The Fed is about to raise rates on an inflation it can't fix

The number due to decide next week's Federal Reserve meeting is not the one most investors watched. American consumer prices rose 3.4% in August from a year earlier, steady against July and in line with forecasts. Yet markets moved anyway, pricing a roughly seven-in-ten chance that the committee raises rates on September 16th — the first hike in more than three years. To read that as mere data-following is to miss the strangeness of the situation. The Fed is preparing to raise rates in response to an inflation that the only tool it owns cannot much affect.

The strangeness is visible inside the report itself. Headline inflation of 3.4% sounds stubbornly high, more than half a point above target. But split the number and the story changes. Gasoline prices jumped 3.9% in a single month and stood 27% higher than a year earlier, pushed up by the Middle East conflict and Brent crude above $100 a barrel; fuel accounted for more than a third of August's entirely ordinary monthly rise of 0.4%. Shelter — the weightiest component and the one most sensitive to interest rates — cooled to 3.0%, its increase easing from July. Strip out energy and food, and the "core" index that the Fed treats as its window on underlying demand ran at 2.4%, the lowest reading since March 2021 and close enough to target to matter.

That is the discomfort at the heart of the decision. The inflation driving the hike is not the inflation a rate hike can reach. Higher borrowing costs work by suppressing demand — cooling the housing market, the durable-goods purchases, the wage-sensitive spending that core inflation measures. They do little to a barrel of crude that rises because a war has interrupted shipping, or to a chip that costs more because every cloud provider is bidding for the same scarce supply. Monetary policy chases demand; the August headline is largely a supply shock wearing a demand number's clothes.

So why raise rates at all? The answer is not economics but credibility. Inflation has run above the Fed's 2% goal for five and a half years, long enough that a central bank starts to worry not about the current reading but about whether households and firms still expect low inflation in the future. Once expectations de-anchor, disinflation becomes far costlier. The committee's behaviour shows the anxiety: at its July meeting it voted 9-3 to hold, with three officials openly dissenting for a hike, and in June nine of its members' projections pointed to at least one increase in 2026. The chair, Kevin Warsh, has emphasised that inflation remains too high. A hike now is, in large part, a way of proving the resolve that the pause was beginning to call into question.

The honest trade-off, then, is not between virtue and vice. It is between two costs. Tighten into a war-driven oil spike and the Fed risks cracking a labour market that, for all the gloom, added 162,000 jobs in August with wage growth running at around 3.5% — an overtightening that would compress demand precisely when supply cannot respond. Do nothing and it risks the slower erosion of its own anchor. A committee that chooses the first cost is not wrong; it is merely buying insurance with money the economy may have to pay out later.

For an investor, the useful lesson is that not every hike is created equal, and the distinction is the mechanism, not the announcement. A rise meant to check a genuine demand boom is categorically bad for stocks and bonds: it slows earnings and lifts the rate at which future cash flows are discounted. A rise meant to defend credibility against a supply shock is different. It may leave valuations largely intact, but it trades a defensible thesis about inflation for a live risk that policy has gone too far — and the damage, if it comes, shows up first in the labour market and only later in the bond market.

The second lesson is about what is already in the price. Markets have not just contemplated a hike; they have priced a strong chance of one, and futures imply the Fed will deliver one, possibly two, before the year is out. That pricing, more than the report, is the real event risk. For the hike itself to hurt, the committee would have to surprise decisively — by raising more than a quarter point, or by signalling a sustained campaign. The reverse surprise, a hold that contradicts the hawks, would be a repricing in the opposite direction. The August report settled nothing in the Fed's boardroom; it merely confirmed that the choice is real.

The larger point is institutional. A Fed that raises rates to fight a war-driven oil price is admitting that its tool cannot fix the thing the market is watching, and that its credibility has become a separate, fragile asset worth defending at some genuine economic cost. There is a worse version of this admission — the one where the Fed, having swung from cutting rates through last year to holding and now to threatening the opposite, discovers that markets have stopped believing its projections at all. That is the failure the committee is really manoeuvring against. The hike, if it comes, will be the first in three years; the question it answers is whether the Fed's word still counts for more than the price of a barrel of oil.