Evaluation of Banks' Interest Rate Risk: An Alternative Approach

Abstract

Interest rate risk is often assessed through parallel yield curve shifts of 100, 200 or 400 basis points. In order to provide a more realistic view, we did simulations based on periods of growing interest rates that actually occurred in the past. These simulations show that non-bank deposits and non-bank loans react more strongly to rising interest rates than certain interbank and security positions. Existing research usually overestimates related risks slightly as it does not take the interest-elastic reactions of non-banks into account. We found three types of effects. Firstly, the direct earnings effect stems from changed market interest rates applied to constant balance sheet positions. This effect is typically measured by straightforward models. Secondly, to increase accuracy, we identified an indirect earnings effect. Customers react to interest rate changes, and therefore balance sheet positions increase or decrease. The size of this effect depends on how strongly they react, i. e. their interest elasticity. Thirdly, the induced earnings effect results from a bank’s reactions in an attempt to compensate for the changed business volume.