Inflation is mostly temporary – according to the gilt market. The five-year breakeven inflation rate (the gap between conventional and index-linked yields) is now 4.7 percentage points. That compares to a current retail price index (RPI) inflation rate of 9 per cent, implying that markets are pricing in a big drop in inflation. In fact, by comparing three-, four- and five-year breakeven inflation rates we can calculate the implicit breakeven inflation rates for the 12 months to May 2026 and May 2027. For both years it is 3.7 per cent. If we allow for the fact that RPI inflation is typically above consumer price index (CPI) inflation (by an average of 0.9 percentage points a year in the last 10 years) this implies that the market expects CPI inflation to fall below 3 per cent. That’s above the Bank’s target, and above the Bank’s own forecasts, but no great problem.
Which poses the question: what does the gilt market know about future inflation anyway?
My chart helps answer this. It plots actual five-year annualised RPI inflation against the five-year breakeven inflation rate five years previously. It shows that there have only been two occasions when the market got inflation badly wrong.
One was in the 1990s when inflation was significantly lower than expected. Initially this was because the market under-estimated the disinflationary effect of the recession of the early 1990s, but then in the late 1990s it under-estimated the degree to which cheap imports from China would reduce inflation.
The other was during the financial crisis in late 2008 when the market feared a deflation which never actually materialised.
On average, though, inflation has been close to the breakeven inflation rate for much of this century. Since the Bank of England was given operational independence in May 1997, inflation has on average been only 0.2 percentage points above the breakeven rate.
Which seems impressive.
Except that the gilt market has been very wrong recently. In March 2017 it was pricing in annual RPI inflation for March 2022 of 2.8 per cent. It turned out to be 9 per cent. This is the biggest forecast error since records began in 1985. To some extent this is forgivable; nobody could have foreseen then that utility bills would jump by over 20 per cent in 2021, with much more to come this year.
What’s more curious is that inflation has, on average, overshot breakeven inflation at all. In theory, it shouldn’t. Yields on index-linked gilts are usually lower than those on conventionals not just because investors expect inflation but also because they want protection from the danger of unexpected inflation. Breakeven inflation rates thus comprise both a pure inflation expectation and an inflation risk premium. Which means that if inflation risk doesn’t materialise then actual inflation should be consistently above prior breakeven inflation rates.
Which it often has not been.
One reason for this is that commodity prices have sometimes exceeded expectations. There’s a significant correlation between changes in these and the gilt market’s inflation errors. When commodity prices rise a lot, such as in the mid-2000s, actual inflation exceeds market expectations. And when they fall (such as in in the mid-2010s, inflation falls below expectations.
Also, the market under-estimates the extent to which inflation rises early in economic recoveries because of mismatches between the patterns of supply and demand. In 2010-11, for example, there was demand for software engineers but excess supply of construction workers. And now there are shortages of lorry drivers but unemployed car mechanics. Such mismatches, rather than excess demand generally, can push up inflation.
But they tend to fade away as supply and demand eventually respond to price signals: market forces often work, if not as quickly as people think.
Which gives us a reason to believe the gilt market’s current optimism about inflation. If those mismatches fade away then not only will inflation fall, but so too will one historic cause of the market’s under-prediction of it.
This leaves the main threat to its inflation expectations being the possibility of another commodity price spike. We cannot quantify the danger of this because (as today’s unexpectedly high inflation shows) markets cannot foresee wars or political disruptions.
What we can say though is that barring such risks the gilt market’s forecasting record – while not perfect – is not enough that we should attach credence to its belief that inflation will fall sharply. Markets aren’t perfect, and the future is in large part unknowable by any single mind, but there is sometimes wisdom in crowds.
