The tricky judgments on when to loosen

Stay informed with free updates

Simply sign up to the US inflation myFT Digest — delivered directly to your inbox.

The course of disinflation never does run smooth. At the end of last year, futures markets had priced in six interest rate cuts for the US in 2024. My own expectations had also become quite optimistic. Yet now, after three successive quarters of stubbornly high inflation, US Federal Reserve chair Jay Powell warns that it is likely to take “longer than expected” for inflation to return to the central bank’s 2 per cent target and justify cuts to interest rates. Market forecasts for rate cuts have duly been transformed. Some suggest they will be postponed to December, partly to avoid cuts before the presidential elections in November. Yet no similar rethinking has emerged in the eurozone: the first cut is still expected to be made in June.

Lessons come from this story. One is the inherent uncertainty of any disinflationary process. Another is the difficulty of reading the data: in this case, a part of the explanation for the robust recent figures for “core” consumer price inflation is “Owners’ Equivalent Rent of Residences”. Yet this is just an imputed figure. It is not clear, as yet, that any fundamental change in the US disinflationary process has occurred. A final lesson is that, while there have clearly been some common factors in the inflationary process across the Atlantic, the US and eurozone economies have been different: the former is far more dynamic.

The latest World Economic Outlook from the IMF provides an illuminating quantitative comparison of the inflationary processes in the US and eurozone, derived from annualised three-month average inflation. Labour market tightness has been far more significant in driving inflation in the US than in the eurozone and, crucially, this continues to be the case. At the same time, “pass-through” effects from higher world prices, notably of energy, were far greater in the eurozone. This has made eurozone inflation more credibly “temporary” than that of the US. This has implications for monetary policy. (See charts.)

Two more pieces of data help elucidate what has been happening. One is on nominal domestic demand. In both the US and the EU, aggregate nominal demand sank far below 2000-2023 trend levels of growth during the pandemic. In the second quarter of 2020, nominal demand was as much as 12 per cent below trend in the US and 14 per cent below trend in the eurozone. By the fourth quarter of 2023, in contrast, it was 8 per cent above trend in the US and 9 per cent above trend in the eurozone (where trend growth was also weaker). This explosive growth in demand in these two crucial economies must have caused supply shocks as well as merely accommodating them. This is the past, however. In the year to the fourth quarter of 2023, nominal demand expanded by only 5 per cent in the US and 4 per cent in the eurozone. The former is still a bit too high, but it is still getting closer to what is needed.

A second relevant piece of data is on money. I remain of the view that these quantities should not be ignored in judging monetary conditions. The pandemic saw not just huge increases in fiscal deficits, but also explosive growth in broad money. In the second quarter of 2020, for example, the ratio of US M2 to GDP was 28 per cent above the 1995-2019 linear trend. By the fourth quarter of 2023 it was back to just 1 per cent higher. For the eurozone, these ratios were 19 per cent and minus 7 per cent, respectively. These numbers show a huge monetary boom and bust. In future, disinflationary pressure might prove excessive.

So, what needs to be done now? In answering that question, the main central bankers have to remind themselves of four crucial points.

The first is that ending up with inflation well below target is, as we have by now learnt, pretty bad, because this risks making monetary policy ineffective. Central banks should act on the assumption that the consequences of being too tight could turn out to be almost as bad as of those being too loose. Moreover, it is not a trivial matter that the former could be particularly damaging for vulnerable debtors worldwide.

A second point is that uncertainty cuts both ways. It is evidently true that demand and so inflation might prove to be too great, especially in the US. But it could also turn out too weak. The policies that would eliminate the mere possibility of the former might make the latter a certainty. Thus, while the aim is rightly to get inflation to target, it makes no sense to pay any price to achieve this objective: it is not infinitely valuable.

A third point is that there are problems created by being determined to eliminate the very possibility of having to change course. If one starts from the assumption that the first interest rate cut must be followed by many more in the same direction, the degree of certainty needed before starting will be too great. The price of waiting until certain is likely to be that of waiting too long.

The last point is that being data-dependent indeed makes sense. But new data matter only if they materially affect forecasts of the future. What matters is not what is happening right now, but what will happen in the months or even years ahead, as past policy works through the system. New information must be seen through that lens. There is good reason to suppose that the recent inflation news in the US is not very significant. Unless the Fed is reasonably confident that it is, it should ignore it.

It is now that decisions start to become really tricky. Two years ago, it was clear that monetary policy had to be tightened: the risk of moving into a high-inflation world were too high. But now it is clear that the ECB should start loosening quite soon. The underlying situation in the US is more evenly balanced. But the Fed, too, cannot wait forever.

martin.wolf@ft.com

Follow Martin Wolf with myFT and on X