Interest rate derivatives and asset-liability management by commercial banks

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

Bank participation in derivative markets has risen sharply in recent years. The total amount of interest rate, currency, commodity, and equity contracts at U.S. commercial and savings banks soared from $6.8 trillion in 1990 to $11.9 trillion in 1993, an increase of 75 percent. A major concern facing policymakers and bank regulators today is the possibility that the rising use of derivatives has increased the riskiness of individual banks and of the banking system as a whole.

Banks have long used one type of derivative instrument, namely interest rate futures, to manage interest rate risk (Koppenhaver 1986; Booth, Smith, and Stoltz 1984; Parkinson and Spindt 1986; Franckle and Senchack 1982). However, the development of newer instruments, such as swaps, caps, collars, and floors (see the glossary in Box 1), has greatly expanded the menu of financial technologies available to banks for asset-liability management. In particular, interest rate swaps have become the preferred tool. According to a recent market survey of derivative users, 92 percent of responding financial institutions report using interest rate swaps to manage the interest rate risk of their lending portfolios (Group of Thirty 1993, pp. 40-41).

More recently, studies have focused on the determinants of the broader derivative activities of banks. Sinkey and Carter (1994) studied the determinants of bank use of derivatives between 1989 and 1991. They found that measures of maturity gap and liquidity are consistently significant across banks of different asset sizes. Brewer, Minton and Moser (1994) focused on the relationship between derivative use and bank lending, concluding that the growth of business lending is positively related to the presence of swaps on the bank's books, though the presence of futures had no significant effect on bank lending.

This article contributes to the growing literature on bank derivative use by analyzing the determinants of banks' use of interest rate derivatives between 1988 and 1993. It begins by explaining the use of gap models to measure interest rate risk and the way interest rate derivatives can be used to manage that risk. The article goes on to describe bank use of various interest rate derivatives in more detail and trace growth in recent years. Then it outlines the empirical specification of a model of derivative use and describes the data set used in the analysis. The subsequent section presents the estimation results and interprets parameter estimates. The article concludes by drawing policy implications from the analysis.

I. Interest Rate Risk and Gap Analysis

Banks use derivative products mainly to manage interest rate risk. The last 15 years have seen an increased volatility of interest rates compared to the earlier post-World War II era, making the need for accurate measurement and control of interest rate risk particularly acute. At the same time, financial innovations in the field of interest rate derivatives have given banks new and effective instruments for managing that risk.

Interest rate risk arises in bank operations because banks' assets and liabilities generally have their interest rates reset at different times. This leaves net interest income (interest earned on assets less interest paid on liabilities) vulnerable to changes in market interest rates. The magnitude of interest rate risk depends on the degree of mismatch between the times when asset and liability interest rates are reset.

One way to measure the direction and extent of the asset-liability mismatch is through gap analysis, which derives its name from the dollar gap that is the difference between the dollar amounts of rate-sensitive assets and rate-sensitive liabilities. A maturity gap is calculated for a given time period and includes all fixed-rate assets and liabilities that mature in that period and all floating-rate assets and liabilities that have interest rate reset dates in that period.

A bank that has a positive gap will see its interest income rise if market interest rates rise, since more assets than liabilities will exhibit an increase in the interest rate. Similarly, a bank with a negative gap will be hurt by rising rates but will benefit from falling rates.

For example, a bank that issues a 3-month certificate of deposit, and uses the funds to buy a 2-year Treasury note, will see its net interest income eroded if interest rates rise after the first three months because it will have to roll over the CDs at a higher rate, while the rate on the Treasury note will remain the same. In general,

[Delta]NII = (A - L) x [Delta]r, (1)

where NII is net interest income, A is rate-sensitive assets, L is rate-sensitive liabilities and r is the market interest rate. The problem with this simple gap measure is that unless the time interval chosen is very small, assets and liabilities will have their rates reset at different times within that interval. In an extreme case, if the chosen interval is three months, a bank that issues 3-month CDs and funds them by borrowing federal funds overnight will show a three-month gap of zero, even though that bank is exposed to a substantial interest rate risk.