Policy Rule Changes
One area of active interest for both policymakers and market watchers is to find a simple rule (or rule of thumb) that approximates Fed policy on interest rates. John Taylor came up with the first such rule in 1993, and since then, a number of variations have been proposed. One variation suggests that the Fed responds positively to increases in inflation above target (currently 2 percent) and negatively when unemployment increases.
While this simple rule of thumb tracks broad movements in the federal funds rate, the average absolute value of the miss is 87 basis points. To put this another way, if we assumed that our best guess of today’s funds rate was yesterday’s funds rate, the average absolute miss would only be 32 basis points.
For this reason, it is frequently argued that the Fed responds not only to inflation and unemployment, but also to last quarter’s funds rate. A rule incorporating all of these elements is known as a partial adjustment rule. In practice, this rule would mean that the central bank uses the Taylor rule as its intermediate target and only partially moves the level of the funds rate to this value at every meeting.
At first glance, this rule tracks the funds rate remarkably well. But looks can be deceiving. The deviation of the funds rate from its predicted value is given by the vertical distance in the chart below. Take the end of 2001, for example. The miss on that date was a whopping 133 basis points. Since the average absolute funds rate change is only 32 basis points, this 133 basis point miss is huge. Even more disconcerting is that it does not beat the naive rule, where the funds rate today is given by yesterday’s funds rate and the average absolute miss is 32 basis points. As the chart shows, the partial adjustment rule is essentially a simple phase shift of the actual funds rate.
Obviously, there is still something important missing from this partial adjustment rule. One possibility is that the rule assumes that the Fed is adjusting the level of the funds rate to the level of the Taylor rule. Implicit in this way of thinking is that no change in the funds rate translates to no policy change. But if the Federal Open Market Committee (FOMC) had been steadily reducing rates by 25 basis points over the past few meetings, keeping rates constant would probably be viewed by most as a change in the course of policy. This type of thinking focuses on changes in the funds rate and not the level of the funds rate.
Metaphorically, if a boat is traveling east toward the harbor at five knots, should we think of the constantly changing location as a change in the skipper’s policy, or should we think of a change in the skipper’s policy as a change in the boat’s speed? With a level policy, the current speed is independent of the past speed and depends only on his current distance from his destination. But such a policy may imply a very sharp acceleration or deceleration, which could be uncomfortable for the passengers (markets). The change in policy we consider is one where the skipper considers both the distance from the destination and his recent speed. This implies a smoother path into the harbor.
We explore a description of monetary policy expressed in terms of funds rate changes, instead of the level of the funds rate. Here the change in the funds rate moves gradually toward an intermediate target. There is a subtle but important distinction between the traditional level rule and our change rule. Suppose that at the previous meeting the FOMC had increased the rate to 3.25 percent. Under the traditional level rule (a partial adjustment rule based on the level), the FOMC’s choice today is independent of how the FOMC arrived at 3.25 percent at its last meeting. In contrast, under our change rule (a partial adjustment rule based on the change in the rate), the FOMC would also consider the rate changes that led it to 3.25 percent.
Our change rule expresses the change in the funds rate as a weighted average of yesterday’s change in the funds rate and the deviation of yesterday’s funds rate from a simple Taylor rule, or the “intermediate target.”
Compared to the level rule, the improvement in fit with the change rule is substantial: a reduction of 10 basis points in the residual. This reduction is sizeable, given that the average rate change is 32 basis points. There is one problem with the change rule: It often overshoots the actual funds rate at the end of sustained policy movements. The change rule is trying to proxy for the idea that the FOMC does not like to change the course of policy abruptly. That is, other things equal, the FOMC would not want to decrease rates if there is a likelihood that it would need to increase rates in the near future.
Focusing on shorter subperiods highlights some of the differences between the level and change rules. The phase shift under the level rule is quite evident, while this shift is largely eliminated with the change rule. For example, during the sustained increase in rates starting in early 2005, the level rule is always a quarter behind, while the change rule is on target. There is also the overshooting under the change rule, overshooting at both the end of 2006, and the fall of 2008.
These two episodes are almost certainly a manifestation of the fact that the FOMC does not mechanically follow a simple policy rule but responds to unusual developments in the economy. The 2006 overshooting is likely a reflection of the FOMC’s desire to limit funds rate increases in the wake of the substantial change in the behavior of house prices. As for the fall of 2008, almost certainly the FOMC moderated the funds rate decline (relative to the change rule) because of the near proximity of the zero bound (where the funds rate approached zero). This moderation may have derived from the FOMC’s desire to save some policy ammunition for a later date. A similar argument likely applies for the change rule’s overshooting in the spring of 2002. A review of FOMC minutes reveals that there was discussion of the zero bound at the January 2002 meeting.