Before the presser on Wednesday, Ben Bernanke’s vague definition of “substantial improvement” in the outlook for labour markets resembled the old line about porn: he’ll know it when he sees it.
The phrase was originally intended to represent the scenario under which asset purchases would end, not when they would be slowed (or “tapered”). And the purpose of this round of quantitative easing was to “increase the near-term momentum” of the economy until growth was self-sustaining, and conducted in the context of price stability.
But the Fed chairman had also said previously that he would prefer a variable policy, one in which the amount of monthly asset purchases would be raised or lowered depending on fluctuations in the labour market outlook.
(The reason that the FOMC hadn’t embraced numerical guidance for asset purchases, as it has for rates, was that the costs and effects of asset purchases remained uncertain. It was a topic that the FOMC had continued to debate without resolution.)
Bernanke’s decision on Wednesday to confirm his earlier congressional testimony — when he said that the Fed would begin scaling back asset purchases later this year — can be understood as a move in the direction of such a variable policy.
The idea was that given the FOMC’s heightened expectation of better growth later in the coming years, a substantial improvement in the outlook for labour markets is likely to have occurred by roughly the middle of next year. And it would be represented by a 7 per cent unemployment rate.
But rather than continuing with the current pace of asset purchases until mid-2014, “it would be appropriate to moderate the monthly pace of purchases later this year”. Presumably by then, the near-term momentum will have increased such that fewer and fewer asset purchases are needed to sustain the quicker decline in the unemployment rate that the Fed is now forecasting.
Bernanke emphasised the point repeatedly that if the economy disappointed the Fed’s outlook, then this would factor into his decision on whether to reduce the pace of LSAPs — and if it disappointed by enough, then he would even consider increasing the pace.
But he didn’t offer any hints on when “later this year” would be (the first tapering could begin anytime between the July and December meetings inclusive), nor did he suggest how much he would scale them back at the start, how quickly he would scale them back in subsequent meetings, or anything specific on how he would make such decisions.
So it turned out to be a confusing signal, one of several, and the day marked Bernanke’s worst communicative performance in some time — though it’s also true that the task he set for himself was a tricky one.
For instance, the Fed now sees diminished “downside risks” to the economy relative to what it saw in May, and its forecast for unemployment became more optimistic relative to its forecast three months ago. The expected reasons are offered: less fiscal drag, vigourous housing activity, a more stable Europe.
Yet inflation is forecast to be lower this year, and roughly the same next year. Bernanke downplayed the recent disinflation by referencing transitory and “non-market” factors. But he also said in the press conference that he was concerned about getting inflation back up to target, and that it “will be a factor in our thinking about the thresholds [for raising rates]. It will be a factor in our thinking about asset purchases.”
Meanwhile, the recent rise in yields and the decline in inflation expectations is placing upward pressure on expected real rates — very likely a response that Bernanke neither anticipated nor wanted.
And as Credit Suisse economists note, compared to March:
FOMC participants on the margin prefer a slower pace of policy firming today than they did three months ago. Again, this is despite somewhat faster growth and lower unemployment rate projections. It may be that distrust of the unemployment rate is growing, while concern about below-target inflation could be stronger than today’s policy statement would have us believe.
There is good reason for the distrust of the unemployment rate. An unresolved issue is the extent to which currently depressed levels of labour force participation are down to cyclical or structural factors. The more it’s the latter, the more it is expected that labour force participation will recover — and keep the unemployment rate elevated. (And vice versa.) But there is research from Fed and private sector economists pointing in different directions: try here, here, here and here.
This too will factor into the decisions about scaling back asset purchases and when to first raise rates, and it’s yet another variable to watch without knowing exactly what to watch for.
So to summarise the above signals, as one might reasonably receive them…
The economy is expected to improve and unemployment come down faster than I previously thought, but inflation is also expected to be lower this year than our earlier forecasts despite this improvement. But not to worry: inflation should bounce back in the direction of 2 per cent by next year, as its current downward trend is likely the result of specific temporary factors, even if to everyone else it seems broad-based. But if it doesn’t, or if unemployment doesn’t come down as I expect, I’ll take that into account in deciding when to start and how fast to taper. And by the way, my colleagues on the FOMC are leaning towards a later start date for the first rate rise, which seems to contradict the more-optimistic outlook. Two possible reasons, in case you’re still following, are that we don’t necessarily trust the unemployment rate and that we’re actually a bit more worried about disinflationary pressures than we let on.
Markets have freaked out a bit. That’s okay: markets can be silly, even stupid, and they can change direction quickly. In this case, they might be disappointed at having to price in a probability of earlier tapering than expected, or they might doubt the Fed’s rosier forecasts and Bernanke’s commitment to continuing current policy if those forecasts prove wrong. Or maybe they’re just confused.
The argument against these interpretations is nicely explained by Gavyn Davies:
The Fed has taken a calculated risk by starting the exit process before there has been any significant improvement in the labour market, only the expectation of one. Past experience suggests that the bond market might be very sensitive to the first signs that the central bank is losing its enthusiasm for further easing. In 1993-94, and to a lesser extent in 2003, bond yields rose sharply, and the equity market fell for a time, on the first indication that the Fed might be contemplating a change in direction. Since central banks tend to manoeuvre very slowly, somewhat like supertankers, this is a rational response, and it has already shown signs of taking hold this time.
The big risk is that bond yields will rise too far, before the improvement in the labour market has been given enough chance to gather momentum. When Bernanke commented in May that tapering could take place “in the next few meetings”, it seemed that the Fed might be reneging on its commitment that asset purchases would continue until there had been a substantial improvement in the outlook for the labour market, in the context of price stability. Many were ready to criticise the Fed for this, but tapering has now been made explicitly contingent on further, specified improvements in the labour market, which is a much better place to be. …
Recent data suggest that this [substantial improvement] might take quite a while. …
If the Fed viewed the labour market as a major national problem late last year, it is not clear from the bulk of this information why it should have changed its minds since then. It seems from the new unemployment thresholds introduced on Wednesday that it basically agrees with this.
So the exit has started, and risks to all asset classes have risen as a result. But the Fed is working very hard to persuade the market that it really will be different this time, and the labour market data certainly suggest the process will be a long one. The response of the bond market shows that investors are aware that the supertanker is turning, but I suspect that the conditions required for a major bear market in bonds are not yet in place.
If so, then given how much context was needed to arrive at this conclusion, it’s no wonder the message was difficult to communicate — and appears thus far to have been misinterpreted.
Bernanke answered the market’s question about whether he would start slowing the pace of asset purchases soon. He was much less clear about when, how, and even why.
Copyright The Financial Times Limited . All rights reserved. Please don't copy articles from FT.com and redistribute by email or post to the web.