Via Citi's Steven Englander,
The consensus expectation is overwhelming that Fed Chair Yellen will deliver a dovish message at Jackson Hole. Macro investors have largely eliminated their short Treasury position and look to be long risk, particularly via equities and EM. FX positioning is long USD and long EM, the long USD largely because the euro zone economy is slipping again and the ECB is hinting at further ease. Our question is whether Yellen can be more dovish than what is now priced in, not whether she will be dovish on the Richter scale of dovishness.
We would go into this week long USDJPY. There is upside to USDJPY if geopolitical tensions ease further or US rates back up. Given the focus on Yellen’s talk, we are also a bit worried that investors will be surprised if FOMC Minutes read somewhat hawkishly.
EM would be the big winner if Yellen succeed in surprising on the dovish side. However we worry that dovishness is increasingly anticipated and that by the time we get to her talk, anything less than 'full dovishness' will be a disappointment.
‘Full dovish’ means moving the goal posts on the targets. Keeping the current targets, even accompanied by rhetoric and optimism, is hawkish because it suggests that normalization is coming as well get closer to the targets.
There are three ways by which Yellen can express dovishness, but only one that breaks new ground:
i) Full dovish
1) Argue that the natural rate is less than 5 ¼ - 5 ½ %
2) Advocate for a temporary overshoot of the inflation target
3) Emphasize the uncertainty around NAIRU estimates that tightening can wait till there is real evidence of accelerating inflation.
4) Introduce a soft wage target of about 3.5%, consistent with aspirational 1.5% productivity growth and 2% unit labor cost growth
ii) Semi dovish
1) Make the case that there is no sustained inflation likely without accelerating wage growth and there is little broad evidence for such a pickup, but keep existing inflation and unemployment targets
2) Introduce a new labor market indicator that captures the slack she feels that the unemployment rate misses, but again keep to existing targets
iii) Contingent dovish
1) Forecasting a pickup in productivity and labor force participation that will limit the need for tightening
Full dovish goes beyond anything she has stated explicitly in her comments. It would give stimulus more room on unemployment, inflation or both, and lead to yields dropping even further, taking the USD with them. There are straightforward arguments to justify full dovish, but the Chair has not advocated any of them so far, so it would clearly plant her among the most committed doves.
Semi dovish may generate a strong initial market reaction if it looks as if it is introducing new factors into the policy equation but is much more ambiguous. 'Low wages imply low inflation' is a property of most inflation models but much weaker than saying that wage growth is now a target. Were inflation to pick up for others reasons the Fed would still tighten, even if wages remained soft. Similarly it is unclear what it means to say that the unemployment rate understates slack, but 5 ¼ - 5 ½% is the target anyway. While we would focus on whether the goalposts are being shifted, semi dovish can sound very dovish until it becomes clear whether there is any functional shift in the FOMC’s goals.
Contingent dovish is the argument she has put forward for a long time. It sounds more dovish than it is because no one has a real handle on the drivers of trend productivity growth and the US supply side has disappointed badly. Both the participation rate and productivity growth are at weak levels and there is no compelling case that either will pick up. The safest assumption is that trend productivity growth over the next three years is what it was the last three years, and that participation rates are unlikely to surge.
Moreover, if the supply side does not improve, the Fed will have to start withdrawing stimulus as soon as the inflation and unemployment targets are approached. So even though the stress on supply side response sounds dovish, at this stage of the cycle it may actually turn out to be hawkish.
The hawkish surprise would be an acknowledgment that they were approaching their dual mandate targets faster than expected. Even repeating the FOMC statement would be something of a hawkish surprise given how far markets have moved in the dovish direction. Dec 2015 and 2016 Eurodollar rates are at the low end of their range of the past 15 months so there is no market concern on the pace of tapering. Were she simply to say that the targets are the targets and they will begin to reduce stimulus as they are approached, it would be a tremendous let down and viewed as very hawkish versus expectations.
The case for full dovish
The full dovish case basically argues that we do not have a direct measure of full employment or the NAIRU. It is inferred from the behavior of wages and prices. With this imprecision, the Fed should not tighten before there are indications from market-based wages and prices that the US economy is pressing against capacity limits. The NAIRU has fluctuated between around 6.5% and 4.5% over the last 50 years, but is difficult to determine ex ante and there is no reason to guess. (The counterargument is that inflation is a lagging indicator so by the time we see it, it is too late for a soft landing, but that is not the argument she is likely to make).
Productivity growth has been dropping since the mid 1990s.
In fact where we are now is pretty much in line with the trend that began in 2004, and if anything the risk is that we are falling below. Most cyclical productivity gains come during the early recovery period and that was in 2009-10.
Labor force participation
The US has averaged 195k jobs for the last 40 months, so it is harder to make the case that workers are discouraged by the absence of jobs. The employment to population ratio of older works is higher than before the crisis, unlike that of so-called prime aged workers (Figure 2).
In the past we have seen labor force gains towards the end of recoveries. It is hard to pin down what drives the gains but the pickup is not so fast even in the initial stages of strong recoveries when output and employment gains are rapid.
Figure 3 shows the participation rates of men aged 25-54 years. There has been a downward trend the last 50 years, often exacerbated by recession and with modest subsequent rebound in participation. (we use male participation rates because the influx of women into the labor force in the 1970s and 1980s would make it very hard to see the cyclical behavior.)
There is a difference between the argument that strong growth will encourages workers to rejoin the labor force and keep a lid on wages and prices and the more pessimistic argument that higher wages for everyone are needed to increase participation rates by one or two percent. The latter leads to higher unit labor costs in general, the former to lower.
* * *
With opinions mixed as to whether or not Jackson Hole will be the forum for Yellen to say something new, many are trying to figure out if it is a buy the rumor and then buy more after the fact event, a buy the rumor sell the fact event, or a do nothing with the rumors and then buy the fact if the USD is actually rallying after the fact event.
* * *
As The WSJ reports, it's going to be hard for the Fed to justify more dovishness in light of their own exuberant slack-less data...
But we are sure they will find a way