Net interest income is the difference between revenues from interest-bearing assets and costs of servicing liabilities. Bank assets typically consist of commercial and personal loans, mortgages, investment securities, and construction loans. Liabilities, on the other hand, are the customer deposits. The ratio of assets to liabilities is referred to as the “asset-to-liability ratio.”
Basis model overpredicts net interest income
The results of the base model are summarized in Table 3 in the first (1) column. For example, a 100-basis-point increase in the yield curve results in an estimated 14.5-basis-point increase in net interest income within two years. Likewise, a change in any one factor causes a cumulative effect on net interest income of 12 basis points in two years. Nonetheless, the results of the base model are not as strong as those of the FCB model.
The basis model overpredicts net interest income by overestimating the amount of change that will occur if interest rates rise or fall. As such, it is necessary to take into account the interaction between interest rates and balance sheet items to understand the reasons for the difference in predictions. This is where the basis model can prove to be particularly useful. This method requires a number of assumptions, including the existence of balance sheet items and the composition of the financial statement.
The quality of a bank’s loan portfolio can significantly affect net interest income. Loans are riskier than marketable securities, so they carry a higher interest rate. On the other hand, non-maturity deposits carry the lowest interest expense, and they pay lower rates than comparable deposits. If the loan-to-asset ratio increases, net interest income should rise. So, if the economy is improving, the bank should expect to earn more than $50 million in net interest income.
RMSE of FCB model is 8 basis points
The FCB model can be compared with the base model in terms of its performance. The base model predicts net interest income more accurately than the FCB model. The RMSE of the base model is 8 basis points, whereas the FCB model’s RMSE is 18 basis points. The RMSE, or root-mean-square error, is a summary measure of the predictive errors.
The FCB model underpredicts community bank net interest income over the postcrisis period. Compared to the first half of 2007, the model predicts a decline of 62 basis points. The actual decline was 15 basis points higher. However, the FCB model predicts net interest income at a rate close to 40-year low. In other words, the model is only eight basis points more accurate than the base model.
Effects of changes in the slope of the yield curve on net interest income
The effect of short-term rate changes on bank net interest income is dependent on the sensitivity of the bank’s assets and liabilities. In general, higher long-term rates lead to higher net interest income. However, the effect is not the same for banks with different asset-liability sensitivity. For example, higher short-term rates lead to higher net interest income for asset-sensitive banks.
Banks publish data on the sensitivity of their net interest income to interest rate changes. These changes are measured in terms of parallel, steep, and flattening shifts on the yield curve. The Big Four U.S. bank holding companies all reported positive changes in net interest income in 2014. These results assume that banks are earning higher interest on their lending than on deposits. Thus, when the slope of the yield curve increases, the banks’ net interest income should increase.
The base model predicts higher net interest income in the first half of 2017, or 38 basis points above current levels. However, this level is only marginally above the average. As the economy improves, it is likely to rebound. So, the study focuses on the effect of steeper yield curves on net interest income. This is likely the income component most sensitive to interest rate changes. So, higher yields do not necessarily mean a return to the high levels of the early 1990s.
