Latest / The Federal Reserve Podcast with Fexingo: Interest Rates, FOMC Meetings, and Monetary Policy / How the Fed Reads Inflation Data It Admits It Cant Forecast
Transcript
- Lucas: So the May core PCE reading came out yesterday — three point four percent year-over-year. That is the highest print since October 2023, and it is notably above what the Fed's own dot plot projected just a few weeks ago. Luna: Right, the June Summary of Economic Projections had core PCE ending this year at two point five percent. We are already overshooting that by nearly a full point with seven months to go. Lucas: Exactly. And this isn't a one-month blip. The three-month annualized core PCE rate is running around three point seven percent. The six-month is near three point five. This is a trend, not a surprise. Luna: So why does the Fed keep missing? I mean, these are the people with the best models and the most data in the world. Lucas: It's a great question. I think part of the answer is that the composition of inflation has shifted. During 2022 and 2023, the main drivers were goods — supply chains, energy, food. Those are relatively easier to model because they respond to global prices and shipping costs. Lucas: Now the stickiness is coming from services, specifically services excluding housing. That category includes things like portfolio management fees, auto insurance, medical care, and recreation. These prices are driven by domestic labor costs and administered pricing — much harder to forecast with standard models. Luna: And portfolio management fees — that's tied to asset prices, right? When the stock market goes up, those fees tend to rise with a lag. Lucas: Exactly. The S&P 500 is up about twelve percent year-to-date through yesterday. That feeds into wealth management fees, which show up in the PCE services category. The Fed's models probably didn't fully capture that linkage. Luna: And auto insurance is another one. Premiums are up something like fifteen percent year-over-year. That's partly because repair costs and replacement parts are still elevated from the supply chain mess, but also because insurers are playing catch-up after years of under-pricing risk. Lucas: Right. So the Fed is in a tough spot. If they tighten further to fight this services inflation, they risk crushing the parts of the economy that are actually slowing — like manufacturing and small business investment. But if they ease too soon, they lock in this three-point-five percent core inflation floor. Luna: And the market is already pricing in a cut in September, maybe another in December. But yesterday's PCE number — three point four — that makes the September cut look less certain, doesn't it? Lucas: It does. The CME FedWatch tool showed the probability of a September cut drop from about sixty percent to forty-five percent after the release. The bond market is also repricing — the two-year yield ticked up six basis points yesterday to four point one four percent. Luna: So the market is essentially saying: the Fed is going to be more patient. But the Fed itself has been saying it needs to be data-dependent. And the data keeps surprising to the upside. Lucas: That's the paradox. The Fed is data-dependent, but its own forecasts are consistently wrong. That suggests either the models are flawed, or the data itself is being revised in ways that make forecasting even harder. Luna: Speaking of revisions — the Bureau of Economic Analysis revised up the Q1 core PCE reading from three point six to three point seven percent annualized in last week's GDP revision. So the starting point keeps getting higher. Lucas: Yeah. And that makes the Fed's path even more complicated. If the starting level is higher, even if inflation decelerates from here, the year-end print will be above the SEP projection. Luna: So what should the Fed actually watch? If the usual models aren't working, what's a better signal? Lucas: I think they need to focus more on wage growth and unit labor costs. Average hourly earnings are running around four percent year-over-year. Productivity growth has been decent — around two percent — so unit labor costs are rising roughly two percent. That's consistent with two percent inflation, not three point four. Lucas: But the disconnect is that services inflation is running hot even though wage growth is moderating. That suggests there's a pricing power story — companies are maintaining margins by passing through higher costs, and in some cases, raising prices more than costs warrant. Luna: So margins are a lagging indicator too. Companies might be reluctant to give up the pricing gains they got during the pandemic. Lucas: Exactly. The Atlanta Fed's business inflation expectations survey shows firms still expect to raise prices by about three percent over the next year. That's down from four percent last year, but still above the Fed's target. Luna: So the Fed is stuck between a rock and a hard place. They want to be done with rate hikes, but the data isn't cooperating. Lucas: And the global backdrop doesn't help. The Strait of Hormuz situation, the Venezuela earthquakes — those are supply-side shocks that could push energy and commodity prices higher, adding to inflation pressure. Luna: But those are transitory, in theory. The Fed has said it looks through supply shocks. But if they keep happening, at what point do they become persistent? Lucas: That's exactly the debate inside the FOMC. The hawks are saying: don't cut until you see sustained evidence that services inflation is cooling. The doves are saying: the economy is slowing, real rates are already restrictive, don't over-tighten. Luna: And the chair — Kevin Warsh — this is his first year. He's got to navigate this without the credibility that Powell had after years of experience. Lucas: Right. Warsh is still building his track record. The market is watching every word. His press conference after the June meeting was notably cautious — he said the committee needs 'several more months of data' before gaining confidence. That's not a strong signal either way. Luna: Speaking of the data, we're looking at a pretty packed calendar next week: ISM manufacturing, JOLTS, the June jobs report. Those will be key inputs for the July meeting. Lucas: Absolutely. If the jobs report shows another month of solid payroll gains — say above two hundred thousand — and wage growth stays around four percent, the hawks will have ammunition to hold rates steady. Luna: But if payrolls come in soft, under one hundred fifty thousand, and the unemployment rate ticks up, the doves will say the labor market is cracking. Lucas: And the wildcard is the inflation expectations data. The University of Michigan survey last week showed one-year expectations at three point three percent, the highest in over a year. If that keeps rising, it becomes a self-fulfilling prophecy. Luna: You know, all of this — the data, the forecasts, the uncertainty — it's exactly why podcasts like this exist. We try to make sense of it for people who don't have time to read every Fed speech. Lucas: Honestly, if walking through the economy with us has made something click for you today, we'd be grateful if you considered supporting the show. It's just a small gesture — the price of a coffee — but it keeps us independent and ad-free. That link is buy me a coffee dot com slash fexingo. Luna: Yeah, and no pressure at all. We're just glad you listen. Every little bit helps us keep doing this. Lucas: So back to that September cut probability — I think the next big test is the July FOMC meeting. They won't cut in July, but the statement language will be crucial. If they remove the phrase 'elevated inflation' and replace it with something softer, that signals a cut in September is still on the table. Luna: And if they keep it, then the market will push expectations further out into December or even 2027. Lucas: Exactly. So we've got about four weeks of data between now and the July thirty-first decision. That's four employment reports, two CPI prints, and one more PCE reading. By then, we'll have a much clearer picture. Luna: And we'll be here to break it down. Thanks for joining us. Lucas: Take care, everyone.