Should the Fed Hike?
Today's post runs through key metrics on whether the Fed should hike - it should not!
Thank you for subscribing to my posts. If you’re not yet a paying subscriber, please consider becoming one. That’ll allow you to DM me with questions and comments. I plan to move my posts behind a paywall in coming months. Subscribing now locks you in at a low rate. Thank you so much for your support and feedback!
Today’s post runs through how I think people inside the Fed are looking at inflation and whether to hike or not. I do a step-by-step analysis of actual inflation data, what inflation expectations are doing and whether they’re becoming de-anchored, and end with a discussion of the Philipps curve, an empirical relationship that describes how much the unemployment rate would need to rise to bring inflation down by a certain amount. The bottom line is that - as of today - there is no reason for the Fed to hike, in contrast to market pricing, which continues to be for more than one hike in 2026.
Actual inflation: there’s always shocks hitting the economy that’ll move prices for certain things one way or another. If these “relative” price shocks start to broaden out and look like generalized inflation, that’d be a reason for the Fed to hike. This is what my inflation generalization indices look at. The black dotted line in the chart above is the combined weight of items in the PCE - the Fed’s preferred inflation gauge, with month-over-month inflation above two percent on an annualized, seasonally-adjusted basis. The blue line is the same thing for year-over-year inflation. I focus less on the latter because it lags the former. The month-over-month generalization index was around 60 percent in May. That’s at the lower end of things in recent years, so there’s no evidence that the oil shock and AI-linked price hikes are translating into broad-based inflation.
Inflation expectations: the chart above shows survey-based inflation expectations over the next year (black line) and for the next five years (blue line). The latter rose sharply in May, but reversed that jump almost entirely in June amid the fall in oil prices. There’s thus no sign that longer-term inflation expectations are getting de-anchored. The same is true for market-based measures. The blue line in the chart below shows 2-year break-even inflation, which is down sharply in recent months as oil prices have tumbled. There’s no case for hikes based on any kind of worry about inflation expectations.
The Phillips curve: the plot below shows the relationship between core PCE inflation (vertical axis) and the unemployment gap (horizontal axis), which is the difference between the actual unemployment rate and what the Congressional Budget Office (CBO) says is full employment. If this gap is negative, it signals a tight labor market. I’ve highlighted in red the post-COVID inflation shock and in black the current oil shock. The blue dots are the rest of the data and suggest that the Phillips curve is very flat. This means - if you want to bring core (currently around 3.5 percent) down to two percent - the unemployment rate would have to rise by five or six percentage points to make that happen, i.e. it’d take a very deep recession. I can’t think of anyone who’d advocate for that, especially since we know inflation is currently elevated due to tariff and oil shocks.
Market pricing: markets price around 40 bps in hikes through end-2026 as the chart below shows. Most likely, this reflects the heavy emphasis on price stability in the first press conference, which was largely performative in my opinion, and the move up in the “dots,” which are likely stale because oil prices were falling so fast at the time. Given how flat the Phillips curve is, this amount of tightening would do almost nothing to bring inflation down. It’d take a much bigger hiking cycle to drive unemployment up sufficiently. In other words, there’s no economic justification for hiking 40 basis points at this point.
There’s really no economic rationale for why the Fed should hike, but markets are sticking to their guns. The most common explanation I hear is that this is about Warsh establishing his credibility by hiking in 2026. Once he’s done that, he can then cut in 2027. I’m not a big fan of that argument. There’s no crisis in Fed credibility that needs fixing, as the inflation expectations data I show above document. Next week should reveal a sharp slowing in CPI inflation in the June print that comes out on July 14. My best guess is that this will end markets’ flirtation with rate hikes.
Excellent analysis, thank you!
Does the Fed really dynamically weight the PCE component of the inflation number they look at? That is a surprise to me (physicist, software engineer, not economist).