For example, the 5y5y forward real (TIPS) rate actually fell about 100bp during the first year of the tightening episode and generally remained in a 2-2.5% trading range for years thereafter. The long-run real rate remained around that level (with more volatility as the financial crisis and its aftermath played out) even after the dramatic easing of (conventional) monetary policy brought the fed funds rate to nearly zero, in early 2009. It was only in early 2011 that the long end of the US curve cracked decisively and plunged to unprecedentedly low levels.
More recently, the long end of the curve has displayed a sensitivity to monetary policy announcements that is as pronounced as it was muted during 2004-09. We think that this is at least as puzzling as the insouciance with which the yield curve reacted to the Greenspan policy tightening. True, the sell-off from extreme levels that ‘safe haven’ bond markets reached late last year can be shrugged off as a correction of frothy conditions in fixed income markets; much of this predates the May tapering shock and last week’s FOMC surprise decision to postpone the tapering.
We do not view the sensitivity as a reaction to a signal by the FOMC about where monetary policy is likely to be during that period. If the Fed’s communication to the market since late May proves anything, it is how uncertain the outlook for monetary policy remains within the Fed over a 3-6 month time frame. This is no criticism of the Fed, just an acknowledgement of the uncertainty it faces in the fundamental drivers of monetary policy over that short horizon. Given this, the FOMC is in no position to send credible signals about where policy rates (never mind real interest rates) will be five years from now, nor do we think it had any intention of doing so this week.
It also makes little sense to extrapolate the economic weakness that underpinned the Fed’s decision not to taper into 2018-23. Five years is a long time, and although the 5y5y forward period could wind up being another episode of economic weakness, a more plausible base case is that the economy continues to heal and at some point between now and then returns to a more normal state that calls for more complete normalization of monetary policy. Much healing can take place during five years. If, for example, the unemployment rate were to continue to drift down at the rate that it has done during the past four years of disappointing but reasonably steady recovery, it would reach 4.5% in five years, close to the level it reached during the 2005-07 economic and financial boom. This is an extrapolation, not a forecast, but it offers a reminder how dramatically the cyclical context can change in five years.
One explanation of the ‘excess sensitivity conundrum’ may be that the Fed’s direct intervention in the US bond market is so powerful that it overwhelms fundamental drivers. This may well be an important part of the explanation. To the extent that it is, we resign ourselves to reminding investors that what QE gives, it will eventually take away.
A complementary explanation may lie in a powerful lesson that the recent financial crisis taught investors, which is how little visibility we really have about how the world will look in, say, half a decade’s time. Before the financial crisis, and indeed well into the subsequent recovery, investors seemed comfortable pricing bonds as though financial conditions displayed enough mean-reversion to justify pricing 5y5y forward rates for historically ‘normal’ market conditions, shrugging off transitory influences of business and monetary cycles. Hence, the Greenspan “conundrum” of low sensitivity to monetary policy.
After the unprecedented events of 2007-08, and the deep and long-lasting recession that has ensued, many investors may now feel more comfortable extrapolating current conditions into the future, rather than relying on a powerful degree of mean reversion in economic and financial conditions. This response is understandable, but seems to underestimate the forces of economic recovery that are slowly but reasonably steadily playing out and that seem likely to call for a normalization of monetary policy during the coming half-decade that is more complete than is priced in the market.
But this week’s roughly 25bp decline in the 5y5y forward real rate looks like a direct response to the FOMC announcement. This is a large move, even if it is only a partial reversal of the year-to-date sell-off. From a fundamental economic point of view, it is hard for us to accept that the strong sensitivity of real rates in 2018-23 to monetary policy makes sense.