Home improvement loan interest rate

Interest rate risk at commercial banks is the risk that changes in interest rates will adversely affect income or capital. Such risk is an inherent part of banking because banks typically originate loans with longer maturities than the deposits they accept. This maturity mismatch between loans and deposits causes the net interest margin (NIM)--the spread between loan rates and deposit rates--to fall when interest rates rise, because interest rates on deposits adjust more quickly than interest rates on loans. Further, when interest rates rise, the economic value of longer-term instruments (assets) falls by more than the economic value of shorter-term instruments (liabilities), thus reducing the bank's capital.

Bankers became increasingly concerned about interest rate risk following the savings and loan (S&L) crisis. In the early 1980s, many thrifts became insolvent after interest rates rose sharply, setting off a crisis that eventually required a $150 billion taxpayer bailout (Curry and Shibut, 2000). Thrifts were particularly vulnerable to interest rate risk because of the large maturity mismatch that resulted from using short-term deposits to fund long-term home loans. Nevertheless, banks devoted considerable resources to measuring and managing their exposure to interest rate risk. Many regional and money-center banks implemented elaborate models to measure their exposure and began to use sophisticated asset and liability management to manage their risk.

Bank supervisors also were challenged to stay abreast of the industry's ability to take on interest rate risk, and they responded with three related initiatives. First, bank examiners received capital markets training to help them understand better the techniques for measuring and managing interest rate risk. Second, bank supervisors explicitly incorporated interest rate risk into their ratings system in 1997, transforming the "CAMEL" rating system into "CAMELS." (1) The "S" rating stands for a bank's sensitivity to market risk, which includes interest rate risk and exposure to trading account assets, exchange rates, and commodity prices. (2) The third supervisory initiative was to develop a measure of interest rate risk that examiners could use to risk-scope a bank--that is, to pinpoint the areas of the bank that warrant closer scrutiny--and to conduct off-site surveillance. Economists at the Board of Governors of the Federal Reserve System developed a proprietary economic value of equity model called the economic value model (EVM), which is a duration-based estimate of interest rate sensitivity for each U.S. commercial bank (Houpt and Embersit, 1991; Wright and Houpt, 1996). The Federal Reserve operationalized the model in the first quarter of 1998 by producing a quarterly report (called the Focus Report) for each bank. The Focus reports are the confidential supervisory reports that provide the detailed output of the Fed's EVM.

The EVM's interest rate sensitivity assessment is most relevant for community banks, which we define as those with less than $1 billion in assets and no interest rate derivatives. Larger banks often have derivatives or other balance-sheet complexities that the EVM ignores, making the output from the EVM more questionable. The EVM also is more appropriately applied to community banks because community banks are examined less often than larger banks and the EVM is usually the only tool for off-site interest rate risk assessment available to examiners for those banks. Community banks devote fewer resources to modeling and measuring their interest rate risk than do regional and money-center banks, which normally have full-time staff devoted to such tasks. Consequently, examiners of larger institutions usually have access to more sophisticated and often more timely information than that provided by the EVM.

This paper investigates the effectiveness of the EVM by examining whether model estimates are correlated with community bank measures of interest rate sensitivity during recent periods of both rising and falling interest rates. Because the model is relatively new, it has yet to be validated against actual bank performance. The Federal Open Market Committee (FOMC) increased the federal funds rate six times in 1999 and 2000, and then lowered the federal funds rate 12 times in 2001 and 2002. A strong correlation between the EVM's estimate of interest rate sensitivity and measures of interest rate risk during these periods would suggest that the model provides a useful surveillance tool to community bank supervisors.

We find that estimates from the Fed's EVM are correlated with the performance of U.S. community banks in the manner the EVM suggests. Specifically, the banks that the EVM identifies as the most liability sensitive--those most sensitive to rising rates--show the biggest deterioration in performance during the period of rising interest rates between 1998 and 2000. The most liability-sensitive banks also show the greatest improvement in performance measures during the 2000-02 period of falling rates. The evidence indicates, then, that the EVM is a useful tool for supervisors interested in identifying the minority of banks that are highly sensitive to interest rate changes.

RELATED LITERATURE

Researchers have examined the interest rate sensitivity of depository institutions in some detail. There are two general lines of inquiry. The first line of inquiry asks whether depository institutions are exposed to interest rate changes and, if so, how large is that exposure on average? The motivation for this research often is to assess the impact that monetary policy or unexpected inflation might have on financial intermediation. Most studies measure interest rate sensitivity by regressing the firm's stock return on a market index and an interest rate. Flannery and James (1984), Aharony, Saunders, and Swary (1986), Saunders and Yourougou (1990), Yourougou (1990), and Robinson (1995) find that bank stock prices react to (unexpected) interest rate changes. A major limitation of this research is that the vast majority of U.S. banks are excluded from the analysis because they have no publicly traded equity. Flannery (1981, 1983) constructs a model that estimates the effect of rate changes on a bank's net operating income. The model has the added advantage that it indirectly estimates the maturities of the assets and liabilities. Flannery finds that the impact of rate changes on long-run bank earnings is small, averaging only 5.6 percent of net operating earnings. He also finds that banks are slightly asset sensitive; that is, profits increase with rising interest rates. These results, however, contradict much of the literature--including some of Flannery's later work--which shows that banks tend to be exposed to rising rates.

A second line of inquiry attempts to isolate a bank-specific measure of interest rate risk to separate banks by their interest rate sensitivity. Regulators are interested in this process because bank-specific measures provide opportunities to identify high-risk banks. Flannery and James (1984) construct a one-year gap measure and quantify the correlation between this measure and the portion of a bank's stock return driven by interest rate changes. They find that this simple maturity variable has statistically significant explanatory power. Gilkeson, Hudgins, and Ruff (1997) use output from a regulatory gap model for thrifts between 1984 and 1988. They also find a statistically significant correlation between net interest income and the one-year gap measure. Robinson and Klemme (1996) find that bank holding companies with relatively high levels of mortgage activity have higher degrees of interest rate sensitivity than other bank holding companies, as reflected by changes in stock prices. Finally, Lumpkin and O'Brien (1997) construct a comprehensive measure of thrifts' portfolio revaluations caused by interest rate changes. They fail to find evidence that such revaluations influence stock returns beyond the influence already captured by more general movements in interest rates.

This article adds to the evidence that banks are liability sensitive, though the interest rate sensitivity, on average, is small. Our results are consistent with Gilkeson et al. (1997) and show that even accounting-based measures of interest rate sensitivity can have significant explanatory power to aid bank supervisors in risk-scoping and monitoring the interest rate exposure of commercial banks. Our results imply, as well, that large rate increases are unlikely to have significant adverse effects on the banking industry, which is also consistent with previous literature.

A MEASURE OF RATE SENSITIVITY: THE EVM