A common expectation among academics and financial market analysts is that a steeper yield curve raises bank net interest margins. This is because bank liabilities (e.g. deposit funding) are usually priced-off of short-term interest rates and assumed to adjust relatively quickly to rate changes, while bank assets (e.g. mortgages) tend to be priced-off of longer-term rates and assumed to adjust relatively slowly to rate changes. Net interest margins (NIMs) for banks would also be affected by the level of interest rates – higher rates for an unchanged yield curve slope is often assumed to reduce NIMs if yields on liabilities rise faster than on assets.
Yet in South Africa, the correlation (both contemporaneous and lagged) between the banking sector profitability and the slope of the yield curve has been low since 2009. One interpretation is that banks have been successful at limiting the exposure of their profitability to market rates through the management of their balance sheets. We discuss the pricing of assets and liabilities in South Africa’s banking sector in Diesel et al (2022). Both assets and liabilities are priced based on short-term rates, but different premia are added to incorporate liquidity risk (and credit risk and regulatory charges in the case of assets), operating costs and commercial margins/discounts. We show in that work that funding costs (for liabilities) have not followed the policy rate cycle, while the funding price transfer rates applied to assets have been slightly higher on average over recent years than those applying to liabilities. This likely explains a part of the low correlation between NIMs and the yield curve slope.
NIMs are also affected by regulatory changes (such as Basel III) that affect bank balance sheet composition and interest-rate pass-through to loans or deposits (Greenwood-Nimmo et al 2022). In Diesel et al 2022 we suggest that Basel III’s stable funding regulations served to increase banks’ funding costs by increasing the duration of banks’ funding liabilities and the relative cost of deposit funding, potentially putting pressure on bank NIMs. On the other hand, we argue that Basel III’s liquidity requirements created incentives for banks to hold higher levels of government securities that have provided relatively attractive returns recently. There are other factors that also affect bank NIMs. It is possible that banks have increased their margins in response to higher non-performing loans or that the high level of concentration in the sector has buoyed profitability. Lastly, the aggregate data in the figure mask important heterogeneity across banks, so a bank-level examination of NIM drivers would be important to understand how individual bank’s have managed their balance sheets and interest rate risk.