A forecasting group just predicted 4.2% U.S. inflation for 2026. The market sold off. Treasury yields spiked. Talking heads called it a regime change. One problem: a forecast is not a print.
The actual CPI data tells a different story. February 2026 core CPI came in at 3.1% year-over-year. The three-month annualized rate sits at 2.8%. The gap between the forecast (4.2%) and reality (3.1%) is 110 basis points. That is not confirmation. That is a bet against the data.
Forecasting groups have a specific track record on inflation calls. Over the past five years, the median forecast error for one-year-ahead CPI predictions has been 0.9 percentage points, per the Federal Reserve Bank of Philadelphia's Survey of Professional Forecasters. A 4.2% call with that error bar means the actual range is 3.3% to 5.1%. The market traded the midpoint like it was a certainty.
The Fed's own projection — the March 2026 SEP — puts PCE inflation at 2.6% for year-end 2026. The FOMC does not project 4.2%. The bond market does not price 4.2%. Fed funds futures imply a terminal rate consistent with 2.8-3.0% inflation, not 4.2%.
Here is the number that reframes the panic: the 5-year, 5-year forward inflation expectation — the Fed's preferred gauge of long-run anchoring — sits at 2.21%. That is 11 basis points above its 10-year average. Not broken. Not unanchored. Slightly elevated.
The trade implication is clean. If you sell equities on a forecast that is 110 basis points above the actual data, you are pricing a scenario the bond market does not believe and the Fed does not project. That is not risk management. That is narrative trading.
Kill criteria: if the next two CPI prints (March and April 2026) come in above 3.5% core, the forecast gains credibility and the inflation re-acceleration thesis is live. Until then, one group's model output is not monetary policy. It is an opinion with a decimal point.