After the latest inflation data, headline numbers were still running hot. Yet markets moved on the idea that the Fed might be softening — not because inflation fell, but because the specific metric highlighted in the commentary implied a different story. That metric was trimmed mean PCE, and the trimmed mean PCE reading gave policymakers considerably more room than the headline figure.
The trimmed mean strips out outliers — the biggest price increases and the biggest decreases — and averages what remains. The argument for using it is that it filters noise. The problem is that deciding what counts as noise is itself a choice. Different trimming methods produce different numbers. Different numbers support different conclusions. And the metric that gets elevated tends to be the one that gives policymakers the most flexibility at that moment.
This is the structural feature of central bank communication that investors most often underestimate. It isn’t that the Fed misrepresents data. It’s that there are enough legitimate inflation metrics to support almost any narrative — and the one spotlighted shifts with the policy need. Investors who build positions around “the inflation number” without tracking which inflation number the Fed is currently watching are playing a game where the rules can change without announcement.
The practical implication becomes clearest during drawdown periods, when monthly performance targets feel like pressure and every macro release looks like a signal. That’s precisely when reacting to the Fed’s current framework as if it’s permanent becomes most dangerous. Inflation metrics will keep evolving. The preferred measurement will shift again. A portfolio thesis built on a specific metric’s trajectory isn’t a thesis — it’s a leveraged bet on the Fed’s current ruler.
The measurement changes. Your framework shouldn’t.
Central banks don’t move the goalposts. They move the field.
Not advice. Just how I see it.