is an integral element of the assessment of whether setting-aside orders should be granted and if so on what terms. Compensation of this kind, ie as an antidote to the harshness of a setting-aside order, may well not be the sort of compensation contemplated in s 8(1)(c) at all, instead being sourced in the broad just and equitable jurisdiction conferred by the Constitution and PAJA. The applicants’ argument is that, in the absence of a counter-application by the respondents, the court is faced with a binary choice: set aside the decisions and contracts, or allow them to stand. That would turn the court’s just and equitable jurisdiction into a blunt tool rather than the flexible instrument envisaged by the Chief Justice in the passage I quoted earlier from Mhlope.

Constitutional damages?

[356] For similar reasons, compensation here is conceptually different from constitutional damages considered in cases such as Fose v Minister of Safety and Security 1997 (3) SA 786 (CC) and Modderfontein Squatters, Greater Benoni City Council v Modderklip Boerdery (Pty) Ltd (Agri SA and Legal Resources Centre, amici curiae) 2004 (6) SA 40 (SCA). In those cases constitutional damages represent relief in favour of a party complaining that his or her fundamental rights have been violated. In Fosie, constitutional damages were claimed in addition to delictual damages. The claim was under s 7(4)(a) of the interim Constitution, the equivalent of ‘appropriate relief’ under s 38 of the current Constitution.

[357] In the present case, one is dealing with a self-review. The aggrieved parties are the applicants, not the respondents. None of the respondents claim that their fundamental rights have been violated. The question of compensation arises as a potential antidote to ameliorate the harsh consequences of the remedy which the applicants claim.

Delictual liability?

[358] The applicants’ counsel submitted that if the impugned transactions were set aside, the respondents might have delictual claims for damages, and that it would be preferable to allow recovery of losses to be adjudicated in the context of such claims than as an exceptional constitutional/PAJA remedy. I disagree. I cannot determine in the present proceedings whether delictual claims would succeed. The courts are reluctant to impose delictual liability for the improper but honest exercise of powers (Steenkamp NO v Provincial Tender Board, Eastern Cape 2007 (3) SA 121 (CC) paras 38-56; South African Post Office v De Lacey & another 2009 (5) SA 255 (SCA) para 14), which might apply to the MoE and SFF’s board. To the extent that Gamede was dishonest, questions would arise as to whether SFF is vicariously liable.

[359] This case is distinguishable from many review cases, where the question is simply whether an administrative decision should stand. Here, contracts have actually been concluded and have been in existence for some years. I am not called upon to decide merely whether certain decisions should be set aside but whether these contracts should be nullified. The respondents might or might not have delictual claims if their contracts are nullified, but they certainly have contractual claims unless their contracts are set aside.

[360] Since I cannot prejudge delictual liability, I need to determine just and equitable relief in the light of the possibility that, as a matter of law, delictual claims would be found not to be competent. The applicants certainly do not accept that delictual claims are available. Since the respondents might not receive compensation through delictual claims, I need to consider whether, in the light of all relevant circumstances, I should allow the contracts to stand (so that the respondents can exercise their contractual rights) or whether I should set them aside. In striking that balance, it seems to me that I must be entitled, if justice and equity so demand, to set the contracts aside but on the basis that the respondents receive compensation, to an equitable extent, for the deprivation of their contractual rights. If this course were not available to me, it might strongly tilt the balance in favour of allowing the contracts to stand.

Other common law remedies

[361] Counsel for the applicants and OUTA spoke about contractual and unjust enrichment remedies. If the contracts are set aside, the respondents will not have contractual remedies against SFF. The extent to which CTSA and Glencore would have contractual remedies against Taleveras and Venus is something I shall discuss presently. Unjustified enrichment would not assist CTSA, Glencore and Vitol to recover out-of-pocket losses.

Direct and indirect parties

[362] Unlike Vitol, Glencore and CTSA did not buy oil from SFF but from Venus and Taleveras respectively. Two interrelated questions arise. First, can an ameliorating just and equitable remedy be made in favour of Glencore and CTSA, bypassing Venus and Taleveras? Second, is the prejudice suffered by Glencore and CTSA, as distinct from the prejudice suffered by Venus and Taleveras, relevant in determining the remedy that would be just and equitable?

[363] As I have said, s 172(1)(b) of the Constitution and s 8(1) of PAJA confer a very wide jurisdiction. The unlawful exercise of public power has the capacity to harm persons in addition to those who are its direct targets or subjects. In my view, any person who stands to suffer harm if an impugned decision is set aside may in principle be the beneficiary of a just and equitable remedy. In principle, this could include granting a restitutionary or compensatory remedy in favour of Glencore and CTSA. Similarly, the potential of the unlawful exercise of public power to harm persons in addition to those who are its direct targets or subjects means that in deciding whether to condone delay or in determining a just and equitable remedy, a court should have regard to the prejudice suffered or likely to be suffered by such other persons.

[364] Glencore and CTSA, although they were not direct parties in the purchase of the oil from SFF, were not distant from SFF. The latter knew that Venus and Taleveras were immediately on-selling the oil to Glencore and CTSA. There were tripartite agreements involving Glencore and CTSA in which SFF acknowledged the transfer of ownership of oil to them and pursuant to which SFF issued tank warrants to them. The Glencore tripartite agreement incorporated SFF’s right to repurchase the oil. These tripartite agreements are among those which the applicants seek to have set aside on review.

[365] In relation to restitution, the applicants’ counsel submitted in their heads of argument that CTSA should look to Taleveras and that Glencore should look to Venus (though this is inconsistent with the subsequent settlement between the applicants and Glencore which proposes the payment of restitution directly to Glencore). The remedies which CTSA and Glencore respectively have against Taleveras and Venus are certainly relevant considerations, but they do not bar the granting of restitutionary and compensatory relief to CTSA and Glencore.

The remaining factors

[366] I have already discussed delay. In the next three sections of this judgment I deal with the respondents’ financial prejudice and the question of reprehensible behaviour by the respondents on the one hand and by the applicants on the other. Thereafter I shall draw the threads together in arriving at my decision on a just and equitable remedy.

Commercial prejudice to respondents


[367] Venus has not participated in the proceedings and has thus not put up prejudice as a basis for the refusal of any relief claimed by the applicants. Nevertheless, a measure of prejudice to Venus is self-evident from the facts. If the impugned decisions were set aside on the basis that SFF retains ownership of the oil, Venus would, at a minimum, have to refund the purchase price of $100,761,000 and storage fees of $17,200,075 it received from Glencore. Venus could also face additional claims from Glencore for contractual damages.

[368] If, as the applicants propose, the amounts received by SFF from Venus ($90,225,000 plus $7,458,371) were refunded either to Venus or to Glencore, Venus would still have to refund Glencore the differences on the purchase price and storage fees ($10,536,000 plus $9,741,704). These differences are Venus’ profit, so if it had to refund them to Glencore, Venus’ position would be neutral.


[369] Glencore paid $10,536,000 more for the oil than SFF received from Venus, and paid storage fees of $9,741,832 in excess of those SFF received from Venus. SFF would have been aware that Venus was charging a mark-up but would not have known its extent. Since the purchase price was settled upfront, delay on the applicants part did not affect the extent of Glencore’s loss of $10,536,000, though delay could have affected Glencore’s ability to recover the shortfall from Venus. Delay certainly affected the storage fees, since these were payable monthly.

[370] Glencore paid inspection costs of $29,887 over the period April 2016-November 2018. If the applicants had not delayed, these inspection costs would have terminated sooner.

[371] Glencore incurred a cost of $43,630 to establish a letter of credit. Since this was incurred upfront, it would not have been avoided by prompt review proceedings.

[372] Glencore insured the oil over the period January 2016-December 2019. It says its insurance costs totalled $631,050 though the debit notes attached to support this assertion total $614,355. Glencore has continued to insure the oil to cover the eventuality of the applicants’ case failing. Clearly this is an ongoing cost the extent of which will have been materially affected by delay..

[373] Glencore’s suffered hedging losses totalling $113,729,139 incurred when it unwound its hedges in June and July 2018. The hedges were taken out to cover the risk that oil prices would not increase as Glencore was anticipating. The market rose in accordance with Glencore’s expectation, which meant that although it was losing on the hedges, it was making on the oil. When the market changed from contango to backwardation in August 2017, Glencore gave notice to terminate the SGA at the end of December 2017, continuing to roll over the hedges until then.

[374] In the latter part of 2017 SFF alleged that the SFF/Venus transactions were invalid. Since there was no certainty that the applicants would adhere to this position or institute review proceedings, Glencore considered it unwise to terminate its hedge straightaway. In 2018 the market continued in backwardation, meaning that Glencore’s hedging losses kept growing. Accordingly, and a couple of months after the institution of the review proceedings in March 2018, Glencore decided to unwind the hedges. The prudence of Glencore’s conduct is supported by an expert report from Ms Jago, and the applicants do not criticise the hedging strategy or the timing of the decision to unwind it.

[375] There is no doubt that the applicants’ delay created a predicament for Glencore. It so turns out that the sooner the applicants had adopted a firm position and instituted proceedings, the less costly would unwinding the hedges have been. With hindsight, one can see that unwinding the hedges in June/July 2018 was particularly expensive.

[376] Glencore alleges that it incurred ‘clearing costs’ of $213,692. According to Ms Jago’s report, these were routine costs incurred in executing the hedging strategy. Since hedges were rolled over from time to time, the extent of these costs was increased by delay.

[377] In summary, Glencore’s out-of-pocket expenses total $232,394,781, of which $97,683,371 is covered by the tender of restitution, the balance being $134,711,410.


[378] Taleveras is willing to settle for the applicants’ proposal of restitution. Taleveras paid SFF $112 million for the oil but received a net amount from CTSA of $134,568,426, leaving it with a ‘profit’ of $22,568,426. If the transactions are set aside on the basis that SFF remains the owner of the oil, and if SFF refunds $112 million to Taleveras or to CTSA, Taleveras would still be exposed to a claim from CTSA to refund the difference of $22,568,426. As with Venus, if Taleveras repaid this amount, its position would be neutral – no profit, no loss. But as with Venus, Taleveras could perhaps face additional claims from CTSA for contractual damages.

[379] Unlike Venus, Taleveras did not charge a storage fee mark-up to CTSA, so if applicants refund the storage fees to Taleveras or to CTSA, Taleveras will not face any additional refund claim from CTSA in respect of storage fees.


[380] Since CTSA paid the purchase price due to SFF for Taleveras’ oil ($112 million), this represents an actual outgoing. If the applicants’ tender of restitution results in payment to CTSA, this financial prejudice would be eliminated.

[381] In principle the same applies to the storage fees. According to CTSA, it paid $12,345,600 for the months January 2016-January 2018.[55] According to PwC’s most recent report, SFF received only $11,865,600.[56] The difference of $480,000 is represented by a payment CTSA made to Taleveras in January 2016. Since CTSA cannot prove that Taleveras on-paid this amount to SFF, I must accept that the applicants’ version. The applicants’ proposed storage refund would thus leave CTSA $480,000 out-of-pocket.

[382] If the applicants’ proposed refunds were paid to Taleveras rather than CTSA, the latter would be out of pocket unless Taleveras in turn refunded the money to CTSA. CTSA has put up evidence calling the Taleveras group’s financial soundness into question. Judgments and freezing orders have been granted against Taleveras companies and Sanomi in various jurisdictions. It is not possible to reach a definite conclusion, but there is a fair risk that money refunded to Taleveras would not find its way back to CTSA.

[383] After settling the purchase price of $112 million, CTSA paid Taleveras and net amount of $22,568,426. This represents an actual outgoing which is not covered by the applicants’ tender. Since the amount was paid out at the commencement of the transaction, it would not have been avoided or reduced by prompter action from the applicants.

[384] In terms of the MPA, Taleveras was obliged to repurchase the oil on 5 April 2018. To mitigate the risk that Taleveras would not do so, CTSA concluded a put option agreement with Total Oil Trading SA (‘Total’) in terms of which the latter would be obliged to buy the oil if CTSA gave due notice. Total charged CTSA an option fee of $800,000 (the ‘Offtake Premium’ previously mentioned as one of the transaction costs). This cost was incurred on 5 February 2016, and would not have been avoided by prompter review proceedings.

[385] CTSA incurred insurance costs and inspection costs. CTSA’s insurance costs totalled $408,853 by way of monthly premiums paid over the period February 2016-March 2018, while the inspection costs totalled $28,195 by way of monthly debits over the period March 2016-February 2018. These amounts could have been reduced if the applicants not delayed.

[386] CTSA incurred a hedging loss of $83,680,000 incurred when its hedges expired on 6 April 2018. Dated Brent at expiry was $65.903 – this was the price with reference to which Taleveras or Total would have been obliged to buy back the oil from CTSA. This was thus also the float rate of the CTSA hedge, while the swap rate replicated the MPA’s forward price of $44.98. This meant that CTSA had to pay Natixis GMC the difference of $20.92 ×4 million = $83,680,000, and Natixis GMC had to pay an identical amount to the external counterparty.

[387] If SFF had not repudiated the transactions, CTSA would still have incurred this hedging loss of $83,680,000. However, because it would have had the physical oil, CTSA on 6 April 2018 would have sold the 4 million barrels to Taleveras, Total or to a third party for a market price determined with reference to a Dated Brent value of $65.90/bbl. In the papers, CTSA refers to an amount of $65.903 × 4 million = $263,612,000. However, the repurchase price was stated to be based on Dated Brent for March 2018 ‘minus haircuts not paid/utilised’. In the MPA confirmations, the price differential between the 4 million barrels of Basrah and Bonny was a discount of $2.55 to Dated Brent.

[388] At any rate, the sale of the physical oil on 6 April 2018 would have covered the hedging loss of $83,680,000 while leaving CTSA with at least $164,322,400 (ie the forward price in the February 2016 transaction of $179,920,000 minus the ‘total haircut’ of $15,597,600). In this way, CTSA would have recouped all actual outgoings and been left with its profit from the transaction (the treasury fee and premium of $10,776,811, the hedging premium of $2,816,800 and the flat transaction fees of $1,314,579).

[389] It needs to be emphasised that recovery by CTSA of the hedging loss of $83,680,000 would not leave it in the same position as if the transactions had been fully honoured. Because CTSA was not allowed to uplift and sell the oil, it could not realise the additional amounts from which to fund its other outgoings or its profit. Although CTSA states that just and equitable relief should cover its other outgoings, it does not contend that the applicants should compensate it for the profit it would have made had the transactions been fully honoured. Those profits it calculates as totalling $13,714,965,[57] although those amounts do not exactly correlate with the fees and premiums contained in the sale confirmation.[58]

[390] The extent of CTSA’s hedging loss has been affected by the applicants’ delay. CTSA states that if its hedging position had been unwound on, say, 27 October 2016, the loss would have been only $42,270,476. If the applicants had firmly repudiated the transactions and launched proceedings within six to eight months after the board meeting of 5 February 2016, CTSA might well have terminated its hedge at around this time.

[391] In summary, CTSA’s out-of-pocket expenses total $231,751,074, of which $123,865,607 is covered by the tender of restitution, the balance being $107,885,474.


[392] If the transactions were set aside without more, Vitol would be out of pocket in respect of the purchase price of $78,606,000 and storage fees of R8,22 million. The applicants have tendered to restore these amounts with interest.

[393] Vitol incurred a cost of $37,530 to establish a letter of credit for payment of the purchase price. This was an upfront cost unaffected by the applicants’ delay.

[394] Vitol paid insurance premiums of $933,487 over the period January 2016 to May 2020.[59] The extent of this cost has clearly been affected by the delay.

[395] Once again, the biggest item of cost is a hedging loss. I have explained how Vitol concluded short-term hedges during the pricing window and long-term hedges to coincide with the expiry date of the SGA on 31 January 2019. With the litigation pending, Vitol extended its hedging, based on market conditions and its expectation of when the case might finally be determined. Vitol could not know whether the transactions would or would not be set aside.

[396] Vitol says that it was only fully apprised of the applicants’ case after service of the supplementary papers at the end of February 2020. Vitol also discovered, in late 2019, that SFF had pumped some of Vitol’s oil from Tank 2 into another tank. Finally, there was the fact that more than 700,000 barrels of Vitol’s Basrah was umpumpable (being Vitol’s aliquot share of the unpumpable 1,2 million barrels). And so Vitol concluded that its best course was to cancel the contracts and pursue an action for damages.

[397] In consequence of this election, Vitol unwound its hedges in May 2020, incurring a loss of $18,078,928. Because the unwinding happened after the oil price plunged in reaction to the Covid-19 pandemic, the hedging loss was relatively modest. There were times when Vitol’ exposure under the hedges had exceeded $100 million. Fortuitously, therefore, the applicants’ delay reduced rather than increased Vitol’s hedging loss.

[398] In summary, Vitol’s out-of-pocket expenses total $105,875,944, of which $86,826,000 is covered by the tender of restitution, the balance being $19,049,944.

Misconduct by respondents?

[399] The applicants submit that all the respondents have at least been guilty of reckless or utterly unreasonable conduct, and more serious allegations are levelled at Taleveras and Venus.


[400] I have found that there must have been an improper basis for Venus’ participation in the purchase of oil and for the unreasonably low price at which it bought the Bonny, even though the details have not been uncovered.

[401] Venus was aware that in a memorandum of 8 February 2016 Nkutha had questioned the low price for the Bonny. Gamede sent the memorandum to Venus and Taleveras on 10 February 2016. The next day Venus sent a letter of credit to Gamede as if nothing had changed. The inference is justified that from Gamede’s perspective , too, nothing had changed in substance, but he reminded Venus that he still needed to go through the motions, because he wrote to Van der Vent as follows on the afternoon of 11 February 2016 (I reproduce it exactly as written):

‘We have to respect the process that we have initiated to deal with this transactions. our enemies. I will appreciate it if we can hold this matter in abeyance as per my yesterdays letter. Continuing with other processes undermines this. The process that I have initiated seek to deal with the issues of accountability and transparency. Let us not do things that give an impression that I have instituted this process as a facade, as I am still continuing doing what the people have raised as concerns. I will appreciate it if you work with me on this.’

[402] Although this email is elliptical, it seems that Gamede regarded those who were raising queries as enemies (in another email to Van der Vent, also on 11 February, he spoke of the ‘vultures that want to devour on as’), and that he and Venus should be careful not to give them ammunition. That Gamede’s purported participation in the SFF’s internal processes of scrutiny was a facade is proved by the fact that privately he continued to progress the transactions and signed amendments without the internal processes having run their course. It was just a few days later, on 16 February, that Gamede sent Van der Vent the email which set out precisely how he had in mind to amend the pricing formula so as to give an appearance of reasonableness while still achieving a sale at a low price. Venus did not resist these endeavours.

[403] When Venus concluded its first SPA with Glencore on 8 January 2016, it knew that it would be making a profit of $18 million by immediately on-selling the oil to Glencore.[60] It also knew that it would be earning a storage mark-up of $420,000 per month over the five-year period of the SGA, ie $25,2 million in total.[61] Venus knew that in terms of the agreements as finally amended, it would still make a profit of $10,536,000 on the sale of the oil and that its storage profit was unchanged (though in the event, and because SFF later pulled out of the transactions, Venus’ storage profit was ‘only’ $9,741,704). Venus could not conceivably have thought that transactions on these terms with SFF were proper.

[404] In the circumstances, Venus must be regarded as a complicit party. This weighs heavily against it when assessing just and equitable relief.


[405] Glencore only received redacted versions of the Venus/SFF SPA and SGA. Glencore was thus unaware of the manifestly improper profits which Venus was making from the transactions.

[406] Glencore supplied Venus with a draft SPA and conditions precedent for the Venus/SFF SPA, which was not unreasonable bearing in mind that Glencore would be concluding a back-to-back SPA with Venus. These draft terms did not specify pricing. The applicants have not alleged that Glencore’s terms and conditions were improper. In the event, Glencore’s SPA was not used. Instead, SFF and Venus on 15 December 2015 executed what Glencore’s Mr Anthony Stimler described to his colleagues as an ‘SFF crappy SPA version’, onto which SFF and Venus tagged Glencore’s proposed conditions precedent.

[407] Glencore was unaware of the amendments made to the Venus/SFF SPA on 19 and 29 February 2016 (it saw them for the first time as annexures to the founding papers), and was not privy to the improper communications between Gamede and Van der Vent. Glencore’s initial and subsequent SPAs with Venus set prospective pricing windows in accordance with normal commercial practice. According to Glencore’s expert evidence, the price at which it bought the Bonny from Venus was market-related. There is no evidence that the shenanigans between Venus and SFF were designed to benefit Glencore.

[408] By 20 January 2016 Gamede, Mayaphi, Ndlela, De Wet and Beukes were in Geneva.[62] The applicants say that Gamede was staying at the Swissotel Metropole, a luxury hotel. This they gleaned from an email sent by Van der Vent to Gamede on 21 January. If SFF had paid for this trip and accommodation, Gamede would have needed to fill out various forms and get approvals, which was not done. The applicants thus inferred that the cost was paid by Venus or Glencore.

[409] In its answering papers, Glencore says that it took this allegation seriously and investigated it. It got details from its travel agent of all Glencore bookings made at the hotel during January and February 2016. These revealed that there were no bookings for Gamede or anyone else from SFF. On the papers, it cannot be found that Glencore funded the expenditure. Venus was certainly expecting to make enough money to render this expenditure a mere bagatelle, though there is no evidence that Venus paid for the trip. Representatives of Vitol, Taleveras and CTSA also met with Gamede in Geneva at this time, so the field for suspicion is wide open.

[410] In early June 2017 SFF notified Venus about the legal review and that the MoE had prohibited the upliftment of oil from Saldanha Bay. Despite the tripartite agreement and tank warrant issued to Glencore, SFF did not give Glencore a similar notification, and Glencore remained unaware of it. Glencore first learnt of the problem in late September 2017 after giving notice that it wanted to uplift its oil.

[411] The applicants have criticised Glencore for making no investigations about the propriety of the process by which the strategic stock was sold. Glencore, the applicants complain, did not take steps to satisfy itself that SFF had complied with its statutory and internal procurement processes. It also did not take the precaution of insisting on an SFF board resolution.

[412] It may be accepted that Glencore did not know the process which SFF had followed beyond being aware that Venus had evidently acquired a right to purchase 3 million barrels of Bonny. One must bear in mind, however, that the disposal of the oil did not involve procurement. While I have found that it was irrational for SFF not to have followed a competitive disposal process, a privately negotiated disposal by a public entity would have been less plainly irregular to an outsider than a privately negotiated acquisition. Vitol’s Mr Foster has alleged that over the period 2003-2008 he assisted PetroSA in five strategic stock rotations. There is no evidence that these were done by a tender process. The evidence is that while competitive disposal processes by state-owned oil companies are common, privately negotiated deals are not unusual.

[413] The applicants’ counsel submitted that ‘any reasonable businessman operating in South Africa would foresee that a project for a public entity must follow a procurement process’.[63] If the need for a competitive tender process in relation to disposals was so obvious, one wonders why SFF’s board so tamely approved the disposals on 5 February 2016 or why CEF, in its media statements and annual report later in the year, defended the transactions or why it took the applicants until March 2018 to launch review proceedings. According to one of Gamede’s memoranda submitted to the MoE in June 2016, SFF had always used private negotiation in its storage agreements.

[414] As to Glencore’s failure to insist on an SFF board resolution, it was dealing with the CEO, Gamede, and another senior executive, Mayaphi. The failure to call for a board resolution may be a minor point of criticism, but I do not think it shows that Glencore was turning a blind eye to rogue conduct. It was parting with a lot of money, which it would not have done if it suspected foul play. The evidential presumption of regularity, omnia praesumuntur rite esse acta, which courts continue to apply in appropriate circumstances, rests on the notion that, in the absence of warning signs, one is not bound to view the acts of public officials with constant suspicion.

[415] Moreover, if a board resolution had been called for, it may have been forthcoming. After all, on 5 February 2016 the board approved the Taleveras and Vitol transactions, and seems to have done the same for the Venus transaction provided the latter complied with its obligations.

[416] In support of their criticism of the failure of the respondents to call for a board resolution, the applicants’ counsel referred to City of Tshwane Metropolitan Municipality v RPM Bricks (Pty) Ltd 2008 (3) SA 1 (SCA), where Ponnan JA emphasised (para 11) the distinction between ‘an act beyond or in excess of the legal powers of a public authority’ on the one hand and ‘the irregular or informal exercise of power granted’ on the other. Estoppel can apply in the second category but not the first, provided the counterparty was acting in good faith without knowledge of the lack of authority. The applicants’ counsel also reminded me of the trite point that an estoppel regarding authority cannot be based on a representation by the unauthorised agent.

[417] I accept the foregoing points, but in my view they are not germane to the present enquiry. In RPM Bricks the counterparty was seeking to enforce the contract. The conduct of the municipality’s officials in varying a contract was held to fall into Ponnan JA’s first category, so that estoppel could not operate. The present case, insofar as board authorisation is concerned, appears to me to fall into Ponnan JA’s second category, because Gamede as CEO was the proper person to conclude contracts of the kind in question once an approving board resolution had been passed (see the passage from National and Overseas Distributors Corporation (Pty) Ltd v Potato Board 1958 (2) SA 473 (A) quoted by Ponnan JA in para 22; and see also s 20(7) of the Companies Act and SA Express Limited v Bagport (Pty) Ltd [2020] ZASCA 13; 2020 (5) SA 404 (SCA) paras 52-54).

[418] But more importantly, for present purposes the enquiry is not whether the respondents could successfully have raised an estoppel. The question is whether they were at fault in not insisting on sight of a board resolution. The fact that a particular requirement causes a decision to fall into Ponnan JA’s first category rather than his second does not necessarily tell one anything about whether the counterparty reasonably and in good faith thought that the requirement had been met.

[419] I thus consider that Glencore should be treated as an innocent third party. There is no material misconduct which weighs against it in the assessment of a just and equitable remedy.


[420] I have found that Taleveras paid bribes to Gamede. Like Venus, it must have known that it was buying its 2 million barrels of Bonny at below market price. Taleveras was given with a copy of Nkutha’s memorandum of 8 February 2016. On 22 February 2016 Taleveras and SFF executed the SPA amendment which sought to give a fig leaf of respectability to the unreasonably low price.

[421] Taleveras’ complicity in wrongdoing must thus weigh heavily against it in the scales when assessing a just and equitable remedy.


[422] CTSA emphasised in its opposing papers that its model of funding a trader’s transaction by a purchase and repurchase transaction means that little reliance need be placed on the creditworthiness of the trader and its guarantors; it is the ownership of the commodity that gives CTSA comfort. In mid-2015 Taleveras had approached CTSA in respect of a different transaction between itself and SFF. CTSA had declined the business because the security was unsatisfactory.

[423] In respect of SFF’s impugned transaction with Taleveras, CTSA was only approached on 29 December 2015, by which time Taleveras and SFF had already concluded their SPA and SGA. On this occasion the idea was to use the purchase and repurchase structure. CTSA was able to assess the matter with reference to contracts already concluded between SFF and Taleveras. However, because SFF, as the storage operator, would need to recognise CTSA’s title, the latter insisted on a face-to-face meeting with SFF and Taleveras.

[424] This meeting took place in Geneva on 22 January 2016. CTSA denies that it payed for the SFF’s trip to Geneva. Taleveras was represented by Mr Guillaume Coupez while present for SFF were Gamede and Mayaphi. CTSA’s representatives were Messrs Pierre Vanlerberghe and Nicolas Mignot. Gamede assured CTSA that it would recognise CTSA’s ownership of the oil, explaining that SFF often operated with international institutions and trading firms. CTSA asked why SFF was selling its strategic stock. Gamede answered that although SFF was obliged to keep 10,3 million barrels of oil as strategic stock, it operated a ‘quality swap programme’ in order regularly to rotate the stock ‘in order to ensure a good quality mix’. SFF had decided to buy 8 million barrels of new crude oil (4 million from Taleveras, 2 million from Glencore and 2 million from Vitol).

[425] Among the contracts CTSA had seen was the back-to-back SPA in terms whereof SFF purchased 4 million barrels of crude in Tank 1 from Taleveras, so the explanation made sense to CTSA. Vanlerberghe states that it is not unusual for state oil companies to rotate their stock, a view confirmed by CTSA’s expert, Ms Bossley.

[426] On 2 February 2016 Natixis’ credit committee approved a credit line of $165 million on the basis of a purchase/repurchase structure. By this stage SFF and Taleveras had already concluded the first amendment to their SPAs. The side letter agreement involving CTSA was signed on 4 February 2016, and on the same day SFF issued tank warrants to CTSA.

[427] CTSA states that until receipt of the review application, it was unaware of the amendment of 22 February 2016 to the Taleveras/SFF Bonny SPA (the one which implemented Gamede’s improper scheme to mask the discount at which Bonny was being sold).

[428] In May 2016 CTSA saw media reports questioning the legality of the stock disposal. This was the first inkling it had that something might be amiss. CTSA took this up with Taleveras which advised them that Gamede would be travelling to Geneva to provide clarification. The Geneva meeting took place on 10 June 2016. Vanlerberghe and Mignot for CTSA, and Coupez for Taleveras, met with Gamede, Mayaphi and a third SFF official whose name CTSA could not recall.

[429] Gamede told CTSA[64] that the media reports were ‘rumour and noise’ for political reasons connected to the upcoming elections. The media reports were a misrepresentation of the facts. He explained SFF’s evolving views of strategic stock. The sale of the oil to Taleveras had been done by the book. SFF had received authorisation from the MoE. Although National Treasury had to be notified, its consent was not needed, and notification had been done. Gamede assured CTSA that South Africa wanted to be an investment location. The Taleveras/CTSA transaction was important to CFF, and CTSA was ‘protected by local investment law’. Gamede suggested that CTSA write to SFF about its concerns, and SFF would send the letter to the MoE ‘to get high level answers and to be comforted’ as to its ownership. He would also facilitate a meeting with the ministry in Cape Town.

[430] CTSA read about Gamede’s resignation in early July 2016. CTSA seems not to have taken up the suggestion of writing to SFF or meeting with energy officials in Cape Town. However, Vanlerberghe and Mignot were in Saldanha Bay in January 2017, and took the opportunity of meeting with SFF representatives in Cape Town, as I mentioned before. They were given comfort as to CTSA’s title. They were not told that a legal review had just been completed and that the institution of review proceedings was in contemplation.

[431] It appears that SFF’s notification to Taleveras of 6 June 2017 regarding the legal review came to CTSA’s attention. They wrote to the new Acting CEO, Zulu, on 4 July 2017, pursuant to which a meeting with the latter took place on 25 July. Zulu was cagey and did not commit SFF to a definite position. Although Zulu’s letter of 12 September 2017 seemed to sound a note of optimism, this was quickly dashed by SFF’s letter of 26 September 2017.

[432] On 12 October 2017 CTSA wrote to SFF, stating that CTSA had bought the oil from Taleveras in good faith and at arm’s length, and that SFF was obliged to comply with its obligations under the tank warrants. CTSA sought details of the legal requirements supposedly not complied with, asked for a copy of the legal review, and asked when court proceedings would be instituted. In this latter regard, CTSA recorded that to the extent that the review relief impacted on CTSA, it insisted that the application be dealt with on an urgent basis. CTSA made itself available for a without-prejudice meeting.

[433] The only without-prejudice meetings involving CTSA took place in the first two weeks of November 2017. Several weeks later, on 29 November 2017, NRF (CTSA’s attorneys) wrote to CDH (SFF’s attorneys), stating that CTSA’ rights under the tank warrants and side letter remained in existence until set aside by a court, and that delay in instituting proceedings had already caused, and would continue to cause, prejudice to CTSA, particularly in relation to its hedging position.

[434] There having been no progress for more than another two months, on 5 March 2018 CTSA wrote directly to SFF, repeating that delay was causing it harm, and stating that if SFF did not comply with its obligations under the tank warrants and side letter, CTSA would have to meet its hedging obligations in early April 2018, causing it to suffer significant loss. In the then prevailing market conditions, CTSA expected its damages to be around $270 million.

[435] The review was launched a week later. With matters proceeding at snail’s pace, NRF wrote to CDH on 20 March 2019, rehearsing the chronology, complaining that there had been no attempt to expedite Gobodo’s forensic report, and stating that CTSA had been significantly prejudiced because of the applicants’ delay. Although CDH in subsequent correspondence complained of a ‘false narrative’ that the applicants were dragging their heels, I consider CTSA’s criticisms to have been justified.

[436] The applicants make the same criticisms of CTSA as of Glencore, namely a failure to investigate the propriety of the disposal process and the failure to insist on a board resolution. What I have said in respect of Glencore applies a fortiori to CTSA. It was dealing with SFF’s most senior officials, including its CEO. It insisted on a face-to-face meeting and received explanations which were plausible, even if they were false. CTSA, as a French financier, had no reason to assume that an open tender process was a legal requirement. As a fact, on 5 February 2016 SFF’s board approved the Taleveras transaction. When rumours surfaced in May 2016, CTSA again met with senior executives of SFF. At that stage the irregularities were evidently about procedural matters such as non-compliance with the PFMA, so CTSA can perhaps be forgiven for seeking assurances from a CEO who turned out to have been complicit in far worse. It is far-fetched to imagine that CTSA would have gone into a transaction of this kind suspecting that it was a house of cards which might collapse at any moment.

[437] One criticism of CTSA is that the extent of the profit which Taleveras stood to make must have been apparent to it (CTSA’s position in this respect is distinguishable from Glencore’s), and that this should have caused it to question the terms on which Taleveras was buying the oil, particularly the Bonny. Quite what CTSA could and should have done about this, however, is debatable. It was a financier, and Taleveras was its client. CTSA, unaware of the dishonesty which infected the Taleveras/SFF transactions, might have been exposed to a claim for damages by Taleveras if it had challenged SFF to explain why it was selling the oil to Taleveras at a price which allowed the latter to make an immediate profit of some $22,5 million.

[438] In my view, CTSA must be regarded as an innocent third party which was to some extent proactive in seeking to ensure that things were in order.


[439] Vitol’s principal representative in its dealings with SFF was Harvey Foster. He was previously with Masefield Energy Holdings AG (‘Masefield’), in which capacity he had dealings with PetroSA and SFF dating back to 2003. He became head of Vitol’s South African operations in 2010, and actively pursued its relationship with SFF. Despite insinuations of improper ‘wining and dining’ of SFF officials, there is nothing in the evidence, assessed in accordance with the Plascon-Evans rule, to show that Foster strayed beyond the legitimate promotion of his employer’s business.

[440] Vitol was aware by late 2014 that SFF was giving thought to optimising its strategic stock by freeing up storage space. This was pursuant to the First Directive of August 2014. Following correspondence between Vitol and SFF in February 2015, Gila identified Gamede as SFF’s contact person for purposes of the commercial relationship with Vitol.

[441] In an email of 11 August 2015, addressed to Gamede and copied to Cynthia Beukes and Susanna Pistorius of SFF, Foster made a proposal that Vitol lease 1 to 2 million barrels of Bonny in Tank 6, simultaneously pledging, as security, an equal quantity of Qua Iboe oil which Vitol was storing in Tank 5. The proposal was that for every barrel Vitol uplifted from Tank 6 (to be replenished within three months), it would pay SFF a fee of $0.05 (ie $50,000 per million barrels).

[442] Although this proposal did not come to pass, Gamede invited Foster to participate in a task team to advise the MoE on stock optimisation. Foster accepted the invitation in an email of 4 September 2015, listing the core aspects on which he thought the task team should concentrate. These included the involvement of the private sector in helping to achieve the aims of strategic stock.

[443] The applicants complain that Foster had a conflict of interest, since his advice might have been tailored to Vitol’s commercial interests. However, the task team did not come to much. Vitol, like other industry participants, was entitled to make proposals from which it might benefit and was entitled to promote them on the basis that they would also be beneficial to SFF. The latter would not have been so naive as to treat private-sector participants in a task team as wholly neutral. The applicants were surely not expecting Vitol to altruistically supply free expertise at the cost of being excluded from commercial transactions with SFF. (According to Gamede, Vitol was not the only global entity whose brains SFF tried to tap. He, Mayaphi, De Wet and Ndlela visited Mecuria, one of its biggest storage clients, to gain expertise in trading.[65])

[444] On 4 September 2015 Foster sent Gamede a proposal intended for the MoE, asking whether it was acceptable. This was a modification of the proposal of 11 August 2015. The leasing proposal now related to parcels of 1 million barrels of Bonny or Basrah, with the leased barrels to be replenished within 12 months. Vitol would pay SFF a fixed fee to be negotiated for every 1 million barrels leased. As before, an equivalent quantity of stored oil would be pledged.

[445] On 15 September 2015 Foster’s colleague, Marc Ducrest, wrote to Gamede summarising the leasing proposal, including a fee of $50,000 per 1 million barrels leased. He emphasised that this was a once-off fee, not a monthly fee. The applicants inferred that the ‘fee’ of $50,000 was a euphemism for a bribe to Gamede. The applicants made this allegation without having had sight of Foster’s email of 11 August 2015, which was dealt with in Vitol’s answering affidavit. In reply, the applicants denied that Ducrest’s ‘fee’ could be explained with reference to Foster’s email of 11 August 2015, because Ducrest was dealing not with the proposal of 11 August 2015 but with the proposal of 4 September 2015.

[446] Ducrest’s email did not explicitly refer back to Foster’s proposal of 4 September 2015, but it does not matter. The simple point is that Vitol’s leasing proposal involved a once-off fee per 1 million barrels leased, and the fee Vitol had in mind was $50,000. Vitol’s version that this was not a disguised bribe must be accepted on the papers and is consistent with the documents. If the transaction had come to fruition, the fee would have been paid to SFF, not to Gamede.

[447] Earlier in this judgment I mentioned the two letters Vitol wrote to the MoE on 15 September 2015 and her reply of 10 October 2015.

[448] Vitol was one of the five companies to which Gamede issued an RFP on 13 October 2015. In view of the prior commercial relationship between the parties, it is unremarkable that Vitol was selected to receive an RFP. The applicants portray Vitol as the party which effectively planted and nurtured in Gamede’s mind the idea of disposing of the strategic stock. The papers do not justify this picture.

[449] In response to the RFP, Vitol requested more information and permission to draw samples from Tank 2. In the event, says Foster, an assay of the Basrah was not possible because there was a 4-metre layer of sludge.

[450] Vitol’s first proposal was dated 16 October 2015. Vitol sought the option to rotate a maximum of 5 million barrels of oil in Tanks 2 and 6, with replenishment to take place within 36 months, subject to earlier emergency for strategic reasons. Vitol states that this proposal was similar to transactions Vitol had undertaken with PetroSA.

[451] Following a meeting in Milnerton on 26 October 2015, Vitol submitted an expanded proposal on 29 October 2015. The stock rotation proposal remained unchanged. The other aspects of the proposal are not relevant. I do not consider there to have been anything improper about the submission of further proposals. The RFP was not subject to a time-limit.

[452] Vitol messaged Gamede from time to time during October and November 2015, indicating a desire to progress the matter. No improper pressure was placed on SFF. In the event, if SFF had moved more quickly it would have avoided the negative impact of the price drop in January 2016.

[453] On 17 November 2015 Ducrest emailed Gamede, recounting market chatter which Vitol had picked up about SFF’s dealings with other oil companies, inter alia in connection with rotation. For example, he had heard that SFF gave Mecuria a letter of support in connection with a recent NNPC[66] tender, adding: ‘We requested the same but we did not receive SFF backing. Did we do something wrong?’ He had heard that SFF was discussing rotation with ‘Mecuria, Chinese and others’, and a claim was doing the rounds that Skydeck had a mandate to buy strategic stock. His concluding point was that ‘we need to move as soon as possible to close our proposals as some vultures are turning around’. I do not regard this email as improper. Vitol was entitled to promote its own commercial interests. There was nothing dishonest in urging Gamede to move quickly.

[454] In his reply, Gamede told Ducrest that SFF had sent out RFPs to a number of people, but that Vitol ‘and the Chinese’ (Mecuria?) were the only companies that responded. (Gamede’s answer seems not to have been truthful, because Taleveras had also submitted a bid.) Gamede wrote that SFF was setting up a trading department, and would be seeking Vitol’s assistance. Until the trading department was in place, there would be no rotation. The fact that SFF had signed cooperation agreements with other people should not be negatively viewed by Vitol. SFF needed strong partners like Vitol, but Gamede did not understand Foster’s proposal, and the proposal was not comprehensive.

[455] Vitol’s bid was more comprehensive than Taleveras’ one. Gamede received his first bribe payment from a Taleveras source one week after his exchange with Ducrest. One cannot help thinking that Gamede might have wanted Ducrest to read between the lines that the logjam might be broken by a gratification. If so, there is no evidence that Vitol ever offered any.

[456] I previously mentioned the expanded RFP which Gamede sent Ducrest on 20 November 2015. In reply, Ducrest wrote:

‘If you feel you need to justify your actions by detailing such to the market, then I suppose you need to do it … However, the more detail and reasoning you provide to the market the more this can be criticized.’

[457] He carried on by saying that it made sense for SFF to conclude sale/rotation transactions per grade with only one company, because SFF would not want lots of different parties buying crude oil in the same tank. Furthermore, SFF should avoid flooding the market with Bonny or Basrah, as this would erode its value: ‘Strongly believe that a single large placement of grade to Vitol/Vesquin works the best.’ He encouraged Gamede to emphasise ‘to the top’ (presumably the board and/or MoE) the importance of the counterparty’s ability to pledge an equivalent quantity of crude.

[458] Ducrest’s reply may justify the inference that he realised that Gamede’s justification for selling the strategic stock was dubious. On the other hand, he may have meant no more than that an elaborate explanation, even though sound, would merely give an aggrieved party material for criticism.

[459] The rest of Ducrest’s email was aimed at persuading Gamede that SFF’s best interests would be best served by selling all the Bonny or all the Basrah (but not both, as I understand it) to a single buyer, preferably Vitol. It was natural that Ducrest should promote his company’s interests. The applicants’ trading expert, Mr Driscoll, said that it was a general phenomenon that traders preferred private negotiation to tenders and auctions. After quoting Ducrest’s email, he said that it was not his intention to impugn Ducrest’s motives: ‘The exchange highlights the predilection of traders to steer SFF away from competitive tenders and seek optimum advantage through bilateral negotiation.’[67] The idea that SFF should favour a party which could pledge an equivalent quantity of oil (this would be on the basis of uplifting the strategic stock) was not sinister. On the contrary, it was consistent with meeting SFF’s need for security of supply.

[460] Vitol says that it submitted a further proposal to SFF on 1 December 2015, acknowledged in Gamede’s letter of 3 December 2015 in which the latter asked Vitol to afford SFF the opportunity to engage with its internal processes, including board approval. Neither the applicants nor Vitol have produced this document.

[461] Not long afterwards, Gamede phoned Foster to say that Vitol had been awarded 3 million barrels of Basrah. (As previously mentioned, Vitol says that it did not receive the award letter of 8 December 2015 until 13 January 2016.) Foster was disappointed, because Vitol had been hoping to get 5 million barrels of Bonny. (That Taleveras and Venus received all the Bonny would be consistent with Gamede having had some improper reason to favour them.)

[462] Despite its disappointment, Vitol wanted to close the deal promptly. Foster engaged with Chili and Gamede on terms, but there was no urgency from Gamede. In her email of 23 December 2015, Chili said that the agreement was fine from a legal point of view but that she would leave the commercial terms ‘to the Trading Department because this is their domain’, creating the impression that the trading department already existed. The agreement vetted by Chili included a provision, also to be found in the final agreement, that the parties had agreed to work together closely to assist SFF with a skill’s transfer programme for SFF’s trading department.

[463] Having in the new year received Gamede’s letter dated 8 December 2015, Foster renewed his engagement with him on 13 January 2016, proposing an ITT in the period 16-31 January 2016 at Vitol’s nomination, the pricing window to be five days after ITT date. A discount to Dated Brent would have to be negotiated but he proposed $10.

[464] The $10 discount not find favour with Gamede or Mayaphi, and they asked Beukes to look into prevailing prices for Basrah. She could not get spot prices. For March 2016 futures, the Basrah discount was $1 and the Bonny premium $2.50. The parties met on 18 January 2016 at which they agreed on a discount of $8 and a storage fee of $0.11.

[465] I have already dealt with the conclusion of the contracts on 20 January 2016 and the subsequent haggling over the pricing window and discount. Ultimately the parties compromised on a pricing window of 25-29 January and a quality discount of $5.50.

[466] The applicants criticise Gamede for disposing of the strategic stock at a time when oil prices were at an 11-year low. There is merit in that criticism, and I have dealt with it elsewhere. However, and on the assumption that the sale of the Basrah was not in principle objectionable, the evidence does not show that SFF sold it to Vitol at less than market value. Foster, although he is not independent, has expertise to opine on the matter. He states that Basrah is a relatively heavy high-sulphur crude. Normally it trades at a significant discount to Dated Brent. The oil in Tank 2 was even less desirable, because it had become mingled with various grades and had lost its identity as a true Basrah. It was not located close to a refining centre.

[467] He states that Basrah (assumed quality 34 API) is supplied by Iraq’s state oil company, SOMO. For Europe-destined Basrah, SOMO’s January 2016 price was Dated Brent less $4.75 FOB at Al Basrah. SOMO would give a further discount of $0.40 per API unit <34. The API of the oil in Tank 2, as advised by SFF in January 2016 (presumably based on the Intertek certificates of 17 January 2016), was 29.66 API, pointing to further discount of $1.74,[68] or an adjusted total discount of $6.49. If the oil in Tank 2 were destined for the Mediterranean (a major refining centre), the discount would be even greater in comparison with oil transported from Iraq.

[468] In reply, the main criticism by the applicants and their experts is that without an assay of the oil, Gamede could not know whether the price was market related. The applicants’ trading expert, Mr John Driscoll, did not frame his report as a response to Foster’s assertions. He did not deal with the reported API of 29.66. His report does not expressly or by necessary implication challenge Foster’s statements about SOMO’s January 2016 pricing for Basrah or the additional discount for oil <34 API.

[469] Accordingly, while one can probably say that the original SPA with its discount of $8 was skewed in Vitol’s favour, the ultimate discount of $5.75 was unobjectionable, as was the selected window period. Vitol’s belief was that South African refineries would not be ready buyers of Basrah and that Vitol thus needed to factor in the cost of transporting the oil to Europe.

[470] The applicants criticise Vitol on the basis although the RFP of 13 October 2015 attached the Second Directive, Vitol took no steps to check that the MoE’s preconditions had been satisfied. The criticism does not have much force. The conditions were internal matters. Gamede notified Vitol in his letter of 8 December 2015 that the MoE had approved the disposal. Regardless of whether or not Vitol saw the Second Approval Notice, Gamede’s statement was true. Vitol could not be expected to have known what due diligence SFF had undertaken or what motivation it had put up to the MoE. It was not under a duty to demand information in that regard.

[471] On 17 December Foster met with Chili[69] to discuss the anticipated terms of the SPA and SGA. Her failure to alert Foster to possible irregularities would have given Vitol comfort that it was not dealing with a rogue CEO. The terms which Vitol was proposing, and which Chili reviewed, included a provision by which SFF warranted that it was duly authorised and entitled under applicable law to carry out and fulfil the obligations imposed by the SPA.[70] This provision was included in the final SPA,[71] as was an additional anti-corruption clause by which SFF undertook to comply with all applicable law relating to bribery and money laundering (Vitol gave a like undertaking).[72]

[472] Mayaphi and Chili were with Gamede in SFF’s negotiation with Vitol on 18 January 2016, again lending respectability to the process. On 20 January 2016, by which time Gamede and Mayaphi were in Geneva, Chili sent them and Foster her tracked suggestions for the SPA, adding that she had ‘no major legal issues with the contract’. The SPA and SGA were signed that same day, with Gamede, Mayaphi and Ndlela in attendance.

[473] At some stage thereafter Mayaphi and Ngqongwa discussed with Foster the possible hedging of SFF’s position under the SPA. There are some inconsistencies in the affidavits as to when this happened. Ngqongwa’s affidavit, filed as part of the replying papers, states that Ngqongwa and Mayaphi spoke with Foster about this in London in mid-February during the course of the London Oil Week,[73] which seems to accord with Foster’s supplementary answering affidavit.[74] Ngqongwa’s version is that in addition to discussing the potential for hedging, he raised his concerns about the process followed by Gamede. Foster allegedly acknowledged that SFF’s processes had not been followed but that this was unimportant because Gamede had signed the contracts. In response to Ngqongwa’s assertion that Gamede had done so without following due process, Foster retorted that even if that were so, Vitol’s processes had been observed and Vitol had a binding agreement.

[474] Although Ngqongwa does not himself say so, the applicants’ principal replying deponent, Mr Tseliso Maqubela, alleges that Ngqongwa was absolutely certain that he flagged his concerns about the transactions and told Foster that there was no way SFF could have signed away the strategic stock without due process and proper approvals. Ngqongwa told him that the contracts had not been approved by the board, to which Foster replied that this was an internal SFF process and ‘not his problem’. (It would have been factually incorrect for Ngqongwa to claim that SFF’s board had not approved the Vitol transaction. Such approval was given on 5 February 2016.)

[475] The above evidence should have been contained in the applicants’ founding papers. Vitol delivered supplementary answering papers in which Foster dealt with Ngqongwa and Maqubela’s versions. He confirmed having met Mayaphi and Ngqongwa in London in February 2016. They discussed hedging and a request from SFF that Vitol waive its right to be offered further storage space. According to Foster, these were the only two matters discussed, and he denies having ever uttered the ‘not my problem’ remark. In regard to hedging, Foster considered that Vitol, as a counterparty, was conflicted in advising SFF on the subject, but he agreed to collect some reports on anticipated oil price fluctuations so that SFF could take an informed decision. He did so, and sent the material to them on 26 and 29 February 2016.

[476] Foster’s version cannot, in accordance with the Plascon-Evans rule, be rejected on the papers. It follows, on Vitol’s version, that by the end of February 2016 it had in essence dealt with SFF’s entire senior management team without any hint from SFF’s side of irregularities.

[477] In late May 2016 the Financial Mail approached Vitol with various questions about the strategic struck disposals. Vitol responded, and furnished a copy of its responses to Gamede and Mayaphi. As previously mentioned, SFF on 5 July 2016 notified Vitol of Gamede’s resignation but assured Vitol that its crude oil stored at Saldanha Bay was properly safeguarded. Vitol’s requests to uplift its oil in February-May 2017 were met with obfuscation, until SFF’s letter of 9 June 2017 revealed that there had been a legal review as a result of which the MoE had instructed that strategic fuel could not for the time being be uplifted. Judicial review was foreshadowed in SFF’s next letter of 26 September 2017.

[478] In the circumstances, Vitol is entitled to be treated as an innocent third party. There is no misconduct on its part which would justify the court from refraining to grant a remedy which would otherwise be just and equitable.

The applicants’ misconduct

[479] Most of the important features of the applicants have been identified earlier in this judgment. I can thus simply summarise the main elements bearing on just and equitable remedy.

[480] Although Gamede was driving the improper disposal process and to a large extent made decisions on his own, he could not have achieved what he did without the acquiescence or supiness of SFF’s senior managers and directors. It is clear that Mayaphi has not been frank with the applicants about his involvement. He signed the Taleveras SPAs and the BPA as a witness. Ndlela was present with Gamede and Mayaphi in Geneva when Vitol’s contracts were signed (the applicants have not filed an affidavit from him). Chili vetted disposal contracts without raising red flags (she too has not made an affidavit). Although Ngqongwa and Nkutha might have been sceptical, they were not forceful in raising objections. The board was allowed to go in ignorance of what was happening until 5 February 2016. Afterwards, Ngqongwa aided the perpetuation of the transactions by preparing a request for condonation to the MoF.

[481] De Wet, who was appointed General Manager: Commercial, in May 2015, states that her responsibilities included ‘commercialising’ SFF’s assets, particularly storage facilities. It was her task to negotiate storage agreements, yet she was ‘severely sidelined’ by Gamede and Nkutha in late 2015. She accompanied Gamede, Mayaphi, Ndlela and Beukes to Geneva in January 2016. She does not seem to have done anything about her side-lining before she was redeployed, as from February 2016, as General Manager: Corporate Services.

[482] Instead of reading the riot act to Gamede and his team on 5 February 2012, the directors lamely approved the Taleveras and Vitol transactions and effectively did the same for the Venus transactions even though they must have known that they did not have enough information and that such information as they did have had been sprung on them at the last moment. They knew what CEF’s procurement policy required, yet they were supine in the face of a patently uncompetitive process. The board did not proactively intervene to ensure that SFF and the country’s interests were safeguarded.

[483] The applicants have not disclosed when CEF’s board learnt of the matter and how it reacted to the news. No CEF board minutes have been included in the papers. What one does know is that when the transactions began to attract adverse media attention in May 2016, CEF’s initial reaction was to jump to SFF’s defence. The impetus for a legal review came from National Treasury, the MoF and (perhaps belatedly) the MoE.

[484] In his replying affidavit Maqubela alleged (and Ngqongwa confirmed) that Gamede wielded ‘extensive political power’. At meetings he would switch recordings off when mentioning ‘political pressure to execute certain transactions and decisions’. Ngqongwa went so far as to say that he was apprehensive for his physical safety. Gamede, so Ngqongwa claimed, was also able to manipulate the MoE, and was dishonest. If this is true, the unfortunate position is that SFF’s managers and board were in Gamede’s thrall and failed to do their duty.

[485] Accordingly, and quite apart from the gross and unsatisfactorily explained delay, the rot which allowed the impugned transactions to be concluded and implemented was pervasive, even if one man was the linchpin.

[486] Vitol contends that the applicants are not pursuing the review in the public interest, and that a driving factor is a desire to avoid the contractual damages which the respondents could otherwise claim. That would be a serious criticism, but I do not think it is justified. The applicants firmly adopted their stance of taking the impugned conduct on review in September 2017. At that stage SFF was not yet in breach. SFF could have complied with its obligations without facing a claim for contractual damages.

[487] It may nevertheless be true that the applicants’ motivations are ultimately financial. If, in September 2017, the oil price had been at or below the levels at which SFF sold the oil, the applicants may well have concluded that there was little point in pursuing a review and that they might just as well replenish the strategic stock out of the funds received from the respondents. Cost-effective replenishment only became possible when the oil price tumbled in late March 2020, but by then the applicants were too far down the review path to retract, and after another two months the price recovered sufficiently to make this option no longer attractive. This is a consideration to bear in mind. We do not know that the applicants were not delaying their review to see whether market conditions might provide a commercial solution to their problem.

Just and equitable relief in this case – a broad assessment

[488] Because I find Vitol and CTSA to be innocent parties, and because of the applicants’ egregious delay and misconduct, my broad view is that either Vitol and CTSA’s contracts should be allowed to stand (so that they can pursue contractual remedies, at least to recover out-of-pocket losses) or that those contracts should be set aside subject to payment of compensation for the out-of-pocket costs. (The same would have applied to Glencore had it not agreed to more limited relief.)

[489] On balance, I have concluded that the remedy that will vindicate effectively the rights violated by the impugned decisions and transactions while being fair to those affected by those decisions and transactions (cf Steenkamp NO v Provincial Tender Board of the Eastern Cape [2006] ZACC 16; 2007 (3) SA 121 (CC) para 27) is an order setting aside the decisions and transactions against the payment of compensation, such compensation limited to out-of-pocket expenses, ie excluding profit. This will give Vitol and CTSA less than their contractual rights, but will ensure that they have not suffered fruitless expenditure because of the unlawful conduct of SFF and the MoE.

[490] Apart from restitution, most of the out-of-pocket expenses – insurance premiums, inspection costs and hedging losses – were incurred by virtue of the risk that Vitol and CTSA carried for some years as the owner of the oil. (CTSA is still the owner; Vitol was the owner until it cancelled the contracts in June 2020.) It was commercially prudent, indeed necessary, to insure the oil. The evidence shows that SFF would itself have done so had it remained with SFF. As owner, Vitol and CTSA prudently arranged for routine inspections. SFF, if it had remained owner, would have done no less. It was also commercially prudent for Vitol and CTSA to hedge the risk of adverse movements in the value of the oil. SFF has quite rightly not alleged that any of the respondents acted imprudently by hedging. It is manifestly unjust for the applicants to ask the court set aside all the transactions, and retrospectively to nullify Vitol and CTSA’s ownership, while insisting that Vitol and CTSA carry all the costs necessarily or prudently associated with their four years of ownership.

[491] I would not go so far as to compensate Vitol and CTSA for lost profit. Where an innocent party has already earned profit from a contract later declared invalid on public law grounds, a court might be disinclined to order such party to surrender the profit. Here, Vitol and CTSA are not already possessed of a profit. If their contracts stand, they would still need to sue SFF for it. The profit would be a benefit of ownership, not a cost.

[492] I regard the circumstances of this case as exceptional. There were pervasive irregularities in the disposal of the country’s strategic oil reserves. Something beyond a mere declaration of invalidity seems to be called for, despite the applicants’ egregious delay. A setting aside of decisions and transactions, which results in the oil continuing to vest in SFF, achieves this, but only at great cost to innocent parties. The applicants, I must emphasise, have not alleged that they will be unable to pay compensation if it is ordered. Their case is that the circumstances do not justify compensation.

[493] On the other hand, because of misconduct by Taleveras and Venus, I would find that their contracts should be set aside without compensation. That question is, however, academic, because Venus has not participated in the proceedings and Taleveras is content with restitution.

[494] OUTA’s counsel submitted that a compensation order would not be appropriate unless it promoted an efficient and effective public administration grounded in the rule of law. I accept that this is an important consideration, but in my view it does not militate against my broad assessment. I emphasise, again, the distinction between compensation in this case (as an amelioration of the harsh effects of setting-aside orders) and compensation or constitutional damages for the violation of fundamental rights. It is the applicants who seek a public law remedy to vindicate the rule of law, and that remedy is a setting aside of contracts coupled with limited restitution. The question is whether mere restitution would be just and equitable.

[495] It was SFF, not Vitol and CTSA, which acted in violation of the principle of legality in making the impugned decisions and concluding the impugned transactions. It was the applicants who violated the principle of legality by grossly delaying the institution of proceedings (cf Khumalo & another v Member of the Executive Council for Education: KwaZulu-Natal [2013] ZASCA 49; 2014 (5) SA 579 (CC) paras 45-47; Department Of Transport & others v Tasima (Pty) Ltd [2016] ZACC 39; 2017 (2) SA 622 (CC) para 160). If, as I find, Vitol and CTSA are innocent parties, I fail to see how requiring them to suffer the loss of their out-of-pocket expenses, while allowing SFF to keep the oil and do no more than return the money, would promote an efficient and effective public administration grounded in the rule of law. Accountability is one of South Africa’s foundational values (s 1(d) of the Constitution). In terms of ss 195(1)(a) and (f) of the Constitution, the principles governing public administration include the promotion and maintenance of a high standard of professional ethics and that public administration must be accountable.

[496] On the contrary, such an order would send out a message to officialdom that no matter how poorly they administer a State entity’s affairs, the court will see to it that the entity suffers no loss. There would then be little incentive to avoid loss by efficient and honest administration. In the absence of loss, responsible officials might not be held accountable by the State. Such an order would also convey to private parties that they should be hesitant to contract with State entities, because if things go wrong it is they, rather than the State, that will be made to suffer the loss.

[497] I asked OUTA’s counsel in what other way the court might send out the right message. She said that my judgment could embody a ‘stern reprimand’ or recommend an ‘investigation into the officials’ conduct’. In my view, these steps would be hollow and impotent. To fulminate is idle. Improvement will only happen when public bodies are made to bear the adverse consequences of their officials’ conduct.

[498] Compensation for out-of-pocket expenses (but not for lost profit) accords with the no-profit-no-loss principle which the Constitutional Court articulated in Allpay (2) para 67. In a very real sense, such compensation results in complete restitution – the position in which the relevant respondents would have been had they never contracted with SFF.

[499] In a post-hearing note I asked counsel for the applicants, Vitol and CTSA to address certain questions underpinned by the above broad assessment. The primary question was whether, before setting aside the contracts and awarding compensation for out-of-pocket expenses, I needed to be satisfied that Vitol and CTSA’s contractual claims for damages were at least as ample as the proposed compensation. (For convenience I refer to this as the ‘contractual qualification’.)

[500] In response, Vitol and CTSA submitted that compensation as a public law remedy should not be viewed as a substitute for contractual damages and could thus notionally exceed such damages. They argued that in any event their contractual damages would far exceed compensation for out-of-pocket expenses. The applicants’ reply was prefaced by submissions as to why compensation should not be granted at all. However, on the assumption that the court was otherwise satisfied that compensation should be awarded, the applicants agreed with the contractual qualification.

[501] Where compensation is sought as a remedy for the violation of the claimant’s constitutional rights, there might be no reason to link the quantum of such compensation to the value of claims the party might have in contract or delict. However, and as I have already said, compensation in the present case would not serve that function. Rather, compensation would ameliorate the consequences of setting aside the relevant decisions and contracts. Since this would be its function, there must necessarily be, as a matter of justice and equity, a link between the compensation and the value of the contractual rights which the respondents stand to lose. If the contractual rights were or might be less valuable than the out-of-pocket compensation, the public interest might be better served by allowing the contracts to stand.

[502] I thus consider that the contractual qualification should be observed when assessing Vitol and CTSA’s claim to compensation. This does not mean that a contractual remedy is granted under the guise of constitutional relief. It acts rather as a check upon constitutional compensation, ie as a circumstance which might cause one to conclude that the grant of constitutional compensation would not be just and equitable.

Just and equitable remedy – Vitol

[503] Vitol’s out-of-pocket expenses total $105,875,944, comprising the purchase price and storage fees totalling $86,826,000 (the restitution offered by the applicants) and other outgoings amounting to $19,049,944. It is this latter amount which represents the contentious compensation.

[504] Vitol stated in its answering papers that it had no option but to ‘cancel the sale agreement’. It said that it would pursue a claim for damages if the Vitol contracts were upheld. In Vitol’s supplementary answering papers, Foster stated that SFF had ‘repudiated and/or breached its contracts with us and we have cancelled them’, entitling Vitol at least to restoration of the purchase price, storage fees and interest as well as damages for other losses suffered. He made clear that Vitol no longer claimed ownership of the oil.

[505] In their replying papers the applicants attached a notice of cancellation from Vitol dated 8 June 2020.[75] In this notice Vitol stated that SFF had been placed in breach of the SPA by the letter of 6 October 2017; that SFF’s ongoing refusal to recognise Vitol as owner, coupled with the pumping of the Basrah out of Tank 2, was a repudiation of the SPA; and that on the basis thereof, Vitol elected to cancel the SPA.

[506] Although the main answering papers referred to a cancellation of the SPA, one would expect the SGA to fall with it, and this is what Vitol stated in its supplementary answering papers. In their replying papers, the applicants did not contend that SFF had not repudiated or breached the Vitol contracts; clearly SFF did. It follows that unless the contracts are set aside, Vitol’s cancellation stands and it would be entitled to claim damages.

[507] Turning to the contractual qualification, the first question is which of Vitol’s contracts SFF breached. In my view, it was the SGA, a proposition with which Vitol agreed in response to my note. The SPA was fully performed once Vitol paid the purchase price and it acquired ownership by an ITT. If the contracts are not set aside, Vitol became and remained the owner. Vitol’s contractual prejudice arose because SFF refused, as the storage operator, to allow Vitol to uplift the oil which it owned.

[508] What then would be the proper measure of Vitol’s damages? In my post-hearing note I drew attention to the distinction between positive inter esse (‘expectation damages’) and negative inter esse (‘reliance damages’), as discussed in Mainline Carriers (Pty) Ltd v Jaad Investments CC & another 1998 (2) SA 468 (C). Expectation damages would seek to place Vitol in the position it would have been in had SFF performed. Reliance damages would seek to place Vitol in the position it would have been in had it not contracted with SFF. The former would be based on what Vitol could have realised for the oil if it had been duly delivered.

[509] I posed this question because an award of out-of-pocket expenses would be akin to reliance damages, and because there is legal uncertainty as to whether reliance damages can be claimed for breach of contract.[76] Apart from this legal controversy, there are clauses in the contracts which might restrict the recovery of reliance damages. Clause 16 of the Vitol SPA precludes a claim for ‘indirect or consequential losses or expenses’ or for ‘any loss of anticipated profits’. Clause 25.4 of the Vitol SGA stipulates that neither party shall be liable to the other ‘for any indirect, special, incidental or consequential damages’ arising out of or in connection with the agreement ‘unless such damages arose as a result of gross negligence’ by the other party.

[510] As to Vitol’s expectation damages, these would depend on the value of the oil at the relevant contractual date. In my post-hearing note I raised the question of the relevant contractual date. If it fell in the period 2017-2019, the precise date is unimportant, because throughout that time the price of Dated Brent was sufficiently above the disposal prices to ensure an excess more than sufficient to cover the out-of-pocket expenses. The position might be different if the relevant date were more recent, given the effect of the Covid-19 pandemic on the oil price. I drew counsel’s attention to the majority judgment in Culverwell & another v Brown 1990 (1) SA 7 (A), since it occurred to me that, at least in Vitol’s case, the relevant date might be in May/June 2020, when Vitol finally accepted SFF’s ongoing repudiation and cancelled the contracts.[77]

[511] Upon reflection, I am satisfied that Culverwell would not apply in the present case. Culverwell deals with the computation of damages for cancellation following an anticipatory breach. The usual rule, as appears from Culverwell, is that damages for breach of contract are assessed at the date of the breach. Once the notion of anticipatory breach was accepted in our law, a different approach was needed for such cases, since ex hypothesi the time for performance has not yet arrived. In the present case, however, Vitol’s primary damages flow not from a cancellation following an anticipatory breach but from SFF’s earlier failure to perform on due date. The conventional rule thus applies to the primary damages.

[512] On 19 April 2017 Vitol gave notice of its intention to withdraw its 3 million barrels from Tank 2, and nominated export slots in June-August 2017. In its letter of 9 June 2017 SFF refused to allow this on the basis that the applicants were considering their position in the light of legal advice. On 26 September 2017 SFF formally notified Vitol that it regarded the transactions as invalid and was intending to apply for judicial review. Vitol rejected this, and on 6 October 2017 formally gave notice that it intended withdrawing its oil during November 2017, nominating three loading slots in the first half of November. Vitol stated that any continued refusal to allow the withdrawal was a breach of contract. In its response of 20 October 2017, SSF adhered to its stance and refused to allow the removal of the oil.

[513] In view of SFF’s explicit repudiation, Vitol was not required to give a notice to remedy in terms of clause 16.1 of the SGA (Taggart v Green 1991 (1) SA 121 (W) at 125F-126I; Vorster NO v PM Security and Crime Prevention (Pty) Ltd T/A Hermanus [2015] ZAWCHC 64 para 20). If the contracts stand, Vitol would be entitled to be placed in the position it would have been in had the oil been released in the first half of November 2017. At that time Dated Brent was in the range $60 – $63. If one assumes the same discount of $5.50 that SFF and Vitol agreed in their SPA, the oil would have been worth at least $163,5 million.[78] The precise levels for Dated Brent and the discount are unimportant, because at any plausible levels the value far exceeds the out-of-pocket expenses.

[514] When Vitol eventually cancelled the transactions on 8 June 2020 Dated Brent was closer to $40. Again assuming a discount of $5.50, the oil at that time would have been worth $103,5 million, slightly below Vitol’s out-of-pocket expenses. However, the difference between that amount and the value the oil would have had at the time SFF should have released it in November 2017 would be part of SFF’s damages.

[515] The aforesaid damages, based on the value of the oil, are damages flowing naturally and generally from the breach of the SGA, namely that Vitol was deprived of the value of its oil. In a claim to recover these damages, Vitol would not need to allege and prove its out-of-pocket expenses. If the value of the oil covered the purchase price, storage fees and other out-of-pocket expenses and left a surplus, Vitol would have made a profit; if not, it would have made a loss. Either way, Vitol would not need to prove its outgoings, from which it follows that the limitation clauses in the contracts would not be applicable.

[516] Accordingly, the quantification of Vitol’s damages is such that the contractual qualification is satisfied. The remaining issue on the contractual qualification is whether a contractual claim would succeed on the merits. If the contracts were not set aside, SFF might seek to avoid liability by pleading a lack of authority on Gamede’s part. On that question I take a robust view. SFF has not sought to evade the contracts on private law grounds. But for the advice to launch judicial review proceedings, SFF would, I am sure, have complied with the contracts – it would not have risked the damages flowing from repudiation.

[517] Furthermore, Vitol’s contracts were ratified by the board on 5 February 2016. Unless the board’s decision were set aside on review, the ratification would stand. In addition, the applicants defended the disposals a few months later when they attracted adverse media coverage. SFF billed Vitol for storage fees for several years. In those circumstances, it is not plausible that SFF could avoid liability on the contracts in private law proceedings.

Just and equitable remedy – Taleveras and CTSA

[518] CTSA’s out-of-pocket expenses total $231,751,074, comprising the purchase price and storage fees received by SFF totalling $123,865,600 (the restitution offered by the applicants) and other outgoings amounting to $107,885,474. It is this latter amount which represents the contentious compensation. The latter amount includes the one month’s storage fees of $480,000 which CTSA paid to Taleveras but which did not find its way to SFF.

[519] CTSA’s position differs from Vitol’s because of Taleveras’ role as the party which bought the oil from SFF. This requires me to consider certain complications absent from Vitol’s case.

Taleveras’ assertion of a settlement

[520] Taleveras claims that it accepted an offer of restitution made by the applicants in their founding and supplementary founding affidavits. In argument reference was also made to statements contained in the replying papers.

[521] In the founding and supplementary founding papers, the passages on which Taleveras relies are simply statements as to what the applicants considered appropriate if the court were to set aside the impugned transactions.

[522] The replying papers are more helpful to Taleveras. In para 327 the applicants’ deponent said that SFF ‘proposes to pay the respective traders’ a specified amount in restitution. In para 336 he said, with reference to the SPAs, that ‘[a]s part of restitution, the applicants have therefore offered to repay [the purchase prices] to the respective traders’. And in para 342, with reference to the SGAs, he said that SFF ‘has therefore tendered to pay back the respective traders the total amount of storage fees actually received as part of restitution’.

[523] Strictly speaking, Taleveras cannot rely on statements in the replying affidavit, because its claim of an accepted offer is contained in an affidavit made before the replying papers were filed. Anyway, the statements in the replying affidavit are merely descriptions, perhaps inaccurate, of the stance adopted by the applicants in their founding papers.

[524] In order for there to have been a compromise agreement, it would be necessary to find that the statements by the applicants’ deponents were made animo contrahendi. I do not consider that they were. And by the time the replying papers were filed, the applicants were well aware of the dispute which existed between Taleveras and CTSA as to which of them should receive restitution. Whatever the applicants’ view of the appropriate order, the question was sub judice and could not be resolved without the court’s endorsement.

[525] Ultimately, restitution follows as part of the just and equitable relief flowing from a setting aside of contracts. That is a matter for the court, not for private agreement, particularly where there is a third party (CTSA) with an obvious interest in the form of relief to be granted. The submission by Taleveras’ counsel in para 22 of their heads of argument recognises that even if there was a settlement, the court still needs to be satisfied that the settlement is satisfactory.

Position had Taleveras not on-sold the oil

[526] If Taleveras had retained the oil rather than on-selling it to CTSA, its misconduct would, despite the delay and the applicants’ conduct, justify a setting aside of the SPAs and SGAs. The applicants did not distinguish between the various buyers in proposing restitution. Although restitution would ordinarily follow from the setting aside of a contract, Taleveras’ SPAs and SGAs are tainted by bribery.

[527] At common law, a bribery contract (here, between Taleveras and Gamede) is void. A contract concluded between a briber (Taleveras) and the bribee’s principal (SFF) is voidable at the instance of the innocent party (Plaaslike Boeredienste (Edms) Bpk v Chemfos Bpk 1986 (1) SA 819 (A); Extel Industrial (Pty) Ltd & another v Crown Mills (Pty) Ltd 1999 (2) SA 719 (SCA). If the innocent principal rescinds the contract, is it obliged to make restitution? In Extel, where the plaintiffs were the bribers seeking to enforce the tainted contract, the second and third grounds dealt with by Nienaber JA at 730H-735B concerned restitution, and it is the third ground which is relevant here. The second ground was whether the innocent party (the defendant in that case), when giving notice of its election to rescind, has to tender restitution. Nienaber JA discussed this question generally (not specifically in the context of bribery), and concluded that a tender of restitution is not an inflexible requirement.

[528] The third ground was that because the defendant had accepted performance without payment, it had been unjustifiably enriched, so that the plaintiffs were entitled to compensation. Nienaber JA stated that the rule – that the parties should be restored to the respective positions they were in at the time they contracted – is founded on equitable considerations. The fact that the contract is tainted by bribery may be an exception to the general rule, in regard to which the learned judge of appeal said this (734H-735A):

‘If the briber is disqualified from claiming either performance (because of the maxim ex turpi causa non oritur actio) or restitution (because of the par delictum rule) from the party he bribed, there is no apparent reason why he should be treated more leniently when he seeks restitution from the party he duped. It is true that in the one case the agreement is void and in the other it is voidable, but that in itself is no reason for refusing him relief in the one case but granting it to him in the other, since his conduct in both instances is equally culpable. In both instances there may of course be circumstances justifying a relaxation of the rule which would otherwise disqualify him from claiming restitution. Those are points and considerations that may have arisen if the issue had been properly raised by the plaintiffs.’

[529] Accordingly, if Taleveras had not on-sold the oil, it might have been just and equitable to set aside the contracts without more, leaving Taleveras to sue SFF for restitution. If, despite this court’s finding of bribery, SFF had been minded to make restitution, legal action may not have been needed. But if restitution had been resisted, the facts relevant to the relaxation of the rule which would ordinarily require the briber to suffer the loss might have more appropriately been fought out in an action.

[530] However, Taleveras on-sold the oil to CTSA, and I have found that CTSA is an innocent party. In order to assess whether setting aside the contracts with or without additional financial relief would be a just and equitable remedy, one needs to have some sense of how the parties would stand contractually if the court did not set aside the contracts. This involves three contractual relationships: SFF and Taleveras; Taleveras and CTSA; and CTSA and SFF. In regard to compensation payable by SFF to CTSA as a public law remedy, the contractual qualification would again need to be satisfied.

SFF and Taleveras

[531] As between SFF and Taleveras, there seem to be no outstanding commitments. If the SPAs and SGAs stand, Taleveras does not allege a breach by SFF or claim to have suffered any damages in consequence of SFF’s conduct.

[532] For the sake of completeness, I note that the SFF/Taleveras SPAs were in very different terms to the SFF/Vitol SPA. The Taleveras SPAs are governed by English law and subject to the jurisdiction of the English courts. The breach clause (clause 14) contains no limitation on the kind of damages that can be recovered. Whether clause 6.4.3 is a general limitation of liability or a limitation operating only in the circumstances stated in the introductory part of clause 6.4 is something I need not consider. The Taleveras SGAs contain much in common with the Vitol SGA but there are many differences too. The Taleveras SGAs do not include the limitation clause found in clause 25.4 of the Vitol SGA.

Taleveras and CTSA

[533] As between Taleveras and CTSA, the relevant contract is the MRA, which is governed by English law and subject to the jurisdiction of the English courts. On 10 May 2018, about a month after the financing structure was meant to have been unwound with a repurchase of the oil by Taleveras, CTSA gave notice to Taleveras that it was terminating the MRA with immediate effect. CTSA also notified Charmondel that it was holding it responsible for any losses caused by Taleveras’ breaches.

[534] CTSA relied on clauses 12.1(f) and 13.3(c). The former clause provides that if any of the representations and warranties made by Taleveras in clause 8.1 are incorrect or untrue in any material respect, such constitutes a ‘Termination Event’, entitling CTSA, among other options, to terminate the MRA. The latter clause provides for termination by CTSA if it has been delayed or prevented from carrying out any of its obligations due to a ‘Force Majeure Event’.

[535] In regard to termination under clause 12.1(f), CTSA relied upon breaches of the warranties contained in clauses 8.1(b), (e), (f), (h), (i), (j), (n) and (p). It is by no means clear to me that those warrantees were breached. As at 14 May 2018, the position was that Taleveras had obtained good title to the oil from SFF and had passed good title to CTSA. The impugned transactions had not yet been set aside on review, and they have yet to be set aside. It is also not clear to me that clause 8.1(p) is truly a warranty. It is framed as an acknowledgment by Taleveras that CTSA is the owner of the oil. The difficulty in may 2018 was not that Taleveras had not complied with the MRA but that SFF was refusing to comply with the side letter and tank warrants.

[536] Termination in terms of clause 13.3(c) is more straightforward. A ‘Force Majeure Event’ is stated to be any event that is reasonably beyond CTSA’s control and which prevents or substantially delays it from carrying out any of its obligations under the MRA. SFF’s refusal to allow CTSA to uplift the oil, and the pending review proceedings, were events beyond CTSA’s control. They prevented or substantially delayed CTSA from giving effect to the repurchase agreement.

[537] If CTSA was entitled to terminate the MRA in terms of clause 12.1(f), of which I am doubtful, clause 12.2(d) stipulates that upon such termination the right and title to the oil will not revert to Taleveras. Because of the breach, CTSA would be entitled to sue Taleveras for damages. The heads of recoverable damages would depend on a proper interpretation of the MRA. Clause 14.1(c) read with the definition of ‘Loss’ might suggest a wide recovery. On the other hand, clause 14.3 provides that neither party may claim any sum ‘by way of indirect or consequential loss or damage, including loss of profits’, in the event of a failure by the other party to perform any of its obligations.

[538] Clause 13.3(c) simply states that in the event of a ‘Force Majeure Event’, CTSA ‘shall be entitled to terminate its obligations hereunder and under any outstanding Transaction’. The clause is silent on the consequences of such termination. On the face of it, only prospective obligations are affected, meaning that the repurchase component of the transaction falls away, leaving CTSA as the owner of the oil. If CTSA’s only right of termination is the right conferred by clause 13.3(c), it could not sue Taleveras for damages.

[539] On my reading of Charmondel’s guarantee (which is also governed by English law), its liability would not exceed that of Taleveras.

[540] Taleveras did not dispute the termination in its explanatory papers. Taleveras did, though, criticise CTSA for terminating the MRA when the review proceedings were not ripe for determination. Taleveras alleged, further, that the termination reflected an acceptance by CTSA that the sale of the oil by SFF to Taleveras was invalid, something which was inconsistent with the theme presented in CTSA’s answering papers.

[541] I do not accept the criticism of inconsistency. CTSA was faced with a de facto refusal by SFF to recognise its ownership. Whether this gave rise to a breach of warranty by Taleveras is something of which I am doubtful, but one can understand why CTSA chose to terminate the MRA. This did not imply an acceptance by CTSA that the transactions with SFF were invalid and would have to be set aside by the court. It also did not imply that CTSA no longer claimed title to the oil. In the letters which CTSA and its attorneys wrote to SFF on 29 November 2017 and 5 March 2018, CTSA adopted the firm position that it held good title.

[542] It thus appears to me that regardless of which of the two MRA clauses is applicable to the termination, title in the oil has remained with CTSA. For as long as the contracts stand, therefore, there would be no restitution as between Taleveras and CTSA, because CTSA is the owner of the oil.

[543] If Taleveras breached the MRA in the respects alleged by CTSA in its letter of 14 May 2018, the interpretation of the MRA at least casts doubt on CTSA’s right to recover from Taleveras, as damages, some of the losses it has suffered, such as the hedging loss and the premium paid to Total for the put option.


[544] As between CTSA and SFF, the contractual relationship is recorded in the side agreement letter and tank warrants. These documents go further than merely containing SFF’s consent to a transfer of title from Taleveras to CTSA:

(a) In terms of clause 3.1 of the side letter, SFF agreed that CTSA would have and retain ownership of the oil. In terms of clause 5.1, it undertook to act on CTSA’s instructions and not to refuse to carry out any operations requested by CTSA which were in accordance with the SGAs.

(b) In terms of the tank warrants, SFF confirmed that it knew that Taleveras had sold the oil to CTSA (clause 2) and that SFF had full capacity and title to issue the tank warrants (clause 4). SFF undertook to give CTSA unlimited access to the storage facilities for the purpose inter alia of protecting or enforcing its rights over the product (clause 11) and to act at all times in CTSA’s best interests (clause 14). Subject to payment of the storage fees, SFF irrevocably waived any claim to the oil and promised that it would ‘not exercise any right of set-off, counterclaim or retention or possession in respect of’ the oil’.

[545] The side letter agreement and tank warrants are governed by South African law. They contain no limitation on recoverable damages. If the impugned transactions are not set aside, CTSA as the owner of the oil would be entitled to sue SFF for specific performance (delivery) together with any damages suffered due to delayed delivery. Alternatively, CTSA could elect to cancel, forego a claim for specific performance, and claim damages.

[546] As in Vitol’s case, CTSA’s primary damages would be those flowing from SFF’s failure timeously to release the oil. CTSA’s financing structure with Taleveras was scheduled to terminate on 6 April 2018. This was the date on which CTSA needed to be in a position to sell the oil back to Taleveras or to Total or to a third party. In the face of SFF’s repudiation, CTSA’s attorneys on 29 November 2017 informed SFF that CTSA stood by its rights under the tank warrants and side letter, that CTSA had hedged its position, and that if the tank warrants and side letter were not honoured, CTSA would still have to meet its hedging obligations in April 2018, which would cause CTSA to suffer substantial losses.

[547] There having been no response to this letter or to a further letter of 31 January 2018, CTSA on 5 March 2018 again reminded SFF of the looming close-out date in early April 2018. CTSA called on SFF to confirm that the oil was available for export or an ITT, and recorded that any failure to release the oil would be a breach. CTSA sought SFF’s prompt confirmation of the foregoing within five working days. A week later the applicants launched their review application, in which SFF’s ongoing repudiation of the tank warrants and side letter was placed beyond doubt.

[548] In the circumstances, if CTSA were now to sue for specific performance of the contracts, it would get the oil at its current value and would be entitled to claim, as damages, any negative difference between the current value and the value the oil would have had on 6 April 2018. Alternatively, if CTSA were now to cancel the tank warrants and side letter, and abandon the oil to SFF, it could claim damages with reference to the value of the oil as at 6 April 2018. If the court set aside the transactions between SFF and Taleveras (meaning that CTSA did not acquire good title from Taleveras), but allowed CTSA’s contract with SFF to stand (in the form of the tank warrants and side letter), CTSA would still have a good claim for damages against SFF for breach of contract.

[549] Dated Brent on 6 April 2018 was $65.903. Applying the net discount of $2.55 agreed between Taleveras and CTSA, the oil would have been worth $253,412,000,[79] which would enable CTSA to achieve full restitution plus recoupment of out-of-pocket expenses while still leaving it with a profit. The contractual qualification would thus be met if I were instead to set aside the transactions and award CTSA its out-of-pocket expenses.

[550] As in Vitol’s case, I do not think that SFF could successfully resist a contractual claim from CTSA based on Gamede’s lack of authority. The board ratified the Taleveras transactions on 5 February 2016, and the applicants subsequently defended them publicly. Again one has the fact that storage fees were charged for several years. We do not know whether the side letter agreement was placed before the board on 5 February 2016 but I do not think it matters. The important authorisation related to SFF’s disposal of the oil to Taleveras. Once that transaction was ratified, the recognition of a transfer of title from Taleveras to CTSA (or to any other third party) was a standard operational matter which one would expect to fall within the usual authority of the CEO.

Final assessment

[551] The applicants submitted that CTSA was ‘double-dipping’ in seeking compensation from the applicants, because it was also laying claim to damages against Taleveras and Charmondel. I disagree. It often happens that an aggrieved party has remedies against multiple parties. Usually these may be pursued in parallel. The law will obviously not allow double compensation. Any compensation awarded in these proceedings to CTSA will reduce its claims against other parties. CTSA has given a formal undertaking in this regard which I shall record in my order.

[552] I have expressed reservations about whether Taleveras breached the warrantees in the MRA. It is by no means clear, therefore, that CTSA would have a claim for damages against Taleveras. If the contracts between SFF and Taleveras were set aside (so that Taleveras did not pass good title to CTSA), one would need to know what effect such an order would have on the MRA, which is governed by English law. It seems tolerably clear that the effect would be to discharge the MRA in terms of the English doctrine of frustration (see Chitty On Contracts: General Principles 31 ed (2012) chapter 23). This would give rise to essentially restitutionary remedies in terms of the Law Reform (Frustrated Contracts) Act, 1943 (Chitty §23-074 ff). There would be no claim for damages.

[553] Furthermore, there is evidence casting doubt on Taleveras’ financial standing. CTSA insisted on getting ownership of the oil, and on receiving undertakings and assurances from SFF, precisely so that it would not be exposed to Taleveras’ creditworthiness. It is not just and equitable that CTSA should now be cast adrift with only a speculative damages claim against Taleveras, in circumstances where substantially the whole fault – as between CTSA and the applicants – lies with SFF.

[554] I have explained why, at common law, Taleveras would not necessarily have a right to claim restitution from SFF if its agreements with SFF were rescinded. This is not the only reason why a just and equitable remedy should not result in the restitution of the purchase price and storage fees to Taleveras. It was CTSA’s money, not Taleveras’, that went to SFF.

[555] Taleveras’ counsel argued that the MRA between CTSA and Taleveras is not one of the contracts that will be set aside if the review succeeds. In making payment to SFF, CTSA was discharging its obligations to Taleveras under the MRA, and such payments were not sine causa as between CTSA and Taleveras. If the court were to order the applicants to make restitution to CTSA rather than Taleveras, the court would effectively be granting a judgment in favour of CTSA against Taleveras, in violation of the latter’s rights under s 34 of the Constitution.

[556] I accept that in law CTSA’s payments to SFF discharged CTSA’s obligations to Taleveras and were not, as between CTSA and Taleveras, sine causa. However, Taleveras would be under a legal obligation to make restitution to CTSA if it did not pass good title to the latter. This flows from the fact that setting aside the contracts between SFF and Taleveras would discharge the contract between Taleveras and CTSA in accordance with the English doctrine of frustration. This is so self-evident that to give it recognition in a just and equitable remedy does no violence to s 34. Taleveras has not explained on what conceivable basis it would – as between itself and CTSA – be entitled to retain a restitutionary payment if all the contracts were set aside. So no injustice is done by requiring the restitution to be paid directly to CTSA.

[557] In economic reality, if not in law, CTSA’s payments to SFF were sine causa, because they were made against assurances by SFF that CTSA’s ownership of the oil was recognised. Injustice might be done to CTSA if the money were paid to Taleveras, because there is evidence that the latter’s financial standing is questionable. Exposure to Taleveras’ creditworthiness is precisely what CTSA’s structure was designed to avoid.

[558] In any event, it is unnecessary, as between CTSA and SFF, to construe the repayment of the purchase price and storage fees as restitution pursuant to the setting aside of the Taleveras SPAs and SGAs. Let us accept, for the moment, that the money which landed up with SFF was money which Taleveras in law paid to SFF. As between SFF and Taleveras, the setting aside of the SPAs and SGAs would involve restitution between those two parties. But at the same time the setting aside of the CTSA side-letter and tank warrants means that CTSA is out-of-pocket for the money which it paid to Taleveras to fund the purchase price and storage fees. This out-of-pocket expense can be awarded to CTSA as compensation, in the same way that money it paid to Total, to the insurers and to the inspectors may be awarded as compensation.

[559] On this basis, the court is faced with two alternatives: to award restitution to Taleveras or compensation to CTSA. The choice must be exercised on the basis of what is just and equitable. I have no hesitation in preferring compensation to CTSA than restitution to Taleveras.

[560] CTSA has an additional argument, namely that any right which Taleveras might have to receive restitution if the SPAs and SGAs are set aside falls within the definition of ‘‘Rights’ in the MRA[80] and was thus, in terms of clause 3.4, assigned to CTSA together with the oil (and see also clause 8.1(p)). Both parties filed supplementary written submissions dealing with this issue. In favour of CTSA’s argument is that the definition (a) includes rights under or pursuant to judgments and awards; (b) is expansively framed (‘all’, ‘any nature whatsoever’, ‘whether actual, perspective or contingent, direct or indirect’; and (c) expressly includes rights and interests that were not within the contemplation of the parties when the MRA was concluded. The commercial context of the MRA would also justify a generous interpretation. CTSA wanted every right and interest connected with the oil to vest in it; it did not wish to have to rely on Taleveras’ creditworthiness at all. However, in view of the other conclusions I have reached, it is not necessary to reach a final opinion on this issue.

[561] In the circumstances, I consider it just and equitable to set aside the transactions between SFF and Taleveras and between SFF and CTSA, subject to payment of compensation for CTSA’s out-of-pocket expenses. I will, however, exclude the one month’s storage fee of $480,000, which was not paid to SFF. I think it is reasonable to expect CTSA to recover that amount from Taleveras.

[562] There is one further component of CTSA’s expenses which should at this stage be excluded from the compensation, viz the net balance of $22,568,426 which CTSA paid to Taleveras. The provisional disallowance of this item is not concerned with any misconduct on CTSA’s part. However, it represents Taleveras’ profit. In view of Taleveras’ misconduct, the latter should not be entitled to retain it. If CTSA were to recover $22,568,426 from SFF as part of its compensation, the economic result would be that Taleveras would get to keep its profit and such profit would have been paid by SFF. Despite the bribery, it is not clear to me that SFF could recover the money from Taleveras, since it was not SFF but CTSA which parted with this amount. By carving out the profit element from CTSA’s compensation, I would in effect force CTSA to look to Taleveras to recover this out-of-pocket expense. The discharge of the MRA by frustration appears to accord a straightforward basis for recovery. I shall, however, allow CTSA to recoup this amount from SFF if it is unsuccessful in excussing Taleveras.

[563] The total amount to be awarded to CTSA is thus $208,702,648, which includes the amount of $123,865,600 which the applicants have tendered in respect of the purchase price and storage fees. The balance, constituting the disputed compensation, is $84,837,048.[81]

Just and equitable remedy – Venus and Glencore

[564] Since there is no dispute between the applicants and Glencore, and since Venus has not entered the lists, I can deal with Glencore’s position briefly. The only question is whether it would be just and equitable for me to endorse the settlement between the applicants and Glencore.

[565] The position that currently prevails (ie before setting aside any transactions) is that Glencore is the owner of the oil. Like CTSA, it could seek specific performance from SFF or claim damages. On 11 September 2017 Glencore and Venus gave SFF notice under the SGA that same was being terminated as at 31 December 2017. This was due notice under clause 2.1 of the SGA as amended on 31 December 2015. By refusing to release the oil on that date, SFF fell into breach. Dated Brent at the end of 2015 exceeded $66. Any negative difference between the current value and $66 would be recoverable by Glencore as damages. Even at current values, the value of the oil would exceed the limited restitution which Glencore is willing to accept. The contractual qualification is thus satisfied.

[566] I consider that restitution should go to Glencore, not Venus, my reasons being similar to those that apply as between CTSA and Taleveras. The proposed restitution is the price and storage fees paid by Venus to SFF. Venus’ mark-up is not covered, nor does Glencore seek to recover any of its further out-of-pocket expenses. Because Glencore will not be recovering the mark-up from SFF, Glencore will need to look to Venus to recover those amounts. This is similar to the outcome I propose as between CTSA and Taleveras.

[567] I am thus satisfied that the draft order proposed between the applicants and Glencore should be approved. The applicant’s counsel’ submitted that this should have some bearing on what I ordered in relation to Vitol and CTSA. I disagree. I do not know Glencore’s reasons for accepting limited restitution and foregoing a claim for out-of-pocket expenses. Its attitude can have no bearing on what is just and equitable for Vitol and CTSA.

The time-value of money

[568] The purchase price which SFF received has been invested by it in dollar accounts. The applicants have proposed to make restitution plus interest equal to the interest actually earned on the purchase price plus notional interest on the storage fees at a similar rate. Glencore is content with this.

[569] Vitol and CTSA, on the other hand, seek interest at their actual cost of funding, and ask for such interest not only on the purchase price and storage fees but also on their other outgoings. It must be emphasised that one is not here concerned with whether, in a contractual claim for damages, Vitol and CTSA could recover the cost of funding, either as damages or as mora interest. As I have explained, if their contracts were to stand and they claimed damages, such damages would be determined with reference to the value of the oil at the relevant date, and as I have shown that value substantially exceeds their out-of-pocket expenses. If the cost of funding represents an out-of-pocket expense falling within the contractual qualification, it appears in principle to stand on the same footing as other out-of-pocket expenses, and there is no reason not to award it.

[570] Furthermore, the gross delay of which the applicants have been guilty makes it particularly unjust, in this case, that Vitol and CTSA should be out of pocket in respect of the time-value of money. If the applicants had launched and prosecuted their application expeditiously, compensation would have become payable to Vitol and CTSA much sooner than will now be the case.


[571] In Vitol’s main answering affidavit, Foster baldly stated that Vitol’s ‘cost of capital (time value of money cost)’ used to buy, store, insure and hedge the oil totalled $7,939,610. No details were given. In his post-hearing supplementary affidavit, Foster stated that the aforesaid figure was only the cost of capital on the purchase price, not on other outgoings as mistakenly suggested. He explained that Vitol claimed its cost of capital based on its internal cost of capital for the group. The group’s finance department calculated the time value of money based on the ‘blended cost of borrowing’ of Vitol and its affiliates. In para 10 he tabulated the rates from November 2015 to July 2020.

[572] If one compares Vitol’s table with PwC’s annexure ‘TW16’, attached to its post-hearing report of 12 October 2016, it is apparent that Vitol’s cost of borrowing is substantially higher than the interest rates earned by SFF on its dollar deposits, particularly in the period March 2016 to February 2017, where Vitol’s rates are between 2.6 – 4.6 times higher than SFF’s deposit rates. The gap narrows for the balance of the period, though Vitol’s rates are still more than double SFF’s rates.

[573] Vitol does not state that Vesquin or any other company in the Vitol group actually incurred on outgoing at the rates specified by Foster. Vitol is a substantial group. Its blended cost of borrowing presumably reflects an average of the cost of borrowings of a number of different companies. Some might have higher costs, others lower. I do not know what the cost of borrowing was in relation to the specific outgoings in the present case. Indeed, I do not even know that Vitol had to borrow money to fund these outgoings. The fact that a group has a blended cost of borrowing does not mean that every outgoing it incurs is paid with borrowed money.

[574] Particularly since Vitol’s blended cost of borrowing is significantly higher than SFF’s deposit rates, and also significantly higher (as will appear) than CTSA’s borrowing costs, I do not feel justified, in the absence of better evidence, in compensating Vitol at the claimed rates. However, it is fair assumption that if Vitol had not had to incur its fruitless outgoings, it could at least have invested the money at rates equal to those that SFF earned on its dollar deposits. I shall thus allow Vitol compensation for the time value of money at the rates specified in PwC’s annexure. In its report of 12 October 2020, PwC has calculated these amounts on the tendered restitution from date of payment up to 31 October 2020, and I shall incorporate those amounts in my order.


[575] In the answering affidavit, CTSA stated that Natixis had to source funding from the market for the facility of $164,322,400 which it granted Taleveras. The funding cost was 0.695 % over the period February 2016 – March 2018. Because CTSA did not have access to the oil on 6 April 2018, this funding cost has continued to accrue not only on the facility but also the hedging loss. Vanlerberghe states that the amounts reflected in his schedule ‘SA16’[82] have actually been paid by Natixis which has passed on the cost to CTSA at no mark-up. Natixis, he says, would seek to obtain optimal rates for the cost of funding.

[576] A comparison of Vanlerberghe’s schedule with PwC’s schedule shows that except for the period February 2016 to November 2016, when Natixis’ borrowing rates were slightly higher than SFF’s deposit rates (but less than double), Natixis’ rates have been lower than SFF’s deposit rates. In the circumstances, it would be fair and equitable to award CTSA compensation in the amount of its actual funding costs.

[577] The question remains as to the capital amounts on which this funding cost should be allowed. CTSA seeks to recover interest on the full facility of $164,322,400 rather than on the individual components which this facility funded. The facility covered CTSA’s profit (treasury fees and premiums). It also covered the net amount of $22,568,426 which CTSA paid out to Taleveras but which, at least at this stage, I do not intend to award to CTSA as compensation. I think that the interest should be confined to the outgoings I have actually allowed, namely $125,102,648 out of the facility,[83] plus a further amount of $83,600,000 as from 6 April 2018.

[578] Although CTSA apparently seeks interest on the full facility as from February 2018, it is not clear to me why interest on the full amount should run from that time. It is fair to assume, in the absence of further explanation, that the facility was only drawn down as and when various outgoings were paid:

(a) The purchase price of $112 million was paid to SFF on 25 February 2016, so interest will be allowed on that sum from that date.

(b) The put agreement was concluded with TOSA on 5 February 2016, so I shall allow interest on $800,000 from that date.

(c) In the case of storage fees, CTSA’s schedule gives the invoice date, not the payment date.[84] PwC has likewise not given the dates on which SFF actually received the storage fees but has calculated notional interest from an assumed date of payment.[85] I shall thus allow interest from those assumed payment dates.

(d) In the case of insurance premiums totalling $408,853, the information as to when the various months’ premiums were paid is patchy.[86] Where a payment date has been supplied,[87] I shall allow interest from such date. For the balance of the premiums, I shall allow interest as from 1 June 2018, which is a conservative assumption against CTSA.

(e) In the case of the inspection fees, CTSA’s schedule does not give payment dates.[88] However, the inspection invoices required payment within 30 days of invoice, so I shall (again conservatively against CTSA) allow interest on each monthly invoice from the first day of the second month following the month in which the invoice is dated (so that, for example, interest will run on the invoice dated 19 April 2016 from 1 June 2016).

Mora interest

[579] Vitol and CTSA submitted that they are entitled to interest at the rate prescribed in terms of the Prescribed Rate of Interest Act 55 of 1975 from date of demand, which they identify as being the date on which they filed their main opposing affidavits, 22 May 2020. In his post-hearing affidavit, Foster makes bold to claim, on the basis of legal advice, that the court does not have the power to depart from such rate. The prescribed rate of interest as at 22 May 2020 was 8.75%. The rate at the time of this judgment is 7.25%. In accordance with authority, whatever prescribed rate is applicable when the prescribed interest starts running is not affected by subsequent changes in the rate.

[580] Vitol errs in asserting that the court cannot deviate from the prescribed rate. In regard to interest on judgment debts, s 2(1) of the Act states that the prescribed rate will be applicable ‘unless that judgment order provides otherwise’. In regard to interest on unliquidated amounts which were demanded before the judgment date, the prescribed rate applicable in terms of s 2A(1) is made subject to the further provisions of s 2A. Section 2A(5) provides that a court may make ‘such order as appears just in respect of the payment of interest on an unliquidated debt, the rate at which interest shall accrue and the date from which interest shall accrue’.

[581] The compensation which I propose to award Vitol and CTSA does not represent a ‘debt’ within the meaning of s 2A. At no stage before my awarding of the compensation would a debt have existed, liquidated or otherwise, for the payment of such compensation. In terms of the contracts, Vitol and CTSA would have had unliquidated claims for damages, but those are not the debts which will be enforced by my judgment.

[582] In my view, therefore, the interest to be awarded to Vitol and CTSA on their outgoings, up to the date of my judgment, should be at the rates already discussed. As from the date of my judgment, the compensation awarded to Vitol and CTSA will constitute a judgment debt. The question is whether I should exercise my power, in terms of s 2(1), to deviate from the prescribed rate of 7.25%. Although s 2(1) does not specify the test for deviation, it seems to me that it must be similar to the one laid down in s 2A(5), namely whether it would be just to do so. The prescribed rate is more than four times higher than CTSA’s cost of funding and SFF’s deposit rate. Since the applicants will be ordered to pay compensation in US dollars, they will need to fund compensation in international money markets. Vitol and Natixis operate in global financial markets, and it is Natixis’s cost of borrowing which is passed on to CTSA. The time-value of money for them is not determined by the South African economic environment.

[583] Having regard to the foregoing considerations, and to the fact that I am ultimately concerned with making a just and equitable order in terms of the Constitution and PAJA, I think it would be just to deviate from the prescribed rate. During July-September 2020 was earning 1.45% on its dollar investments, which is higher than CTSA’s rate. According to the explanatory note pertaining to the Glencore draft order, the average interest rate which SFF earned from March 2016 to October 2020 was 1.81%, which was used in computing the interest component in the draft order. Since the interest rate I now allow may have to run for some time (appeals being likely), I think justice would be done by granting interest on the judgment debt at this average rate, compounded monthly in arrears. (I am aware that the rate in terms of the Prescribed Rate of Interest Act is simple interest, but the lower rate which I am allowing is based on monthly compounding.)

[584] CTSA’s counsel referred me to Morgan Stanley Capital Group Inc v Strategic Fuel Fund Association; Strategic Fuel Fund Association v Morgan Stanley Capital Group Inc [2019] ZAGPJHC 332, where the court rejected an argument that the common law needed to be developed in regard to interest payable on amounts awarded in foreign currencies. I agree with that view. The Act already gives the court all the flexibility it needs. For the rest, the case decided that there were no grounds to vary the arbitrators’ award of interest at the prescribed rate, since the debtor had not asked the arbitrators to depart from the prescribed rate.

[585] The other two cases mentioned in argument on this aspect likewise do not militate against my view. In Skilya Property Investments (Pty) Ltd v Lloyds of London Underwriting Syndicate Nos 960, 48, 1183 & 2183 2002 (3) SA 765 (T) the debtor’s argument was that the court should order mora interest to run on the rand equivalent of the dollar amount awarded. The court rejected that argument (816H-817B). In NMB Bank Limited v Capsopoulos & another [2017] ZASCA 94; [2017] 3 All SA 765 (SCA), interest at the prescribed rate on a US dollar award was ordered in the absence of any objection from the debtor (para 36). In neither case was the court asked to consider a different rate.

Costs and order

[586] Although the applicants will succeed in having the impugned decisions and transactions set aside, Vitol and CTSA have achieved substantial success. To the extent that they resisted the review on the basis of delay, they had reasonable grounds to do so. The most contentious issues concerned compensation. The applicants must thus pay Vitol and CTSA’s costs. The case warranted the employment of three counsel, given its volume and complexity. The applicants themselves engaged four counsel. Taleveras will need to bear its own costs.

[587] I have made non-substantive adjustments to the formulation of the amended notice of motion and to the draft order handed up by the applicants and Glencore. My order relates to 10 million barrels, rather than 10,3 million barrels, and I have excluded references to Enviroshore/GNI, since the latter parties have not been cited and it would not be appropriate for me to grant any relief in relation to the 300,000 barrels in which they might be interested.

[588] I make the following order:

(1) Dollar amounts stated herein are United States dollars.

(2) The applicants are granted an extension in terms of s 9 of the Promotion of Administrative Justice Act 3 of 2000 for launching this application outside the 180-day period contemplated in s 7(1) of the said Act.

(3) The following decisions and contracts are declared invalid and are set aside:

(a) the decision of the ninth respondent (‘the MoE’), taken on or about 12 November 2015, to approve a disposal by the second applicant (‘SFF’) of 10 million barrels of strategic crude oil reserves (‘strategic stock’);

(b) SFF’s decision, taken during or about late November 2015, to award contracts to buy portions of the strategic stock to the first respondent (Venus), the third respondent (‘Taleveras’) and the sixth to eighth respondents (collectively ‘Vitol’);

(c) the decision of the MoE, taken on or about 7 December 2015, to approve transactions for the disposal of 10 million barrels of strategic stock to Venus, Taleveras and Vitol;

(d) the decision of SFF’s board of directors, taken on or about 5 February 2016, to ratify the conclusion of transactions with Venus, Taleveras and Vitol relating to the strategic stock;

(e) the sale and purchase agreements and storage agreements, and all amendments thereto and novations thereof, concluded between SFF on the one hand and Venus, Taleveras and Vitol on the other, relating to the said 10 million barrels of strategic stock;

(f) the tripartite agreements concluded between SFF, Venus and the second respondent (‘Glencore’), and all amendments thereto and novations thereof, relating to 3 million barrels of Bonny Light forming part of the strategic stock;

(g) the side-letter agreement concluded between SFF, Taleveras and the fourth respondent (‘CTSA’), and all amendments thereto and novations thereof, relating to 2 million barrels of Basrah Light and 2 million barrels of Bonny Light forming part of the strategic stock;

(h) all tank warrants issued to any of the respondents in respect of the strategic stock;

Venus and Glencore

(4) The applicants jointly and severally must pay Glencore $106,223,889.12 on or before 25 November 2020, together with interest thereon at a rate of 1.81% from 15 October 2020 to date of payment, which amount the applicants and Glencore have agreed is a combined and all-inclusive sum constituting just and equitable relief.

(5) No restitution or other compensation shall be payable to Venus in respect of any of the above transactions which have been set aside.

(6) By agreement, the applicants and Glencore shall bear their own costs in the main application and all interlocutory applications insofar as such applications relate to Venus and Glencore.

Taleveras, CTSA and the fifth respondent (‘Natixis’) 

(7) The applicants jointly and severally must pay CTSA:

(a) $123,865,600 as restitution of purchase price and storage fees;

(b) $84,837,048 as just and equitable compensation for out-of-pocket expenses other than interest;

(c) further amounts, in dollars, as just and equitable compensation in respect of interest, compounded monthly in arrears and calculated on the following amounts and from the following dates up to 4 April 2018 at the rate of 0.695% and from 5 April 2018 to date of judgment at the rates set out in the schedule attached hereto as ‘J1’:

(i) on $112,000,000 from 25 February 2016;

(ii) on $800,000 from 5 February 2016;

(iii) on storage fees totalling $11,865,600, on each month’s fee from the ‘assumed date of payment’ reflected in the schedule attached as ‘J2’;

(iv) on insurance premiums totalling $408,853, from each date of payment to the extent that a date of payment is reflected in the schedule attached as ‘J3’, and on the remaining premiums from 1 June 2018;

(v) on inspection fees totalling $28,195, from the first day of the second month following the month of ‘invoice date’ in the schedule attached as ‘J4’;

(i) on $83,600,000, from 5 April 2018.

(8) The total of the compensation set out in para (7) shall bear interest, from date of judgment to date of payment, at the rate of 1.81%, compounded monthly in arrears.

(9) The following undertaking, given by CTSA and Natixis, shall be operative in relation to the compensation awarded to CTSA in terms of this order:

‘CTSA and Natixis will not seek to recover from Taleveras or Charmondel Holdings Ltd any amounts which they receive pursuant to this order, and any such amounts will be taken into account, in accordance with English law, to reduce their claims in any other proceedings they may bring against Taleveras or Charmondel.’

(10) Subject to the fulfilment of the conditions stated in (11) below, the applicants jointly and severally must pay further compensation to CTSA in the amount of $22,568,426 (being the net amount of the purchase price for the oil paid by CTSA to Taleveras in terms of the sale confirmation dated 5 February 2016 issued in terms of the master repurchase agreement concluded between CTSA and Taleveras on 5 February 2016).

(11) The liability to pay the compensation in (10) above shall only come into existence and be enforceable if, and to the extent that, CTSA is unable to recover the said amount from Taleveras after exhausting all reasonable steps to do so.

(12) The applicants must pay CTSA and Natixis’ costs, including the costs of three counsel.

(13) No restitution or compensation shall be payable by the applicants to Taleveras pursuant to the setting aside of the contract between those parties.

(14) No order as to costs is made as between the applicants and Taleveras.


(15) The applicants jointly and severally must pay Vitol:

(a) $86,826,000 as restitution of purchase price and storage fees;

(b) $19,049,944 as just and equitable compensation for out-of-pocket expenses other than interest;

(c) further amounts, in dollars, as just and equitable compensation in respect of interest, calculated at the rates set out in the schedule attached as ‘J5’ up to 31 October 2020 and at the rate of 1.81% thereafter to date of judgment, compounded monthly in arrears, as follows:

(i) on $78,606,000, from 11 March 2016 to 31 October 2020 in the amount of $6,874,030 plus further interest from 1 November 2020 to date judgment;

(ii) on each month’s storage fees, in total $8,220,000, from date of each payment to 31 October 2020 in the amount of $668,817, plus further interest from 1 November 2020 to date of judgment;

(iii) on $37,530 (cost of letter of credit), from 21 January 2016 to date of judgment;

(iv) on each month’s insurance premiums (in total $933,487), from the first day of the month immediately following the month in which such premium was paid as set out in the schedule attached as ‘J6’ to date of judgment;

(v) on $18,078,928 (hedging losses), from 7 May 2020 to date of judgment,

(16) The total of the compensation set out in para (15) shall bear interest, from date of judgment to date of payment, at the rate of 1.81%, compounded monthly in arrears.

(17) The applicants must pay Vitol’s costs, including the costs of three counsel.



O L Rogers

Judge of the High Court

Western Cape Division


For Applicants

T Motau SC (with him R Tshetlo, S Scott and U Gcilishe)

Instructed by

Webber Wentzel

15th Floor, Convention Tower

Heerengracht Street

Cape Town

For Second Respondent

A Subel SC (with him A Smalberger SC and F Ismail)

Instructed by

Werksmans Attorneys

96 Rivonia Road

Sandton, Johannesburg

For Third RespondentL

Kuschke SC (with him U Naidoo)

Instructed by

Knowles Husain Lindsay Inc

10th Floor, 2 Long Street

Cape Town

For Fourth & Fifth Respondents

G Marcus SC (with him K Hofmeyr and M Mbikiwa)

Instructed by

Norton Rose Fulbright South Africa Inc


c/o Norton Rose Fulbright House

8 Riebeek Street

Cape Town

For Sixth to Eighth respondentsW Trengove SC (with him I Goodman and n Nyembe)

Herbert Smith Freehills South Africa LLP

4th Floor, Rosebank Towers

15 Biermann Avenue



For Organisation Undoing Tax Abuse (Amicus Curiae)

Kameshni Pillay SC (with her M Dafel)

Instructed by

Pandor Attorneys

774 Waterval Road

Little Falls




[1] At 5208

[2] At 6.

[3] Jago para 4.27 at 4118.

[4] KPMG report para 1.2.2 at 1305.

[5] The three-year forward curves of the contango market as at 15 and 28 December 2015 and 15 January 2016 are shown at 5301.

[6] The present case concerns 10 million barrels. The other 300,000 had apparently been committed in some way or other to a firm called Enviroshore.

[7] API = American Petroleum Institute. API, or API gravity, is an inverse measure of crude oil's density relative to water. The higher the API, the lighter (and more commercially attractive) the oil.

[8] Crude is 'sweet' or 'sour' depending on whether it has a lower or higher sulphur content. In general, lower sulphur content is more commercially attractive.

[9] The various amending agreements do not have reliable numbering. I number the amending agreements in strictly chronological order.

[10] The reason for this may be that the applicants first saw the Venus/Glencore SPA when the answering papers were filed in May 2020. The storage rate was in an annexure which may have been overlooked.

[11] See clause 3 of the Taleveras SPA at 355-366 and the Taleveras amendment at 396-481.

[12] See Driscoll para 29 at 5273, para 57 at 5282, para 101 at 5299.

[13] There are minor differences in the exact amount appearing at different places in the papers. I take this figure from the explanation submitted together with proposed draft order between the applicants and Glencore.

[14] CFR = Cost and Freight. This would apply if the oil was shipped to Saldanha Bay.

[15] Para 132 at 5309.

[16] Intercontinental Exchange.

[17] Para 8 at 5590.

[18] Paras 11-12 at 5595-5596.

[19] At 5247 and 5251.

[20] At 5248 and 5256.

[21] At 5260-5261.

[22] Basamian’s supplementary report of 10 September 2020 at 6187-6193. Although the sampled Basrah was in Tank 5, this is, as I understand the evidence, the tank into which the Basrah from Tank 2 was pumped in the latter part of 2019.

[23] Para 35.1 at 3200.

[24] Para 93 at 4665.

[25] At 3809.

[26] Contango Trading SA & others v Central Energy Fund [2019] ZASCA 191; 2020 (3) SA 58 (SCA).

[27] My references are to the amended notice of motion at 5208-16.

[28] Para 8.

[29] Paras 3.1, 4.1 and 6.1. The date does not appear in the notice of motion but emerges from the evidence.

[30] Para 2.

[31] Paras 3.2, 3.3, 4.2, 4.3, 6.2 and 6.3.

[32] Para 3.4.

[33] Para 5. (The notice of motion incorrectly refers to Natixis rather than CTSA.)

[34] Para 9.

[35] Para 7.

[36] Para 87 at 5293-5294 and para 136 at 5312.

[37] See paras 1.2.1(c) and 16A.1 of the Treasury Regulations.


[38] Bossley para 18 at 3140, paras 53-54 at 3151. See also Driscoll para 26 at 5272, paras 88-94 at 5294-5297. It is not without significance that according to Bossley she was currently engaged in selling oil on behalf of a subsidiary of a national oil company through a process of direct negotiation with likely buyers. If direct negotiation was to be used by SFF, the skills of an international oil expert such as Bossley might have prevented the rot.

[39] The question of bribery was squarely raised in the supplementary founding papers (paras 79, 82, 87, 89-82 and 100-102) and again in reply to Taleveras' explanatory affidavits (paras 382-384, 389 and 401-406). Although the applicants may have dealt with this in the context of misconduct bearing on just and equitable relief rather than as a distinct ground of review, it seems to me that if it was established as misconduct, it constitutes a ground of review.

[40] This can be seen by comparing the signature on his affidavit (at 5004) with the signatures on the contracts: see, for example, the original SGAs at 2451 and 2502, and the original SPAs at 2536 and 2558.

[41] Paras 80-87 at 3159-3164.

[42] Para 101 at 5299.

[43] Para 108 at 5302.

[44] This is shown by Mr Driscoll’s graph at 5301, showing the forward curve is as at 15 December 2015, 28 December 2015 and 15 January 2016.

[45] Para 111 at 5303.

[46] Foster para 198 at 3262.

[47] Founding affidavit para 285 at 119.

[48] Para 99 at 3167.

[49] Para 24 at 3142-3143.

[50] Para 4.81 at 4131.

[51] Para 122 at 5307.

[52] Record 1323.

[53] Record 1580.

[54] Para 138 at 5312. His table at 5313 assumes that the pollution and nitrogen charges were additional monthly fees. This is incorrect. It is clear from the SGAs that the nitrogen charge was a once-off fee if and when cargo was loaded while the pollution charge was a once-off the if and when cargo was discharged from a vessel. Neither of these charges were applicable to the ITTs, and SFF did not bill the relevant respondents for such charges.

[55] Para 86 at record 2747 reads with schedule at 2962.

[56] Para 32 record 5485.

[57] Para 216 at record 2797.

[58][58] Record 2909. The treasury fees and premium, hedging premium and flat transaction fees in the confirmation total $14,908,190.

[59] See Foster’s supplementary affidavit dated 21 September 2020, para 6.5.

[60] Venus was buying at a discount of $4 to Dated Brent while Glencore was buying at a premium of $2, so Venus would be making $6 × 3 million = $18 million.

[61] Venus was paying a monthly storage fee of $0.11 while Glencore was paying $0.25, meaning that per month Venus would make a profit of $0.14 × 3 million = $420,000.

[62] De Wet para 5 at 5606.

[63] Para 347.

[64] See Vanlerberghe's contemporaneous note at 2964-5.

[65] Gobodo interview at 1594 and 1617. This is confirmed by De Wet para 4 at 5605-5606, who says that similar discussions were also planned with other traders but did not materialise..

[66] Nigerian National Petroleum Corporation.

[67] Paras 92-93 at 5296.

[68] 34 - 29.66 = 4.34 × $0.40 = $1.74

[69] The date can be inferred from Foster’s email to Chili of 18 December at 1132.

[70] Clause 24 at 1143.

[71] Clause 24 at 534.

[72] Clause 26 at 535.

[73] Paras 12-15 at 5596-5597.

[74] Para 18 at 5966-5967.

[75] Record 5603.

[76] In Gauteng reliance damages cannot be claimed (Hamer v Wall 1993 (1) SA 235 (T) per majority judgment; Drummond Cable Concepts v Advancenet (Pty) Ltd (08179/14) [2018] ZAGPJHC 636; 2020 (1) SA 546 (GJ) paras 11-22. In Natal reliance damages can be claimed if the plaintiff has cancelled the contract (Probert v Baker 1983 (3) SA 229 (D)). In the Western Cape and Eastern Cape reliance damages can be claimed even if the plaintiff has not cancelled the contract (Mainline Carriers (Pty) Ltd v Jaad Investments CC & another 1998 (2) SA 468 (C); Emadyl Industries CC t/a Raydon Industries v Formex Industries (Pty) Ltd t/a Formex Engineering 2012 (4) SA 29 (ECP) para 60).

[77] Although it is said that an election to cancel must be made within a reasonable period of time, a guilty party’s persistent refusal to repent is an ongoing repudiation, and in appropriate circumstances the aggrieved party can elect at a later time to throw in the towel and claim damages instead of specific performance (Primat Construction CC v Nelson Mandela Bay Metropolitan Municipality [2017] ZASCA 73; 2017 (5) SA 420 (SCA) paras 25-28).

[78] $60 - $5.50 = $54.50 × 3 million = $163,5 million.

[79] $65.903 - $2.55 × 4 million = $253,412,000.

[80] The MRA defines ‘Rights’ as meaning ‘all of Taleveras' rights (including any rights under or pursuant to any judgment or award in favour of Taleveras), entitlements and benefits of any nature whatsoever in and interests and claims of any nature whatsoever to, under or pursuant to, as the case may be, the Product and the Equivalent Product, howsoever evidenced, whether actual, prospective or contingent, direct or indirect, whether a claim to payment of money, the delivery of Equivalent Product or to the performance of any other obligation, and whether or not the said rights and interests were within the contemplation of the Parties at the date of this Agreement.’

[81] $83,600,000 (hedging loss) + $408,853 (insurance) + $28,195 (inspections) and $800,000 (Total) = $84,837,048.

[82] At 5735-5738.

[83] $112 million (SFF purchase price) + $11,865,600 (storage fees actually received by SFF) + $408,853 (insurance) + $28,195 (inspections) + $800,000 (Total) = $125,102,648.

[84] At 2962.

[85] See annexure ‘TW12’ of PwC's report dated 12 October 2020.

[86] See CTSA schedules at 3099 (containing some payment dates) and 5667 (containing no payment dates).

[87] As per the schedule at 3099.

[88] At 3100.