Risk in banking ends up in only three places: shareholders, depositors (including lenders) or government. If a bank loses money it is one of those three categories that takes the hit.

Ultimately, bank capital rules are simply about nudging the risk between those categories. And a lobby is growing to shift the risk away from shareholders to enable more infrastructure lending.

The Basel capital accords, a globally determined set of regulations about how banks have to put up shareholder money to protect against risks, have taken public criticism recently from Standard Bank CEO Sim Tshabalala and Absa chief strategy and sustainability officer Punki Modise. While their critiques are subtle, the underlying point is that Basel is nudging risks too far onto shareholder shoulders for socially useful bank activities such as financing infrastructure.

Since the global financial crisis the trend has been to shove risk more towards shareholders and away from governments. That is particularly the case in the rest of the world — in SA, capital rules were always more stringent, and shareholders carried more load — one reason SA was relatively unscathed by the crisis.

Basel affects the allocation of risk through two mechanisms: the ratio of capital to assets banks must maintain, but also the calculation of risk-weighted assets (RWAs) banks must hold capital against. Capital is shareholder money that is the first to be lost if loans go bad. Each type of loan gets “weighted” to determine its riskiness. So, loans to risky corporates that do not have a credit grade rating are weighted at up to 150%.

Project finance is also generally weighted at 150% (though this can vary depending on the credit rating and phase of construction of the project). That can be compared to low-risk categories such as residential mortgages (generally weighted at 75% in SA, depending on the loan-to-value ratios) and the ultimate risk-free asset, government bonds, weighted at 0%. You can see quickly that you can encourage banks to shift their lending focus by how those loans are weighted.

The capital requirement is expressed as a percentage of these RWAs. In SA, various layers of capital requirements add up to 8.5%, though the SA Reserve Bank can increase this based on its supervisory reviews of capital requirements, and most banks hold about 12%. Banks have target returns on capital, so they demand that loan interest meets those return requirements.

That drives the pricing of loans — so, all else being equal, a loan to an unrated company or project will be more expensive than a mortgage loan, and certainly far more than parking cash in government bonds. That translates into volumes — if less capital is required, banks can lend more too.

Banks face considerable pressure to increase lending into infrastructure and project finance. That is clear across Africa, so it is no surprise that the voices on this have come from Standard Bank and Absa, which have the largest African franchises. Energy projects, roads, ports and rail all need to be financed, but banks must allocate considerable capital to do it.

Shifting the risk can be done in several ways. Basel simply lowers the risk weighting of infrastructure loans, but governments can act directly by taking the risk onto their own balance sheets. Given that sovereign debt has a zero risk weighting, it is the most attractive from a bank finance point of view. If governments guarantee infrastructure finance the risk weighting can be cut to zero, making it far cheaper for banks to lend.

But there is also an argument about the pricing of risk, one that Tshabalala, as chair of the B20 task force on finance and infrastructure, has been keen to prioritise. Is it rational that a loan for a renewable energy project should have a 150% risk weighting but a portfolio of mortgages should be at 75%? That implies that a renewable project has twice the chance of defaulting than a portfolio of mortgages.

With enough diversification across projects, that should not be the case. Renewable projects, particularly if their revenue lines are secured upfront through long-term contracts, are low risk. It is highly unlikely that customers will stop paying for energy. The Global Infrastructure Hub has estimated that historical infrastructure credit performance means capital charges should be 60%-70% lower.

The argument so far has been focused on Basel — in effect a call to shift the risk weighting of such loans. But regarding SA you can skin this cat in other ways. The use of the sovereign balance sheet to reduce bank risk and therefore capital costs is the most obvious way. Our regulator could also try to exercise some discretion in allowing banks to vary from Basel requirements. For example, the EU created the “infrastructure supporting factor” in 2019 that reduced capital requirements by 25% for infrastructure loans. But a national-level intervention of this sort may be complicated and risk SA’s status as a fully Basel compliant territory. Instead, smarter use of guarantees can solve the problem more directly.

Such a solution is already on the way. Last week the Treasury confirmed it would fund 20% of a $500m credit guarantee vehicle (CGV) that is being supported by the World Bank and its group companies, and will derisk public infrastructure projects. This should be able to secure a strong credit rating, drawing on the credit status of the World Bank and partners, and in turn enable projects to come in at much lower credit risk.

This provides for an optimal reshuffling of the risks, with some of it ending up on the shoulders of global multilateral development institutions. It should enable more infrastructure lending domestically. The CGV has an eye on the independent grid transmission projects that are soon to hit the market and could catalyse bank finance.

One way or another, banks will see the risk equations shift to enable more lending.

• Dr Stuart Theobald is chair of research-led consultancy Krutham. This article first appeared in Business Day.