The Fed Looks Through a Great Inflation Print
After being burned by the unwelcome inflation surprises of the first few months of 2024, they are now justifiably reluctant to be fooled again, even if that means ignoring some encouraging numbers.
The U.S. Consumer Price Index (CPI) rose by just 0.006% between April and May on a seasonally-adjusted basis. That is the smallest monthly increase since July 2022, and before that, March-May 2020. Encouragingly, the most recent number is not a correction following an energy price spike, as in July 2022, nor does it reflect a broad-based collapse in spending on everything from hotels to schools, as in March-May 2020.
Moreover, the disinflation is broad-based. While a few categories experienced notable price declines, including airfares, “subscription and rental of video and video games”, computer software, and gasoline, each of which fell by almost 4%, the really notable feature of the latest data is that there were so few categories that rose by very much, if at all.1
Narrower cuts of the numbers, which attempt to strip out unusually volatile or idiosyncratic components, tell a similar story. The CPI excluding food, energy, used vehicles, and “shelter”2 fell slightly (by 0.07%), while the Federal Reserve Bank of Atlanta’s measure of inflation that only tracks categories where prices change infrequently also went negative on a monthly basis. The Federal Reserve Bank of Cleveland’s “trimmed mean”, which excludes the 16% of components (by weight) with the largest price changes, had its smallest monthly increase since January 2021. Similarly, the prices charged by most American businesses to end customers as tracked in the Producer Price Index (PPI) also rose by the smallest amount on a seasonally-adjusted basis (0.01%) since February-April 2020, when prices were falling.3
Yet this does not seem to be having an impact on the thinking of Federal Reserve officials.
In their latest Summary of Economic Projections, forecasts of short-term interest rates at the end of this year and at the end of 2025 “under appropriate monetary policy” were marked up compared to the projections published in March. At the same time, projections for inflation “under appropriate monetary policy” were also marked up (slightly), while projections for real growth were unchanged. Apparently, the past three months have made Fed officials slightly less confident that inflation will decelerate to the target rate of 2% a year on its own.
These projections were initially submitted weeks ago as part of the preparatory process for the most recent meeting of the Fed’s Open Market Committee (FOMC), which sets monetary policy. That was before anyone had seen the inflation data for May—or the most recent employment and wage data. In theory, FOMC members have the right to revise or update their projections up through the day they are released. But Fed boss Jerome Powell told journalists at the post-meeting press conference that, despite being reminded of this option, almost no one chose to exercise it.4 While the May inflation numbers were good, they were not good enough to change most officials’ minds.
There are two plausible (and complementary) explanations:
Fed officials believe that the interest rate projections from March were so far off that adjustments would have been required even with the most recent inflation data.
Officials believe that May’s CPI print was a one-off and unlikely to be that predictive of future inflation.
The simplest argument in favor of point 1 is that market pricing has changed a lot since the start of the year, especially since the release of the January CPI report on February 13 and then again with the March CPI release on April 10. Rightly or wrongly, Fed officials want to stay relatively close to where traders are betting—and the latest inflation data has not yet caused market prices to reverse the cumulative impact of the February-April moves.
Back in March, officials were close to, but slightly above, the market-implied forecasts of the time. For example, about 26% of officials (5 out of 19) thought that that the target short-term interest rate band should (probably) be 4.75%-5.0%, which was almost identical to the probability implied by interest rate futures and options in the days leading up to the March 20 FOMC meeting. Similarly, 9 out of 19 officials (47%) thought that the target band would end up being 4.5%-4.75% at the end of the year, which was not far off from the market-implied probability of 35%. The difference was that the remaining 5 officials skewed relatively hawkish compared to the market-implied probability distribution, with 4 of the 5 projecting that short-term interest rates would be above 5% “under appropriate monetary policy” and only 1 official projecting short rates below 4.5%.
Since then, market pricing has shifted dramatically. The implied probability of short rates being below 4.75% by the end of 2024 plunged from 60% in early April to 20% on April 10—and has stayed around or below 20% ever since, even with the latest inflation data. (As of this writing, the implied probability is around 26%.) Fed officials have adjusted their own views by much less, but in the same direction.
As of now, there are 4 of 19 officials who think rates should probably hold steady, 7 of 19 think it will probably be appropriate to lower rates by 25 basis points to 5.0%-5.25% by the end of the year, and the remaining 8 officials think it might make more sense to lower rates by 50 basis points to 4.75%-5.0%. Fed officials are slightly more biased to higher rates than what is implied by market prices as of today, but they are perfectly aligned with what was priced in at the end of May, when they initially submitted their projections.
But if Fed officials are more or less just following market pricing, which in turn is responding (mostly) to hard data releases, the question then becomes: why did the May CPI and PPI numbers not move markets as much as the January and March data?