اطلاعیه ها و اثربخشی سیاست پولی: نمایی از نرخ بهره آمریکا

Until 1994, the US prime rate was said to be sticky because of its irresponsiveness to short-term interest rates. After the Fed started the practice of announcing its intended funds rate in 1994, however, the prime rate has come to react immediately to shifts in the target rate. This paper attempts to explain how the Fed’s policy announcements changed the behavior of the prime rate by using a simple menu cost model. It shows that an increase in the expected duration of funds rate targets was essential to the improvement in the target rate pass-through.

In the 80s and the early 90s, a lot of attention was paid to the prime rate stickiness in the US. At that time, the US prime rate was considered to be determined based upon various market interest rates such as the Federal funds rate, CD rates, and T-bill rates. Many empirical studies were conducted in an attempt to explain the source of sluggishness in the response of the prime rate to those market interest rates (e.g., Goldberg, 1982, Forbes and Mayne, 1989 and Mester and Saunders, 1995). Since 1994, however, adjustments of the prime rate have been synchronizing with shifts in the Federal funds target. In response to a shift in the target rate, the prime is moved in the same amount within a few days of the corresponding FOMC. This implies that policy effectiveness has greatly improved since the prime rate is used as a base rate in many of the loan contracts. The empirical models of the sticky prime rate proposed by the early studies cannot account for such one-to-one correspondence between the prime rate and the target rate. Nevertheless, to the best of my knowledge, no formal explanation was given for this phenomenon.1 This paper attempts to explain the reason why the prime rate has become “flexible”. To this end, it should be kept in mind that there were some noticeable changes in the Fed’s practice around 1994. The most obvious institutional change is the start of policy announcements, which was first made at the February 1994 meeting. Before that meeting, the FOMC did not used to disclose to the public whether the intended federal funds rate was changed or not, and because of this secrecy, the FOMC’s intention was sometimes misperceived by market participants. The FOMC started announcing the change in the intended funds rate and its rationale after a meeting at which policy actions were implemented. Recently, many studies have investigated the influences of such an institutional change by closely looking at the behavior of market interest rates, such as the T-bill rates and futures rates. For instance, Lange et al., 2003, Poole and Rasche, 2003 and Swanson, 2006 argue that the predictability of future policy shifts that can be computed from the Federal funds futures or the euro dollar options has significantly improved since February 1994. Demiralp and Jordà (2004) also provide statistical evidence that there was a structural break in February 1994 in the response of T-bill rates to the Federal funds target.2 Given the results of these studies, it is natural to infer that the structural change in the behavior of the prime rate that occurred in 1994 bears some relation to the start of the Fed’s practice of announcing its target. However, there are several other aspects that should be taken into account aside from the beginning of policy announcements. First, most of the policy shifts before 1994 were decided outside the regularly scheduled FOMC meetings. In fact, prior to 1994, only about 30% of all the policy shifts were made within 7 days of the last scheduled meeting. According to Thornton (2004a), only 27 out of 94 policy changes were made at the regularly scheduled FOMC meetings in the pre-94 period. Second, the volatility of the spread between the effective funds rate and the target rate has been significantly reduced since 1994. Although there is some debate as to whether this phenomenon is due to an advancement in the Fed’s controllability (“open market operation”) or to an “announcement effect” (“open mouth operation”), it should be taken into account the fact that the volatility of the spread has been largely reduced.3 Third, the average duration of a newly changed target has been considerably increased since 1994. The average number of weeks between policy shifts was 5.8 in the pre-94 period and 13.3 in the post-94 period, as of March 2007. The main findings are as follows: first, the response of the prime rate to the funds rate over the entire sample period can be well captured by a simple menu cost model once the abovementioned differences in the Fed’s practice are taken into account. Second, according to the stochastic simulations, neither secrecy in the numerical funds rate target nor uncertainty in the timing of policy shifts was a major cause of the prime rate stickiness before 1994. Third, the volatility of the effective funds rate had a non-negligible effect on the response of the prime rate. That is, the greater the volatility of the effective rate, the less frequently commercial banks would react to shifts in the target rate. Finally, an increase in the average duration of targets seemed to have the largest influence on the improvement in the response of the prime rate. Intuitively, if the target rate after the current policy shift is expected to be kept unchanged for a sufficiently long time, then the prime rate will immediately follow the current policy shift. In contrast, if the next policy shift is expected to be carried out in the immediate future, then commercial banks will tend to hold back from reacting to the current policy shift and wait for the next policy shift.

This paper investigates the source of the 1994 structural break in the relationship between the Federal funds rate and the prime rate. To this end, I construct two baseline models: the pre-94 model and the post-94 model. Stochastic simulations are conducted in order to examine the extent to which each of the period-specific features is responsible for the behavior of the prime rate. The simulation results suggest that the average duration of targets matters most to the response of the prime rate. This is because if the target rate is expected to be readjusted in the near future, commercial banks are inclined to hold back from adjusting the prime rate until the upcoming policy change. Therefore, in order to have commercial banks respond promptly to policy changes, the expected duration of targets has to be sufficiently long. The volatility of the effective rate has the second largest influence on the behavior of the prime rate. Over the past several years, a lot of studies have been made on the issue of how central banks should communicate with the markets (e.g., Woodford, 2005, Kleimeier and Sander, 2006 and Walsh, 2008). However, there are only a few studies that address the issue of how often central banks should change their policy rates. For example, Fukuda (2007) argues that recent central banks change their policy rates only infrequently since a shift in the policy rate itself would create further uncertainty, while Mirzoev (2004) insists that a periodic policy regime acts as a commitment device. The results of this paper provide an alternative rationale for central banks’ infrequent policy changes, which have been widely observed in industrialized countries. Given that recent central banks often make the decision to “do-nothing”, the issue of policy shift frequency should receive much attention in future research.