حساب های گردشی، مدیریت ذخیره و نوسانات 1 نرخ بهره

Retail sweep accounts have reduced required bank reserve balances by more than 70% since 1995, raising concerns in some quarters about increased volatility of interest rates and reduced monetary policy effectiveness. We develop a model of bank reserve management and daily payment flows that indicates that the effect of lower reserve balances on funds rate volatility is theoretically ambiguous. We empirically test the relationships among reserve balances, federal funds-rate volatility, and the variation in short-term money market rates. Our conclusion is that reductions in reserve balances have not impinged on the ability of the Federal Reserve to conduct monetary policy in the manner that it has in recent years.

Retail sweep accounts were introduced in 1993 and have led to a 70% reduction in required bank reserve balances with the Federal Reserve since 1995. While this has led to a 70% reduction in the reserve requirement tax, it also has complicated reserve management by banks and interest rate targeting by the Federal Reserve. During this process, the Federal Reserve has suggested that the reduction in reserve balances can lead to a rise in Federal funds rate volatility and hinder the implementation of monetary policy. Sweep accounts shift funds from retail transactions accounts subject to reserve requirements to interest- or noninterest bearing savings accounts exempt from reserve requirements.1 Sweep balances at banks increased from about $10 billion in January 1995 to $382 billion in February 2000. During this period, required reserve balances held in accounts at Federal Reserve banks fell from around $22 billion to $6 billion. This reserve reduction continues a trend initiated earlier in 1990 and 1992 when reserve requirements were reduced and reserve balances, which were then on the order of $33 billion to $38 billion, started their recent decline.2 At present, required reserves are around $6 billion while clearing balances (which earn a return and are used to pay for priced payment services) are $7 billion. Although the sum of these balances ($13 billion) can be used to fund daily bank payment flows over Fedwire, these flows total $1.8 trillion and exceed reserve balances by a factor of over 100. The uncertain nature of daily payments means that unexpected payment flows can significantly affect intraday and overnight bank borrowing, potentially making it difficult for the Federal Reserve to closely target the funds rate. Concerns about potential adverse monetary policy effects of reduced bank reserve balances are not new with sweep accounts. Indeed, after the 1990 reduction in reserve requirements, there was a considerable increase in the volatility of the federal funds rate. In documenting the volatility rise, Feinman (1993) took it one step further and suggested that the elimination of reserve requirements would “engender a significant increase in volatility in the reserves market and seriously complicate the conduct of open market operations.” More recently, in a letter to Congress, Alan Greenspan (1996) stressed this same argument with respect to sweep accounts: “retail sweep” programs, which have further blurred the distinction between transaction and savings deposits,… and the more widespread adoption of sweep programs that is evidently in train could impair the predictability of overall reserve demand and hence adversely affect the ability of the Federal Reserve to gauge the supply of reserves consistent with the FOMC’s intended policy stance. Reinforcing this view, the Federal Reserve announced that sweep accounts “… could eventually suggest changes in the structure of reserve requirements” (American Banker, 1997), raising them, since the reduction in reserve balances “… could impair the predictability of reserve demand and hence adversely affect the ability of the Federal Reserve to gauge the supply of reserves consistent with its intended monetary policy stance” (Wall Street Journal, 1997). This concern has led the Federal Reserve to support legislation which would permit the payment of interest on reserves, if needed to reverse the current erosion of the reserve base (Greenspan, 1998), and also to propose that banks return to a system of lagged reserve accounting to improve the predictability of reserve demand (Wall Street Journal, 1998). However, improving the predictability of reserve demand does nothing to reduce uncertainty in daily payment flows. Are the Federal Reserve’s concerns about the monetary policy implications of sweep accounts and resulting reductions in reserve balances well founded? We develop a theoretical model of optimal bank reserve management to investigate how lower reserve balances, combined with uncertain daily bank payment flows, affects the determination and volatility of 24-hr and overnight funds rates. Since we find that the effects are theoretically ambiguous, we proceed to test empirically the relationships suggested by the Federal Reserve that surround the effective implementation of monetary policy. The model we develop is outlined in the following section. It concerns bank intra- and interday behavior where reserve managers must optimally adjust reserves and borrowings in light of payment-clearing disturbances. Analysis of the model indicates that there are opposing effects of lower reserve requirements on daily funds-rate volatility. As previously reasoned by Sellon and Weiner (1996), Bennett and Hilton (1997), and Clouse and Elmendorf (1998), we conclude that lower reserve requirements reduce the sensitivity of the demand for reserves and funds borrowings to variations in funds rates, which contributes to greater funds-rate volatility. However, we also find that reductions in reserve requirements effectively permit banks to apply a larger portion of any given volume of reserves toward covering unexpected payment-clearing shocks, though at positive opportunity cost. Thus, even though lower effective reserve requirements induce banks to reduce their reserve balances, reserve-requirement reductions can make reserve and overnight funds demands less variable, lowering the variance of 24-hr and overnight funds rates. As the effects from lower reserve balances need not all be in the same direction, in section 3 we evaluate empirically the channels by which lower reserve holdings may hinder the effectiveness of monetary policy. This is a two-part process requiring, first, that a decrease in reserve holdings significantly increases the deviation of the federal funds rate from its target and, second, that this greater funds-rate deviation is associated with greater volatility in the short-term money market rates that anchor the low end of the yield curve. There is some evidence that supports both hypotheses in the time period prior to when the Federal Reserve started to publicly announce the funds rate target. However, there is no such support for the time period after the target was announced, even though this is the period when sweep accounts lowered reserve balances to their current historically low levels. Therefore, we conclude in section 4 that sweep accounts and other regulatory or institutional innovations resulting in lower, or perhaps even zero, required reserve balances are unlikely to pose significant complications for the conduct of monetary policy.

Since 1993, the rapid growth of retail sweep accounts in the U.S. has succeeded in reducing required bank reserve balances, and the reserve requirement tax, by over 50%. We develop a theoretical model for examining how lower reserve balances, combined with uncertain daily payment flows over Fedwire, are likely to influence intra- and interday funds rates and possibly adversely affect the Federal Reserve’s ability to closely target the federal funds rate. Daily payments over Fedwire exceed the value of reserve balances by a factor of more than a hundred. Thus even small unexpected changes in payment flows (deposit shocks) can lead to unexpected bank intraday and overnight borrowing in order to achieve bank targets for end-of-day reserve positions. In this uncertain environment, Federal Reserve open market operations can be imperfect and result in increased volatility of the funds rate from its targeted level. If funds-rate volatility rises, and if this volatility increase is transmitted to the rates on short-term money market instruments that anchor the low end of the yield curve, then implementation of monetary policy can be degraded or made more difficult. Our model of bank reserve requirements, uncertain daily payment flows, and intra- and interday funds rate determination suggests that the effect of lower reserve balances on funds rate volatility is theoretically ambiguous. As previously noted by others, lower reserve balances reduce the sensitivity of the demand for reserves and funds borrowings to variations in funds rates, which contributes to greater funds-rate volatility. However, we also find that reductions in required reserves effectively permits banks to apply a larger portion of any given volume of reserves toward covering unexpected payment-clearing shocks, though at positive opportunity cost. While lower effective reserve requirements induce banks to reduce their reserve balances, reserve-requirement reductions can also make reserve and overnight funds demands less variable, lowering rather than raising the variance of 24-hr and overnight funds rates. This ambiguity leads us to test empirically the relationship between low reserve balances, federal funds-rate volatility, and the volatility of short-term money market rates. For the 1991- (January)1994 period preceding a policy of publicly announcing the funds rate target, lower reserve balances were significantly associated with higher funds rate volatility and with higher volatility of money market rates. For the (March)1994-1996 period after the announcement policy was implemented, this significance was lost. We conclude, therefore, that current reductions in effective reserve requirements caused by sweep accounts do not currently adversely affect the Federal Reserve’s ability to implement monetary policy. Indeed, because the funds-rate target is now known to market participants, increases in funds-rate volatility can be confidently discounted as noise rather than possibly carrying information about changes in monetary policy, which appears (in the very short-term) to have been the case before the Federal Reserve’s public announcements of its funds-rate targets. Elmedorf 1997, Federal Reserve Bank of New York Annual Report 1995, Freeman and Haslag 1995, Kanatas and Greenbaum 1982, Kasman 1992, Meulendyke et al 1988, Thornton 1996, VanHoose 1986, VanHoose et al 1999, Wall Street Journal 1998, Wall Street Journal 1997 and Weiner 1985