How slow should they go?

?

Smooth operator

Most central banks smooth their rate changes. Thus, In a speech in 2004, then-Governor Bernanke explained: “as a general rule, the Federal Reserve tends to adjust interest rates incrementally, in a series of small or moderate steps in the same direction.” He contrasted this with an alternative “bang-bang solution, or what I will refer to today as the "cold turkey" approach. Under a cold turkey strategy, at each policy meeting the Federal Open Market Committee (FOMC) would make its best guess about where it ultimately wants the funds rate to be and would move to that rate in a single step.”

?

Over the years, Bernanke and others have offered several arguments for gradualism. First, if policy makers are uncertain about the direction of the economy (and about the accuracy of the data) they may want to move slowly. Second, since the markets are forward looking the Fed can influence the markets by signaling likely future policy changes as well as changing the current funds rate. Third, gradualism reduces the risks to financial stability. Finally, flip-flopping is bad for the Fed’s credibility.

?

Under Greenspan, Bernanke and Yellen the Fed practiced what it preached (Chart). According to Bernanke a good “representative” example is the June 1989 to September 1992 cycle, with 24 cuts totaling 675 bp, including 21 cuts of 25 bp and just three of 50 bp. Indeed, following Bernanke’s speech the Fed did an even smoother cycle, hiking 25 bp for seventeen meetings in a row from 2004 to 2006.

?

This gradualism has been incorporated into formal policy rules for the Fed. When Taylor invented his “rule” in 1993 it was meant to be a rough description of how the Fed had been behaving under Greenspan. The Fed would set the funds rate equal to the inflation target (2%) plus the neutral real rate(also 2%) and then lean against inflation and output “gaps.” However, many economists showed that the model does better if there is a “partial adjustment” term, taking into account the gradual adjustment of the Fed to changing fundamentals. A good academic piece on this partial adjustment is English, Nelsen and Sack (2002). (See the appendix to this post for examples of the various rules and what they are currently forecasting.)

??

Lighting a fire under the Fed

The Fed has gotten a bit less gradualist over time. The first change was in the 2001-2003 cutting cycle. At the time the Fed was very worried about hitting the zero (or slightly negative) lower bound for interest rates. Hence Bernanke and others argued that moving rates down quickly could help make up for limits on how far they could cut. This justified a somewhat faster cutting cycle, with four 25 bp cuts and nine 50 bp cuts. However, this was unique to the lower bound problem. As I noted above, that faster approach did not apply to the hiking cycle in 2004-6 since the zero lower bound problem does not apply in a hiking cycle.

?

The Powell Paradigm

Under Powell fundamental changes in the Fed’s framework have made it harder to pin down the policy “rule,” including the degree of gradualism. Under the 2020 framework, the Fed promised a nonlinear approach, keeping the funds rate flat and then hiking only after inflation was clearly overshooting the target. While Powell et al offered few details on what the hiking cycle would look like, they seemed to believe that with a very flat and stable Phillips curve and a willingness to (mildly) overshoot the inflation target, the Fed could hike in a gradual fashion. Hence the framework was meant to incorporate both delayed hiking and gradualism when the hikes started.

?

In practice, the Fed's new framework seems like a "stop-and-go" approach.

·????? The funds rate was kept near zero from the fall of 2021 until March 2022 even as core and trimmed mean inflation surged and the unemployment rate dropped rapidly toward estimates of full employment.

·????? In March, 2022 the Fed briefly dusted off the old playbook, rejecting calls for a 50 bp hike and going the “normal” 25 bp.

·????? Then the Fed did one of the most aggressive hiking cycles in history, including four hikes of 75bp hikes, two of 50 pb and five of 25 bp.

·????? Next, the Fed went on hold for eight meetings in a row.

·????? Finally, the Fed opened the current easing cycle with a jumbo 50 bp hike.

?

Inquiring minds want to know

Fed watchers have been scrambling to pin down the Fed’s new rule, including the speed of adjustment. In the run-up to the September meeting, some economists and policy makers argued "cold turkey" would win out. They argued that if the Fed “knows where it has to go,” why not go quickly? After all, with a neutral funds rate of 2.5-3.0% and an actual funds rate of 5.25-5.50% what could go wrong? Others, like myself, leaned toward a more gradualist approach—not as slow as Greenspan, but not as fast as in 2022. It seemed to me that 25 bp was more likely than 50 bp.

Still a good case for gradualism

I was wrong. However, even with the Fed’s 50 bp cut in September, I expect the Fed to lean more toward gradualism than cold turkey in the months ahead. Recent developments are strong reminders to the Fed of why gradualism was adopted in the first place. First and foremost, there have been huge swings in the macro data. Inflation fell sharply in the second half of last year, rebound sharply in the first four months of this year, and then seems to have softened. The latest report out this morning showed core inflation of 3.1%, but the more reliable trimmed mean up 2.5%. It is not clear whether they have “won the war” on inflation.

?

There have been even bigger head fakes from the growth data. Going into the September meeting the employment report was weakening and there was an announcement of a likely large downward benchmark revision to payrolls. By contrast, after the meeting not only has the labor market data improved, but there have been both positive surprises and upward revisions to the GDP accounts. Is the economy weakening or not?

?

A second, and related, reminder of forecast uncertainty has been the sharp differences in signals from the labor market data and the GDP data. In announcing the 50 bp cut the Fed underscored the dominant role of the labor market in their decision, pointing to slowing payrolls and a rising unemployment rate. However, as I pointed out in several posts, the GDP data were telling a very different story, with strong and steady growth and signs of a significant output gap. Is the economy still overheating or is it not??

Third, there is a lot of uncertainty about the “star” variables the Fed uses to guide its ship. Most important, the case for quick cuts is very dependent on the idea that the current funds rate is well above its r-star neutral level. I’m skeptical. R-star is very hard to pin down and econometric models suggest very big standard errors. Most important, current estimates of r-star are heavily impacted by the abnormally weak recovery following the Great Financial Crisis. In my mind, r-star is both higher and a lot more uncertain than the consensus seems to assume. Both financial markets and the growth data are suggesting that they can handle a relatively high funds rate. I think a reasonable confidence band for r-star is 3% to 5%.

?

The Fed's journey into the dark Antarctic night

Looking forward I think the Fed should and will trim its sails as it navigates through fog and icebergs with inaccurate maps and faulty instruments (I'm currently reading a book on the early exploration of the Antarctic). At this stage, I see them leaning toward more traditional gradualism. I expect three 25 bp cuts between now and the Spring, with the Fed likely skipping a meeting along the way. Rising estimates of r-star and a falling funds rate should converge next year at about 4%.

?

Appendix: Skip this section unless you are fluent in Greek

The table below is from the Cleveland Fed, showing a range of policy rules. Under the simple Taylor Rule the Fed immediately responds to the inflation and output “gaps). However, under a more realistic Inertia Rule, the Fed moves gradually. Hence the model includes the lagged funds rate. Looking at data prior to the Great Financial Crisis, the adjustment speed in these models tended to be quite slow.?

The table below shows the results when the Fed economists plug in dat from the Survey of Professional Forecasters into these models (Table below). For the 1993 Taylor rule they find that the Fed should have immediately cut rates to 3.6% and then followed with partial cuts. For the Inertia Rule the Fed should cut to 5.09% and then cut 25 bp at every other meeting. It is important to note that with alternative growth and inflation forecasts the range of outcomes is even wider than this table suggests.

?