November 17, 2003
Reply Hazy, Ask Again
The Wall Street Journal's Steven Liesman stares into his Magic 8 Ball and tries to figure out what the Federal Reserve will do. But his conclusion seems to be "Reply Hazy: Ask Again"--and this makes him mad at the Fed:
WSJ.com - The Macro Investor: ...All of that leaves us where we started: Still trying to figure out what the Fed will do and when will it do it. There are basically two schools of thought. One group believes the Fed is on hold throughout 2004. They believe the line of Governor Ben Bernanke who has argued, rather convincingly, that the inflation rate could actually fall from here even amid strong growth. That's because the economy has a lot of slack, in the form of unemployed workers and idle factories. The economy has to run red hot for several quarters -- as much as a year -- before we have any danger of inflation. Nonsense, say those who think a Fed rate increase could come as soon as this spring. They see higher commodity prices and soaring gold prices as the leading edge of a rebound in inflation. They believe job creation is and will be much stronger than expected.
They have a powerful ally: the Fed-fund futures market has priced in virtually a 100% chance of a rate hike at the March Open Market Committee meeting.
Another part of the argument concerns whether it matters. John Ryding at Bear Stearns sees little harm in the Fed taking back a little of the monetary policy insurance it has provided to the economy. He adds that a strong recovery not only can withstand higher interest rates but also requires it. Bill Dudley at Goldman Sachs, who thinks the Fed is on hold next year, argues that that could do quite a bit of harm. He's worried that once the Fed starts tightening the market will start pricing in further tightening. The resulting rise in interest rates could well hurt the economic recovery.
The best bet is to keep an eye on the data and the statements of Fed officials, and yes, the absence of the "considerable period" phrase. Two more quarters of above trend growth (north of 4%) that produce an average of 300,000 jobs a month will likely lead to a tightening. Such a trend will be relatively obvious by February or March, so the spring theory works. But I think the May meeting is when the likely first hike under that scenario would come, because the Fed will want some assurance that raising rates is the right move.
A big caveat: If Mr. Bernanke is correct and, amid that strong economic performance, the inflation rate actually declines, then all bets are off. If we see a decline in consumer prices, then the "On-Hold in '04" Fed campaign slogan works. A rule of thumb to consider: the Fed will likely want two quarters in a row of strong growth and job creation to begin tightening. So every quarter below 3.5% growth on GDP, along with just so-so job creation of between 100,000 and 200,000 monthly, could give us another quarter beyond the spring of the Fed doing nothing.
That should give them a considerable period of time to refine how they communicate with markets.
But the source of Mr. Liesman's frustration about not knowing what the Federal Reserve will do springs from the fact that the Federal Reserve does not know what it will do.
Take a look once again at actual real GDP quarter-by-quarter and what we think, based on capacity utilization, the economy's stable-inflation potential level of output is:
We see that there is a large gap--a near record gap for the post-World War II period--between the economy's productive potential and actual real GDP. Such a low level of capacity utilization has always in the past been associated with downward pressure on the unemployment rate. That's where the Bernanke view comes from. We also see that output growth is and has been substantial for two years, and that consumer price index inflation no longer appears to be falling. That's where the non-Bernanke view comes from. If the October-March data show a falling unemployment rate, rising capacity utilization, rapid job creation, and upticks in inflation, then the Federal Reserve will start raising interest rates (perhaps in May?). If the October-March data show a steady unemployment rate, stagnant capacity utilization, slow job growth, and stable or falling inflation, then the Federal Reserve will wait and wait and keep interest rates low for a considerable period.
But it is unfair for Mr. Liesman to ask that the Federal Reserve tell him clearly, unambiguously what it is going to do when it does not know itself.
Posted by DeLong at November 17, 2003 05:02 PM | TrackBack
If the economy is growing joblessly because of the effects of information technology, surely it suggests that idle workers and factory space and tools are no longer "slack" in the sense of being available to curtail inflation. Rather they are simply out of use, in the sense that linear programming famously may dictate that it is more productive to leave a machine in a carpet factory out of use.
If so then, quite apart from trying to know the unknowable, Bernacke would be reasoning on false premises.
Surely, Liesman means the May funds contract, not the March.
Posted by: K Harris on November 18, 2003 05:51 AMBy the way, Broaddus has just said pretty much the same thing that David notes from Bernanke. I think the point for all of the Fed guys is that, they do think much of what they are dealing with is unknowable. What they observe is that the trend in most inflation measures is down, and what they know from experience is that these trends don't change in the course of a few months. That gives them some assurance that inflation will not bite 'em in the backside. The output gap is ubiquitous in Fed comment, so I think they really give it credence. It may be, however, that they think there is a combination of factors at work. Productivity is putting people out of work, but those people are surely putting downward pressure on somebody's wage demands. They aren't outdated machines, so they don't become obsolete.
By the way, Broaddus also said rates will go up someday (now part of the standard Fed text) but that he doesn't know when. Getting mad at the Fed for not telling us whether Dudley or Ryding is right does not make the future any clearer. Fed officials don't know, so they cannot tell us, when rates will go up. Their behavior so far, however, suggests they are shifting from a Dudley view of the market response to a Ryding view of the economy. The "considerable period" wording is coming out.
Posted by: K Harris on November 18, 2003 06:12 AM
Something about the image in this post causes scrolling to slow down a *lot* using both Internet Explorer and Mozilla Firebird. The fact that the same image is posted twice on the current page gets irritating for us frequent readers.
If the case isnt clear one way or the other, the cynic in me says that Greenspan the Republican will hold off on a rate increase. That's because it really does matter, particularly given the high levels of consumer debt and the important role of the interest sensitive housing sector(refinancing, home equity loans)in supporting consumer spending.
Posted by: steve kyle on November 18, 2003 07:38 AMSteve,
Though the record suggests otherwise, Blinder has reportedly told a closed meeting that Greenspan will leave rates unchanged next year due to the election. Blinder may be a bit partisan, and calling the kettle black. We'll see.
Posted by: K Harris on November 18, 2003 10:33 AMI had lunch with Cathy Minehan, President of the
Boston Fed, last week and she made one very interesting point. She said with rates at 1%, there is some higher rate -- 2% -- 2.5% (???)
that the Fed could raise rate to without it being a tightening. With her non-economic background this is not the type of idea she would come up with herself. Poole at San Fran has made the same point. One thing that has bothered the Fed since the start of this cycle when they
expected a normal recovery is that funds are so low they do not have any ammunitation to ease if it is needed.
But I think this throws a third variable into the Fed decision making process.
I agree with Harris, the Fed does not know now what they will do next year.
If econ is strong Fed will tighten at some point next year. But that is still not a sure thing.
Presidential election years have never had any impact on Fed policy. In the election years that
conditions called for a tightening -- rising inflation , falling unemployment rate --the Fed tightened.
James L. Kichline ran the Fed's econ research department for a couple of decades in the 1970-1980 timeframe before going into "private practice" at Miller, Anderson & Sherrerd.
Based on Kichline's research at the Fed, MA&S used trends in capacity utilization as the cornerstone of their economic forecasting . . . . . . . except that while trends in capacity utilization were extremely useful tools in the 1950s, 60s, 70s and some of the 1980s, from about 1990 on capacity utilization in the US had no value as an economic statistic.
To my knowledge, Kichline & crew never came up with a conclusive rationale for the that failure of capacity utilization - it could have reflected problems in measuring productive capacity in an increasingly service oriented economy, problems with globalization and trade flows or something else entirely.
Posted by: Anarchus on November 18, 2003 05:33 PMIn many ways industrial production now only represents high teck side of economy as that only
part of ind. prod. to experience significant growth.
For much of econ, now that so much of demand for goods filled by imports the old tight links between consumer spending and ind.prod. have been broken. If retailers have an inventory problem, or surge in demand, most of reaction now shows up in imports rather than ind. prod.
Example of how this works. Use to be that a great signal of relative performance of retail stocks was ratio of real retail sales to industrial production. Retailers outperformed when real retail sales were stronger than industrial production. Normally, a sign that econ in early stage of the cycle. But since 1990 that old signal has quit working.
Part of problem is that old idea of a normal 4 year business cycle and stock market cycle that
much of analysis built on - including cap. use. data -- no longer valid.
I just did a quick calculation.
industrial production: high tech all other
Jan 1980 = 100.0 100.0
Jul 2003 7,824.4 127.4
means growth rate of all other around 1.5%
Posted by: SPENCER on November 19, 2003 07:55 AM