When Will The Fed Change Its Reaction Function?, by Tim Duy: The March FOMC minutes were generally interpretted as having a dovish tenor, contrasting with the generally hawkish reception for the statement and ensuing press conference. Overall, the Fed appears committed to a long period of low interest rates and I continue to think this should be the baseline view. But actually policy seems to remain hawkish relative to the Fed's rhetoric. By its own admission, the Fed is missing badly on both its mandates. Why then the push to reduce accommodation by ending asset purchases and laying the groundwork for the first rate hike? This leaves me wary the Fed could turn dramatically more hawkish with little provocation from the data. At the same time, one can imagine the Fed realizes that the current reaction function remains inconsistent its desired goals, and policy consequently shifts in a dovish direction.
Consider the Fed's take on labor markets:
In their discussion of labor market developments, participants noted further improvement, on balance, in labor market conditions.
Fair enough. But where is the majority of policymakers on the issue of slack?
While there was general agreement that slack remains in the labor market, participants expressed a range of views regarding the amount of slack and how well the unemployment rate performs as a summary indicator of labor market conditions. Several participants pointed to a number of factors--including the low labor force participation rate and the still-high rates of longer-duration unemployment and of workers employed part time for economic reasons--as suggesting that there might be considerably more labor market slack than indicated by the unemployment rate alone.
The opposing view was held by just a "couple" of participants. The "high slack" contingent holds of the upper hand, in my view, given the limited wage pressure to date:
Several participants cited low nominal wage growth as pointing to the existence of continued labor market slack. Participants also noted the debate in the research literature and elsewhere concerning whether long-term unemployment differs materially from short-term unemployment in its implications for wage and price pressures.
It seems fairly clear that the dominant view on the Fed is that labor markets contain more than sufficient slack to contain wage and inflation pressures. And inflation pressures are, by their own admission, nonexistent. But this concern is not as widespread:
Inflation continued to run below the Committee's 2 percent longer-run objective over the intermeeting period. A couple of participants expressed concern that inflation might not return to 2 percent in the next few years and suggested that a protracted period of inflation below 2 percent raised questions about whether the Committee was providing an appropriate degree of monetary accommodation.
Why is the majority not concerned? Because even as they use low wages to justify claims of sufficient slack in the labor market, they use a forecast of higher wages to dismiss the inflation numbers:
A number of participants noted that a pickup in nominal wage growth would be consistent with labor market conditions moving closer to normal and would support the return of consumer price inflation to the Committee's 2 percent longer-run goal.
But how long will the process take? A long time:
Most participants expected inflation to return to 2 percent over the next few years, supported by stable inflation expectations and the continued gradual recovery in economic activity.
The Federal Reserve is clearing communicating the willingness to endure a sustained period of suboptimal outcomes on both the employment and price stability metrics. This suggest that actual policy - entirely directed at reducing accommodation - is considerably more hawkish than dictated by data. It sounds like policy fatigue. The Fed wants out of asset purchases and zero rates and are willing to dismiss the dual mandate to move in this direction. No wonder then that Chicago Federal Reserve President Charles Evans is worried that policymakers will push too hard to normalize rates too early. Via the Wall Street Journal:
“One of the big risks is that we withdraw our accommodative policies prematurely,” Mr. Evans said during a panel discussion at the International Monetary Fund’s spring meetings. “I think it’s just human nature to start thinking we’ve been doing this for a long time.”
The Fed’s benchmark short-term interest rate has been pinned near zero since late 2008, which could prompt some policy-makers to think “that must have been long enough. Maybe it’s time to start the process of renormalizing,” Mr. Evans said. Most Fed officials indicated last month they expect to start raising rates next year.
Consider also the Fed's willingness to continue the taper despite persistent low inflation in the context of this from Federal Reserve Governor Daniel Tarullo:
Last week Chair Yellen explained why substantial slack very likely remains. I would add to her explanation only the observation that, in the face of some uncertainty as to how best to measure slack, we are well advised to proceed pragmatically. We should remain attentive to evidence that labor markets have actually tightened to the point that there is demonstrable inflationary pressure that would place at risk maintenance of the FOMC's stated inflation target (which, of course, we are currently not meeting on the downside). But we should not rush to act preemptively in anticipation of such pressures based on arguments about the potential increase in structural unemployment in recent years.
Arguably, tapering implies that are already acting prematurely. Combine with this commentary by David Zeros via Business Insider:
"As the market prices in higher short-term yields and lower long-term yields, it is really making a bet that the Fed, by tapering our punchbowl drip, is increasing the risk of deflation," says Zervos.
"And at this stage of the game, with inflation BELOW target and plenty of slack in labor markets, that could very well be a mistake. The most important point here is to recognize that low long-term yields are not a sign of a healthy economy."
Indeed, it is reasonable to believe the Fed will make a mistake in the hawkish direction (or already has) given that policy already seems inconsistent with the dual mandate. In other words, the Fed has a hawkish reaction function.
Regarding that reaction function, the now infamous dots were also a topic of discussion. Policymakers knew exactly the implications of the dots:
A number of participants noted the overall upward shift since December in participants' projections of the federal funds rate included in the March SEP, with some expressing concern that this component of the SEP could be misconstrued as indicating a move by the Committee to a less accommodative reaction function.
The next line, however, is not particularly helpful:
However, several participants noted that the increase in the median projection overstated the shift in the projections.
This begs the question of "why?" Some dots moved forward. Why does that overstate the shift? That said, some participants noted that the shift should not be cause to worry:
In addition, a number of participants observed that an upward shift was arguably warranted by the improvement in participants' outlooks for the labor market since December and therefore need not be viewed as signifying a less accommodative reaction function.
This was my interpretation - the upward shift of the dots were consistent with a change in the unemployment projections given the Fed's reaction function. But that doesn't quite explain why the reaction function is so tight to begin with. This is I think the best explanation:
In their discussion of recent financial developments, participants saw financial conditions as generally consistent with the Committee's policy intentions. However, several participants mentioned trends that, if continued, could become a concern from the perspective of financial stability. A couple of participants pointed to the decline in credit spreads to relatively low levels by historical standards; one of these participants noted the risk of either a sharp rise in spreads, which could have negative repercussions for aggregate demand, or a continuation of the decline in spreads, which could undermine financial stability over time. One participant voiced concern about high levels of margin debt and of equity market valuations as well as a notable shift into commodity investments. Another participant stressed the growth in consumer credit to less creditworthy households.
I think the Fed's reaction function now includes some financial stability variable, but the Fed is loath to discuss that variable and the related parameters impacting policy. That said, we are fairly confident that the push to end asset purchases and plan the exit from zero rates were a response to bubbling financial stability concerns at the Fed. They simply hid that behind the "progress toward goals" language.
More surprisingly is that not only did they begin the exit from extraordinary stimulus in the face of clearly suboptimal labor outcomes, they did so in the face of clearly suboptimal inflation outcomes. Now, though, they may be realizing the error of their ways. Via Jon Hilsenrath at the Wall Street Journal:
Federal Reserve officials are growing concerned the U.S. inflation rate won't budge from low levels, the latest sign of angst among central bankers about weakness in the global economy.
So what comes next? To answer that, we again need to divide policy into movements along the reaction function and shifts of the reaction function. We should recognize that the SEP dots will shift in response to the data. If data comes in stronger than anticipated, then the dots will move forward. If weaker, then backward.
A more hawkish reaction function - the dots moving up and forward independent of the forecast - would most likely occur due to heightened financial stability concerns. A less likely cause is that inflation expectations suddenly jump.
What about a more dovish reaction function? I think it was expected that new Federal Reserve Chair Janet Yellen would have already pushed forward a more dovish reaction function given her expressed concerned for the unemployed. So far, she has disappointed such expectations. Factors that could still trigger a downward shift include 1.) a desire to accelerate the pace of improvement in labor markets, 2.) a lessening of financial stability concerns, 3.) a heightened concern about the negative impacts of persistently low inflation.
The inflation concern is my leading candidate at the moment. Still, I would not want to overestimate the chance of such a shift. It is easy to see that ongoing improvements in labor markets could be sufficient to contain inflation concerns to low rumblings.
Bottom Line: Fed policy - dovish those it seems - is maddenly disconnected from their actual forecasts. What does that mean for future policy? Given the relatively dovish forecast, I am concerned that the balance of risk lies on the upside, which implies tighter policy along the existing reaction function. But at the same time I remain open to the possibility that even if the economy evolves as expected, the Fed could extend the low interest rate horizon via shifting the reaction function down. That said, I suspect there is a fairly high bar to such a shift. As unemployment drops further, they will become increasingly concerned about being caught behind the curve given the level of financial accommodation already in place.