Three hot months: what the data now says

January's hot inflation reading could be put down to seasonal quirks. March's cannot. The consumer price index rose 0.4 per cent on the month for the third print in a row, and the annualized pace of core inflation is back above four per cent. The odds on a smooth return to target have fallen sharply.

Three months make a trend

One hot month is noise. Three are harder to dismiss. Core services inflation outside housing is picking up, driven by double-digit rises in insurance premiums and by higher medical costs. The disinflation of late 2023 has run its course, and the Federal Reserve will probably have to keep policy tight for longer to anchor expectations.

What the arithmetic implies

Monthly figures compound. A core increase of 0.4 per cent sustained for a year works out at close to 4.9 per cent, more than double the Fed's 2 per cent objective. To hit the target, monthly core readings need to average a little under 0.2 per cent, and even a steady run at 0.2 per cent would leave annual inflation around 2.4 per cent. The gap between what the economy is producing and what the Fed wants is therefore about half of every monthly print, which is too large to be closed by rounding.

The statistical case for treating the quarter as signal rests on independence. Seasonal adjustment problems tend to show up in one month and wash out in the next, as happened with January readings in several recent years when firms reset prices for the new year. Three consecutive surprises in the same direction are much less likely to be a quirk of one month's adjustment factors. A forecaster who started the year believing in a smooth glide path has to shift that belief a long way after three contrary observations, even if no single observation would have been decisive.

Where the heat is coming from

The detail points to services that are slow to cool. Motor insurance premiums have been rising at their fastest pace in decades, as insurers pass on the higher cost of repairing modern cars and replacing vehicles whose prices jumped after the pandemic. Medical services, which were subdued for much of 2022 and 2023, are catching up as contracts reset. Both reflect earlier price increases working their way through long chains of contracts, which means they fade slowly rather than reversing in a month or two.

Shelter is the other stubborn piece. Owners' equivalent rent, the imputed cost of living in one's own home, is the largest single item in the core index and still rises briskly. Market rents on newly signed leases have slowed sharply, and the models used by many economists predicted that the official shelter series would follow with a lag of about a year. The lag is proving longer than hoped, so the item that was supposed to deliver most of this year's disinflation has so far delivered only part of it.

Lessons from the transitory call

The Fed has been wrong in this direction before. Throughout 2021 officials described rising prices as transitory, arguing that supply bottlenecks would unwind, and the word was only dropped towards the end of that year, by which time inflation was far above target. The subsequent tightening, beginning in March 2022, was the fastest in four decades. That episode left the committee wary of reading the data optimistically, and it explains why three hot months carry more weight in the current debate than they might have done in a quieter decade.

The comparison also marks the limits of the worry. The 2021 surge was broad, with used cars, energy and food all rising at once, and inflation expectations drifted up. The current stall is narrower, concentrated in a few services lines, and survey measures of long-term expectations have stayed broadly stable. That points to a slower return to target rather than a fresh acceleration, but slower is still costly for anyone waiting on lower rates.

The argument against overreacting

There is a fair case that the quarter overstates the problem. The Fed targets the personal consumption expenditures index, which gives shelter a smaller weight and draws its medical prices from a different source, and that measure has run noticeably cooler than the consumer price index. Insurance premiums are a catch-up effect that must eventually stop once insurers have repriced their books. A few soft months in the second half could put the year back on track, and a model that weights recent data heavily will swing back just as quickly as it swung here.

The trouble with that argument is timing. Even if the underlying trend is cooler than the headline suggests, the Fed now needs several good months to regain the confidence it had in December, and those months cannot arrive before the summer. Policy decisions depend on the data the committee has, not the data it expects.

The distribution of the pain follows from the composition. Insurance, rent and medical bills are hard to cut back on, so the services inflation now doing the damage falls hardest on lower-income households, which spend a larger share of their budgets on such necessities. Companies selling discretionary goods face the opposite problem, since shoppers squeezed on fixed costs have less left over. For bond investors, the lesson is that inflation risk has not disappeared from the term premium, and longer-dated yields have risen to reflect it.

Readings that would move the odds

The most useful check is the core PCE figure for March, published at the end of April, alongside the quarterly employment cost index, which shows whether labour costs are still running hot. The April consumer price report in mid-May gives the next monthly read. Two consecutive core prints at 0.2 per cent or lower would restore much of the soft-landing case, while another 0.4 per cent would make a cut before the autumn implausible.

The hope that inflation would fall back painlessly did not survive the first quarter. Bringing prices under control now looks likely to require slower growth and a longer spell of high rates, and borrowers will bear both. On present evidence the Fed will not have the three or four clean months it needs before September, so households and companies with floating-rate debt should plan on paying today's rates well into the second half of the year.