Estimating the Interest Rate Risk of Banks Using Time Series of Accounting-Based Data

Abstract

This paper proposes a new method of estimating the interest rate risk of banks from the perspective of bank outsiders. The key innovation is the inclusion of time series of accounting-based data instead of using only the latest available reports to estimate the maturity structure of banks. Using regulatory accounting-based data, we estimate the model for more than 1,000 German universal banks and compare the results with a unique data set of bank-internal quantified interest rate risk. We find evidence that our model yields a significantly better fit of banks’ internally quantified interest rate risk than standard approaches that rely on one-point-in-time data.