The consensus around Kevin Warsh's Senate hearing was straightforward: here's a guy promising regime change at the Fed who just handed markets a dovish gift. He wants to ditch the standard core PCE — which strips out food and energy by design — for trimmed-mean and trimmed-median gauges that exclude the biggest price swings each month. As of February 2026, the Dallas Fed's trimmed-mean PCE clocked in at 2.3% over 12 months. Bank of America's trimmed-median version sat at 2.8%. Standard core PCE? 3.0%. Warsh called the underlying trend 'quite favorable.' Markets heard rate cuts. Everyone relaxed.
You should not. This switch looks softer right now because the tails it's trimming happen to include the usual volatile suspects. But trimmed metrics don't permanently ban food or energy like core PCE does. They rank every item by how much its price moved that month and chop the extremes symmetrically — often the top and bottom 25% or so. That means moderate but persistent supply-driven spikes in groceries, fuel, or beef can still sneak into the calculation if they're not the absolute wildest movers. Core PCE ignores them outright. Trimmed averages? They judge on the day.
Bank of America economist Aditya Bhave put it plainly: history shows the trimmed-median gauge ran above core PCE in both 2019 and 2020. Had the Fed been staring at those higher readings, policy would have tightened more aggressively. The same dynamic is primed to repeat the moment tariffs bite, geopolitics disrupts shipping, or commodity volatility returns — exactly the 'one-time' shocks Warsh says the Fed should look through. He told senators he's most interested in the underlying rate, not 'the one-time change in prices because of a change in geopolitics or a change in beef.' Yet his preferred method offers no guarantee of ignoring non-extreme supply moves.
Here's the deadpan fact bomb that reframes everything: core PCE excludes food and energy by design. Trimmed PCE excludes whatever happens to be the biggest movers that month — making it agnostic until the month when food or energy is merely the third- or fourth-biggest mover. Suddenly the 'clean' read Warsh wants starts flashing hotter than the metric he's replacing.
Warsh has talked about launching a 'data project' to refine this. Fine. But the mechanical reality is already clear from the Dallas Fed's own series. The trimmed-mean PCE has been trending lower than core lately, creating the illusion of progress. Connect that to the macro backdrop: Trump's tariff agenda is no secret, nor is ongoing geopolitical friction. Those aren't theoretical tails anymore. They're incoming moderate-to-large price pressures that trimmed gauges will partially absorb rather than fully exclude. The result? A Warsh-led Fed staring at readings that force tighter policy than core PCE would have dictated — precisely when markets have priced in easier conditions based on today's flattering numbers.
This isn't abstract macro. It hits capital allocation, business planning, and your portfolio directly. Companies hedging input costs or setting wages will face a Fed that reacts more sharply to supply-side noise under trimmed rules. Rate-cut expectations baked into bonds and equities get repriced higher in yields when those readings spike. The credibility Warsh is trying to rebuild through 'regime change' erodes if the new yardstick delivers exactly the reactive stance he claims to oppose.
Look at the screenshottable stat line for February 2026 data: Dallas Fed trimmed-mean PCE at 2.3% (12-month), BofA trimmed-median at 2.8%, versus core PCE at 3.0%. Today that spread flatters the narrative. In a supply-shock environment, the spread flips and the trimmed versions lead the hawkish charge. BofA's historical backtest confirms it happened before. It will happen again.
Warsh's approach assumes symmetric tail risks that can be neutrally trimmed without bias. Reality doesn't work that way when persistent but non-extreme supply pressures dominate the ranking. The Fed's current framework has flaws, sure, but at least core PCE delivers consistency on volatile categories. Switching to something more sensitive to the middle of the distribution during a tariff-heavy regime is asking for volatility in policy that markets aren't prepared for.
You've been warned. The dovish optics today are real. The hawkish trap tomorrow is baked into the math.