Can the banks survive EWC?

Charles Collocott writes on land expropriation and SA's financial institutions

Land expropriation and South Africa's financial institutions

4 September 2018


With the ANC having resolved to amended section 25 of The Constitution, what will expropriation without compensation mean for the financial institutions of South Africa? Will they be able to weather this storm, and if so, what will be the consequences?


Cas Coovadia, MD of the Banking Association of South Africa (BASA), has called the current narrative around expropriation ‘irrational’ with no concern for ‘investment and growth.’[1] If the government were to expropriate without compensation, land on which the owner still owes money, how does the bank get the funds owed upon this default? As things stand, section 8 of the Expropriation Act [2] holds that expropriation extinguishes the mortgage bond but not the debt. In other words, the owner still owes the bank for what becomes an unsecured loan, typically with a higher interest rate attached.

There will therefore no longer be land/property/security available to the bank, as the bank’s relationship is with the original land owner and not the government. What will be the practical and legal consequences of this? Will the bank be forced on both grounds to lose its claim to payment, and if so, how does this fit into the rule of law?[3] While these legal questions are extremely important, this brief focuses on the practical assumption that the finances owed on expropriated land will no longer be available to the banks.


Recent estimates have the total South African financial institution exposure to the agricultural sector at around R160bn. Economists have the commercial banks exposure at 70%[4], while government have it at 56%[5] - this is a range of R112bn to R89.6bn. For FirstRand this makes up 3.6% of their total loan book, 3.4% for ABSA, 2% for Standard Bank and 1% for Nedbank.

Included in the total R160bn total exposure of financial institutions to the agricultural sector, is the Land Bank - a public entity. The Land Bank’s exposure figures range from R40bn[6] to R46.6bn. R16.2bn of this is secured by mortgages and R25.5bn has been advanced to agribusiness and co-ops. However, much of these advancements to agribusiness and co-ops are loaned to farmers to acquire land, and therefore the Land Bank’s exposure could comfortably exceed 50% of their assets.[7]


Within the ANC’s policy documents, vacant land has been identified as one of the categories for expropriation without compensation. As of March this year, bank loans in the form of mortgages to the private sector came to R1.4 trillion; a figure close to 30% of annual GDP. On the assumption that vacant land makes up 2.5% to 7.5% of the R1.4tn total, the private sector would stand to lose between R35bn and R105bn to expropriation.[8] For FirstRand as our test case, this would be a loss of between R4.9bn and R14.7bn.


Currently, the four major banks have healthy balance sheets, and it is possible that they may be able to cushion some of the consequences of expropriation. For example, if FirstRand were to write off their loans to the agricultural sector and for vacant land (the latter based on our assumption of 7.5% of mortgages), their capital adequacy ratio (CAR) would go from 17.1% to 12.2%[9] - which is still above Basel III’s 10.5% minimum. Despite this, banks will lose faith in land as security for loans – whether one wants to borrow funds to acquire property, grow a business, or to see it through to the next harvest. The consequences from the knock-on effects of this to the economy, and all it encompasses, including food security, are currently uncertain.

The biggest immediate financial – as well as fiscal – threat however, lies with the government and the Land Bank. If commercial farmers were to default on their Land Bank loans secured by land, the government would be left with a bill of around at-least R25bn.


Given that South Africa’s largest private financial institutions are currently well funded, based on our analysis it is likely that they would be able to survive the immediate effects of expropriation of farming and vacant land without compensation. The government’s Land Bank however will not survive, and if it were to collapse, Treasury will have to find somewhere another R25bn to fill the gap. But back to the private banks; their funding models, for at least a large portion of their traditional loan book, run the risk of being turned on their head.

Banks may no longer be able to make loans to farmers or investors who wish to use their farm land or vacant property as security. If so, farmers and a large portion of investors will be ham-strung, and South Africans will run the risk of being left with food insecurity and an economy that will face even greater difficulties in freeing itself from the malaise of mass unemployment.

Therefore, it would be wise for government to seriously consider other models for land restitution, such as the private funding model put forward by Wandile Sihlobo, head of agribusiness research at the Agricultural Business Chamber. Sihlobo argues for a model built on the principles highlighted in chapter six of the National Development Plan (NDP), but at the same times provides space for the private sector to take charge of a large part of the process. Another consideration by government should be a plan put forward more recently by the Banking Association of South Africa, in the same vein as Sihlobo’s, to start a joint fund to accelerate the transfer of land.

Charles Collocott is a researcher at the Helen Suzman Foundation. 

This article first appeared as an HSF Brief. 

[1] https://www.fin24.com/Economy/banks-weigh-new-fund-to-boost-black-farm-ownership-20180829

[2]No. 63 of 1975.

[3] https://www.fin24.com/Economy/could-banks-be-the-biggest-losers-in-expropriation-without-compensation-20180301

[4] https://www.fin24.com/Economy/could-banks-be-the-biggest-losers-in-expropriation-without-compensation-20180301


[6] https://www.fin24.com/Economy/could-banks-be-the-biggest-losers-in-expropriation-without-compensation-20180301



[9] FirstRand Annual Integrated Report, 2017, p137.