Ugly and complicated CPI print this morning -- we had been expecting a beat, and we got a big miss. We can't blame the annual changes in basket weights -- they were trivial. We've been gradually giving up on our monetarist model that predicts CPI based on changes in the money supply, judging that its signal-to-noise ratio has been degrading as inflation has normalized. But its prediction for January CPI was literally perfect, and looking ahead, it's predicting that CPI will keep gently rising over the coming year. Markets are now saying there isn't certainty of a Fed rate cut until December.
What you're not hearing about where inflation is heading, and what it means
Friday, February 21, 2025
Our monetarist model perfectly predicted January's CPI miss, and it sees more misses to come.
Update to Strategic View
January CPI came in much hotter than market expectations, and our expectations as well. But our monetarist model based on M2 growth predicted it perfectly. The model precisely predicted the inflation of 2021 and 2022 when Powell said it would be "transitory," and the disinflation to follow, when Powell said it wouldn't. But then the model predicted an outright deflation that never came. We recalibrated the model about a year ago, but the experience shook us and we have not relied on it. We should have. In January it predicted core CPI ex-OER within 3 bp, and across a broad range of other inflation indicators including Powell's "supercore," the average error was just 3 bp. The model is now calling for inflation to gradually rise over the coming year, with core CPI ex-OER, now at 2.36%, to rise to 3.7%. It this turns out to be true, there will be no more rate cuts from the Fed, and we would have to even start considering the possibility of a rate hike.