اثر نقدینگی: اثر عملیات بازار آزاد رزرو فدرال در سال های اخیر

Previous research indicated that the daily liquidity effect, or the change in the federal funds rate associated with an exogenous change in Fed balances, varies with several factors including the day of the maintenance period. In this paper, we examine data from 1998 to 2007, the recent period of increased Federal Reserve transparency before the financial crisis, and find that the liquidity effect stabilized across days of the maintenance period. We conclude that the liquidity effect may be a function of the uncertainty about banks’ end-of-day balances, as well as pure maintenance period effects. Moreover, we find that increased transparency led to a larger liquidity effect on the days prior to an FOMC meeting.

Whether short-term interest rates change in response to a change in the money supply is a perennially debated issue. Identifying this phenomenon, known as the “liquidity effect,” is central to understanding the Federal Reserve’s ability to implement monetary policy in the reserves regime that prevailed prior to late 2008. Previous literature has discussed two types of liquidity effects, long-range and daily. Evidence of the former has been mixed: over monthly and yearly horizons, some researchers have shown that short-term interest rates respond to a change in the money supply (Bernanke and Mihov, 1998), while others do not find such an effect (Leeper and Gordon, 1992). The differences in the estimated liquidity effects stem from different specifications, sample periods, and measures of the money supply, suggesting that if a long-range liquidity effect does exist, it is not necessarily econometrically robust or stable through time. Indeed, more recently, Carpenter and Demiralp (2008) provide evidence of a liquidity effect at a monthly frequency, using a more relevant measure of the money supply – balances held at the Federal Reserve – for such an exercise. Consistent with this more relevant definition of the money supply, evidence of a daily liquidity effect has been found repeatedly and robustly: on a daily frequency, the effective federal funds rate moves lower in response to unexpected increases in the supply of Fed balances. Research by Hamilton (1997), Hamilton (1998), Thornton (2001), and Carpenter and Demiralp (2006a) indicates that this daily liquidity effect varies according to the day of the maintenance period, becoming particularly pronounced on settlement Wednesday, the last day on which banks can satisfy their reserve requirements.

Our results offer further support for a liquidity effect at a daily frequency from 2004 to mid-2007. Our hypothesis is that the magnitude of the liquidity effect depended critically on three factors: uncertainty on the part of DIs about their Federal Reserve account balance positions, expectations for the funds rate in the remainder of the maintenance period, and proximity to settlement day. The falloff in the liquidity effect after the implementation of lagged reserve accounting and the differential effects of high-payment-flow days support our first hypothesis; the FOMC anticipation-liquidity effect supports the second, and the estimated coefficients on all settlement day effects support the third.