Interpreting the Fed: what Bernanke really told us

Opinion is sharply divided about what the Fed intended to signal in the statement issued on Tuesday. Some have seen the statement as very dovish, because it said that the Fed intended to leave short rates at “exceptionally low levels” until mid 2013 – the first time that a specific date of this sort has ever been set by the FOMC.

Others, however, concluded that the statement contained nothing really new, since the markets had already assumed that short rates would be close to zero for the next two years. Furthermore, the fact that there were three dissents from the majority decision has led some to deduce that the further large step to more quantitative easing (QE3) is still a long way off. On this view, nothing really changed.

One way of trying to assess whether the Fed eased policy on Tuesday is to use the Taylor Rule, which translates economic conditions into the “appropriate” setting for monetary policy.

There are many versions of this rule doing the rounds in economic research, but I am going to use the one which some economists say is the closest to that used by the FOMC. This was published by Glenn Rudebusch of the San Francisco Fed in June 2010. The equation translates the core rate of inflation, and the gap between actual and structural unemployment, into the appropriate short term interest rate using the Fed’s historical response function. (Strictly, this does not give us the “correct” short rate, just the one which the Fed would set if it acts in the same way as it has on average in past 20 years.)

Using the FOMC’s most recent forecasts for core inflation and unemployment in 2011 Q4, we find that the implied fed funds rate is -2.5 per cent. However, in order to discover what rate the Fed should actually set today, we also need to take into account of the expansionary effect of QE. Rudesbusch also enables us to do this. On his central estimates, updated for the arrival of QE2 since he published his research, the increase in the Fed’s balance sheet has been equivalent to a reduction in the short rate of about 3.2 per cent. Consequently, the “correct” short rate right now would be about 0.7 per cent – fractionally higher than where it is today.

This, and the fact that the FOMC believes that inflation expectations are no better than “stable”, probably explains why the Fed has been reluctant to ease monetary policy further as economic activity has slowed down in recent months. Even Fed chairman Ben Bernanke and his dovish supporters have given no sign of any relaxation until the statement on Tuesday. Unlike last year, there has been no indication from them that monetary policy is “too tight”.

What about the future for short rates? After the June FOMC meeting, the committee published its central projections for core inflation and unemployment for 2012 Q4 and 2013 Q4. We can use these to interpolate their projections for mid 2013, the relevant date for the new commitment made on Tuesday. Using the economic projections made in June, the Taylor Rule suggests that the correct level of the short rate in mid 2013 would be 3.1 per cent, if the size of the Fed’s balance sheet remained unchanged until then. In other words, based on what the Fed expected for the economy in June, it was reasonable to expect that there would be a progressive tightening in monetary policy in the next two years, either coming from higher short rates, or a reduction in the size of the balance sheet.

However, this expectation has now changed, mainly because the Fed said that the paths for real GDP and unemployment are now worse than they expected in June, both in the past and in the future.  Furthermore, the downside risks to this forecast are said to have increased, and inflation is thought likely to settle at rates “at or below” the 2 per cent which is deemed consistent with the Fed’s mandate.

We do not know the precise forecasts which the Fed is now using, but we can make a reasonable estimate. I assume that the unemployment rate projected for mid 2013 is now 8.5 per cent (up from 7.6 per cent before), while the core inflation rate is 1.5 per cent (down from 1.7 per cent previously). On this basis, the correct short rate in mid 2013 would be 1.1 per cent, assuming that the Fed leaves its balance sheet unchanged in the meantime.

The Fed has now committed itself to leave rates unchanged at “exceptionally low levels” until mid 2013. So is it intending to pursue a monetary policy which is easier than the Taylor Rule suggests? Not necessarily. For one thing, “exceptionally low” rates could be taken to be consistent with rates rising to 1 per cent. For another, the Fed has not told us what it is assuming will happen to its balance sheet between now and 2013. If the Fed’s security holdings were to shrink by about one third over that period, roughly reversing the effects of QE2, then the appropriate short rate would be around its current level: close to zero.

Either way, the Fed does not seem to have committed itself to any easing in monetary conditions in the next two years. It is not yet ready to get ahead of the curve, and take pro-active steps to prevent the economy from slowing further. It is quite likely that it will ease policy further if unemployment and GDP growth deteriorate, but inflation expectations seem to be standing in the way of that happening pre-emptively.

Mr Bernanke showed that he will not allow a few dissenting voices on the FOMC to stop him easing further if he believes that conditions warrant such a move. That constitutes a smidgeon of dovish news, compared to what we knew before. But by that time, the economy might be in recession.