As JPM's Michael Feroli notes, the September FOMC Taper announcement (which certainly isn't assured, although if the Fed does not taper, it will end up monetizing 0.4%-0.5% of the total private TSY stock per week before year end) may just be a sideshow to a previously undiscussed main event: the Fed's first forecast of 2016 interest rates.
Quote Feroli: "The problem is that at the end of 2016 [the FOMC's] economic forecasts may well show an economy that is close to full employment and price stability. Normally in that situation one would expect the fed funds rate to be close to neutral—which is somewhere close to 4%. However, their end-of-2015 forecasts have a funds rate forecast centered around 1%. An end-of-2016 funds rate of 4%, which implies 300bp of tightening over the course of 2016, is well in excess of what the market is pricing in....If the Fed presents a 2016 interest rate forecast that is well above the market's expectations—and if the market takes any cue from the Fed—this could tighten financial conditions such that the forecasted acceleration in growth fails to materialize" Want to crash the Fed? Just turn off the default Circular option in their excel models.
As Feroli adds, "In some ways the Fed is at risk of being a victim of its own success."
Feeling boxed in yet Chairman Ben? Or just dazed and confused by all the reflexivity you have created, and which even Bill Gross mocked recently?
The 2016 Problem
The Fed faces an interesting situation at the September FOMC meeting. At that meeting they will introduce their 2016 interest rate forecasts for the first time. The problem is that at the end of 2016 their economic forecasts may well show an economy that is close to full employment and price stability. Normally in that situation one would expect the fed funds rate to be close to neutral—which is somewhere close to 4%. However, their end-of-2015 forecasts have a funds rate forecast centered around 1%. An end-of-2016 funds rate of 4%, which implies 300bp of tightening over the course of 2016, is well in excess of what the market is pricing in. If the FOMC were to produce such a forecast, and if the market were to take its cue from that forecast, then the ensuing tightening in financial conditions would undo much of the hard work the Fed has done in getting rates low enough to support the recovery.
In some ways the Fed is at risk of being a victim of its own success. After a few fits and starts, the Fed has finally convinced the market it will keep rates low for a very long time. However, the Fed’s ever-expanding embrace of transparency means it now has to quantify how its verbal attachment to accommodative policy translates into economic and policy rate forecasts. For someone like Michael Woodford, whose paper at last year's Jackson Hole conference arguably had a meaningful impact on the policy debate, such forecasts shouldn't have much of an influence on financial conditions: after all, it is well known that the Fed doesn’t have much of an advantage over the private sector in forecasting accuracy.
That being said, we do think the forecasts convey a policy message, as they indicate how the Fed will react to a given economic outcome. In particular, the 2015 forecasts are already well below what most Taylor rules would prescribe. We expect a similar gap for the 2016 forecasts, which would leave the mid-point of those forecasts around 2.25%, or perhaps a shade above. This would be somewhat above where the market is now pricing the Fed, but not nearly as much as if the funds rate went back to neutral. Such a forecast, well below a Taylor rule, would continue to signal that the Fed will remain “highly accommodative…even after the economic recovery strengthens.” A risk to this outlook is that not enough of the Committee “gets the memo,” so to speak, and pencils in forecasts that are at odds with what the leadership would probably like to see.
The set up
Before jumping straight to the September outlook for 2016, it may be helpful to first review the outlook for 2015 that is captured in the June Summary of Economic Projections (SEP). In those forecasts, the mid-point of the 4Q15 projected unemployment rate was 6.0%, the mid-point of the estimated natural rate of unemployment was 5.6%, and the unemployment rate was falling about 0.5%-point per year over the prior two years. The mid-point of projected 4Q15 inflation was 1.8%, 0.2%-point shy of the 2.0% inflation goal. For the “dots,” which show the funds rate projection in 4Q of each year, we discard the four highest projections. The hawks on the Committee are nice people, and some are good economists, but they have had very little influence on policymaking in recent years, and we expect that to continue. This trimmed forecast has close to a 1.0% funds rate projected at the end of 2015.
The first and most important thing to note about the 2015 forecasts is that the funds rate projection is well below what most Taylor rules would prescribe, given the Fed’s economic projections. For example, the “Yellen rule” (from her June 2012 talk, which is the 1999 Taylor rule with an Okun adjustment of 2.3 to convert unemployment gaps into output gaps) would indicate a 4Q15 funds rate of 2.8%, 180bp above the actual forecast. Most other Taylor rules point to an even higher predicted rate. For example, the "Rudebusch rule" (estimated by the SF Fed's director of research Glenn Rudebusch) indicates a 3.6% funds rate at the end of 2015. The main point is that already in 2015 the Fed is to a large extent obeying the policy prescriptions (as per Reifschneider and Williams, Eggertson and Woodford, and others) of exiting a liquidity trap by keeping rates lower for longer than would be prescribed by a Taylor rule.
On to 2016
Returning to our main issue, the 2016 forecast, if we simply straight-lined the 2013-2015 improvement into 2016, then we would be close to full employment and 2% inflation at the end of that year. There are a few considerations that could come into play, however, when doing such a straight-lining. First, the 2013 GDP forecast will almost certainly be revised down. There is a case for lowering the out-year forecasts as well. It’s not clear how much of the May-June rise in long-term rates was factored into those forecasts, and in most Fed models a rise in long-term interest rates can have a surprisingly large depressing effect on growth. However, much of that growth restraint operates through the housing market, and given the offsetting positive factors supporting housing, the Fed’s judgmental forecast may not have taken too much off of growth due to the rise in interest rates. Second, slower trend growth implies a lower long-run, or neutral, funds rate. This would mean they would need to tighten less over the course of the upcoming cycle. Arguing against looking for this at the upcoming meeting is the fact that the Fed is very slow to change their estimates of unobserved variables—the natural unemployment rate, potential GDP growth, and the neutral funds rate. Third, the ongoing decline in the labor force participation rate has arguably lowered the natural unemployment rate—persons who were previously classified as structurally unemployed (and thus captured in NAIRU) may now be structurally out of the labor force, and thus no longer measured as unemployed, lowering NAIRU. The same caveat as above applies to this argument: the Fed is very slow to change estimates of unobserved variables. Adjustments to all or any of the above variables are likely to be only minor tweaks, and thus leave the economic forecast at the end of 2016 close to the Fed’s goal of full employment and price stability.
Low for long
The economic forecast for 2016 probably won’t give the Fed much justification for keeping rates accommodative, instead there are other reasons to expect the interest rate forecast to remain below what is implied by a Taylor rule. The most important of these is liquidity trap logic: in order to stimulate the economy when rates cannot go below zero, the Fed must commit to keeping rates low even well after the economy recovers. The question then is can the Fed do this credibly? That is, the Fed may tell the market now it will keep rates low in 2016, but why should the market believe it?
One obvious response is: the market already believes it. The 2015 funds rate forecast is around 200 basis points below what a Taylor rule prescribes. If the 2016 forecast had a similar gap it would place the funds rate forecast at 2.0%, pretty close to market expectations. As it is, we expect that the over time the gap between the Fed's forecasts and something like a Taylor rule will close, but it should do so only gradually. Of course, the market may now “believe” the Fed's 2015 interest rate forecast because they don't believe the Fed's optimistic economic forecast, but if so we see no reason why that difference of opinions should change with the release of the September SEP. Another reason the market may believe the Fed’s 2015 forecast is that inflation over the past twenty years has been a slower-moving variable—and less responsive to slack or tightness—than economists of older generations have thought. This means being behind the curve by a year or so on the tightening cycle won’t threaten a 1970s-style inflation risk. A final reason the market may believe the Fed’s forecast of continued easy monetary policy is that the forecasted unemployment rate may not provide a true measure of labor market utilization. Depressed labor force participation rates may mean the Fed has room to remain accommodative even after the unemployment rate gets close to NAIRU, in order to draw discouraged workers back into the labor force. This is a theme that has been mentioned both by Chairman Bernanke as well as in research by senior Fed staffers.
So far, the Fed’s embrace of transparency hasn’t conflicted with its desire to provide extraordinary monetary stimulus. September will provide an interesting test. If the Fed presents a 2016 interest rate forecast that is well above the market's expectations—and if the market takes any cue from the Fed—this could tighten financial conditions such that the forecasted acceleration in growth fails to materialize. Instead, we think the interest rate forecast will be somewhat above where the market is, but not radically so. Such a forecast may appear at odds with the Fed’s economic forecast, but we believe there are good reasons for that gap to exist. So far the market seems to agree.