No abstract

Because of recent structural changes in the balance sheets of banks, regulatory changes in the risk-based capital requirements, and the recent adoption of mark-to-market accounting changes, interest rate risk remains an important issue for commercial banks and an important regulatory concern. Market, interest rate, and foreign exchange risk are estimated for a sample of commercial banks using ordinary least squares from 1986 to 1991. Consistent with earlier studies, the estimated coefficients continue to be unstable. We find that interest rate risk decreases and foreign exchange risk increases. Moreover, the results differ depending on practices of the bank (money center, superregional, or regional). We find evidence consistent with earlier studies that theorize foreign exchange risk is explained by unhedged foreign loan exposure.

I. INTRODUCTIONIN THE EARLY 1950's the Federal Reserve System pursued a "bills only" policy. That is, the System would conduct open market operations.only through the bill market with the effect on interest rates presumably being transmitted gradually to other sectors of the financial system.' This policy generated substantial research efforts. The thrust of subsequent theoretical and empirical arguments has evolved around the manner in which interest rate changes are transmitted from the bill market throughout the financial system. Since the mid-1960's, theorists and economic policy makers have been faced with a bewildering and conflicting array of empirical studies of interest rate relationships and behavior. Since the results of these studies are frequently incorporated into macro-economic models for forecasting and policy choice, it is important that the conflicts in the literature be thoroughly understood if not completely resolved.The purpose of this paper is to present the results of an analysis of the relationship between various interest rates. Specifically, the nature of the lead-lag relationship is examined as is the stability of the relationship over time. The generally accepted view is that long-term rates lag behind short-term rates. This view has, however, come under strong attack in recent years most notably by Cargill and Meyer [2], Pippenger [13], and Phillips and Pippenger [12]. The analysis herein, although employing a different analytical technique, provides strong additional support for these studies and is in sharp contrast to the earlier work of Modigliani and Sutch [8], [9]. The evidence indicates that there is no consistent and meaningful lead-lag relationship between long-term and short-term interest rates and that the bond market is efficient in that knowledge of past interest rates provides virtually no information regarding current rates. Furthermore, the analysis reveals that some of the statistically significant relationships that appear to be present are complex and unstable over time.In the following Section, we provide a brief overview of the conflicting literature. In Section III, we describe the analytical technique and its rationale. In Section IV, we present the results of the analysis and this is followed by a brief summary. II. THE CONTROVERSY One of the seminal efforts to determine the nature of interest rate relationships was the Preferred Habitat model developed by Modigliani and Sutch [8], [9], and later extended by Modigliani and Shiller [11]. Modigliani and Sutch (hereafter M & S), * Assistant Professor, Michigan State University, and Mary Rennebohm Professor of Business, University of Wisconsin-Madison, respectively. 1. For a discussion of the rationale underlying the "bills-only" policy, see [14] and [4]. 93 94 The Journal of Finance concluded that long-term rates depend on a 16-quarter distributed lag on shortterm rates with the weights having an inverted "U" shape. The work of M & S provided theoretical and empirical support for the Federal Reserve policy of conf...

No abstract

The optimal gap of a depository institution is derived using a market value optimization model. The gap is estimated using portfolios of returns on rate-sensitive assets and liabilities and is found to be not significantly different from zero. The estimate is compared to the average gap position of a sample of banks. It is found that the average gap position of a sample of banks is "too positive." This suggests that banks are not showing risk minimization behavior in the positioning of the gap.