Interest rate derivatives and asset-liability management by commercial banks

New England Economic Review, Jan-Feb, 1995 by Katerina Simons

The four "gap" measures are meant to represent a crude measure of the interest rate risk assumed by the bank before its derivative position is taken into account. Larger absolute values of the gap measures indicate a greater sensitivity to interest rate changes on the part of the bank. A bank can reduce its interest rate exposure by hedging with derivative positions.

IV. Results

Table 5 reports regression results for banks with assets below $5 billion, while Table 6 reports the results for the banks with assets of $5 billion or more. The gap measures show no consistent relationship to the use of derivatives. For instance, while a positive and significant relationship exists for large banks between the one- to five-year gap and the use of all derivatives, the relationship between these variables is negative and significant for the small bank subsample.

A somewhat unexpected result of the regressions is the negative relationship between the intensity of derivative use and bank size. Given that large banks use derivatives more frequently than small banks (Table 2), one might have expected the intensity of their use also to be higher.

It has sometimes been suggested that barriers to entry into derivative markets due to economies of scale prevent smaller institutions from participating. However, this argument has more validity for the over-the-counter instruments than for the exchange-traded ones. Over-the-counter derivatives are customized, must be purchased from a dealer, and may have large, indivisible contract denominations. The argument is less compelling for exchange-traded instruments, which are available even to retail investors.

The regression results in Tables 5 and 6 show a negative relationship between the use of derivatives and size for all derivative categories, for both large and small bank subsamples. The only exception is the regression for swaps for banks with under $5 billion in assets, which has a positive and significant coefficient for bank size (Table 5, Column 2). It will be recalled that swaps are over-the-counter instruments, where barriers to entry may indeed be present for smaller banks, while the other regression categories (options, and futures and forwards) include both the exchange-traded and the over-the-counter instruments.

The relationship between the equity-to-asset ratio and the use of derivatives is ambiguous. For both the small and the large bank subsamples, the relationship is negative and statistically significant for futures, but positive and significant for swaps. It is possible that swaps are perceived as more risky and invite greater scrutiny from the regulators in weaker-capitalized banks. This may allow better-capitalized banks to participate in the swap market to a greater extent.

Among large banks, those with weaker asset quality (as measured by a higher ratio of nonperforming assets to assets and a smaller ratio of loan-loss reserve to loans) appear to be bigger users of derivatives than banks with relatively stronger asset quality. In the large bank subsample, the coefficient for non-performing assets is positive and significant, while the coefficient for loan-loss reserves is negative and significant for futures, swaps, and all derivatives.