(The opinions expressed here are those of the author, a columnist for Reuters.)
ORLANDO, Fla., Oct 20 (Reuters) - "Inflation expectations" are probably two of the most important and scrutinized words in financial market and central bank circles, playing, as they do, a pivotal role in determining monetary policy decisions.
Unfortunately, they may also be two of the most meaningless.
Two papers published by Federal Reserve economists recently have poured cold water on the notion that the inflation outlook according to business, market, household, or economists' expectations is a particularly reliable guide to actual outcomes.
This matters - hugely - because tempering and guiding the public's inflation expectations is a key policy tool for any central bank with a price stability mandate like the Fed. How policymakers respond to changes in the perceived inflation outlook has a real impact on millions of people's daily lives.
But according to a Cleveland Fed paper this week ( bit.ly/3nhPLrI ), the predictive relationship between a range of inflation expectations measures and future inflation is, at best, patchy, and at worst, virtually non-existent.
Consumers, in particular, have a poor record in predicting inflation a year out, and financial markets aren't much better.
This follows a paper last month by Fed staffer Jeremy Rudd ( bit.ly/3G2Cg7J ), who went further, warning that belief in this supposed relationship between expected and actual inflation "has no compelling theoretical or empirical basis and could potentially result in serious policy errors".
Well-anchored expectations help ensure that the economy runs smoothly, allowing businesses and consumers to make rational spending and investment decisions. A benign environment like this in turn helps the Fed meet its inflation targets, completing the virtuous circle.
But the Fed is in a quandary right now, and markets in a frenzy, as they grapple with the strongest price pressures in years. Financial market inflation expectations as measured by inflation swaps and breakeven rates are the highest since 2014.
There is a growing view that the Fed will feel compelled to tighten policy earlier and more aggressively than it would like to ensure inflation remains transitory. Many emerging market central banks, most notably Brazil's, are already well down this path, and those in developed economies are getting twitchy too.
The clear risk is that if current inflation pressures do prove to be "transitory", as Fed policymakers still claim, then raising rates could slow economic activity, choke growth and maybe even tip the economy back toward recession.
Can policymakers put their faith in these expectations, and if so, which ones?
The Cleveland Fed paper analyzes inflation expectations from four main groups: households (University of Michigan 1-year consumer surveys); professional economists (Blue Chip Economic Indicators 1-year forecasts); businesses (Atlanta Fed survey); and financial markets (Cleveland Fed model including nominal Treasury yields, survey forecasts, inflation swaps data).
It finds that in some cases over the past decade the correlation between expected and actual inflation has been negative, "suggesting a much more limited ability of the expectations measures to predict one-year-ahead".
Professional economists have consistently had the best track record going back to the 1980s and consumers the worst, while markets have been a "rather poor predictor" of overall inflation since 2011.
Regarding the Atlanta Fed survey of firms' headline and core CPI forecasts, "our results suggest that one is better off simply basing one's prediction on the average of the inflation series", the paper's authors find.
To be fair, making predictions is difficult, especially about the future, so it is understandable that people tend to look back in order to look forward.
Kit Juckes, head of FX strategy at Societe Generale in London, notes that because expectations are a function of past inflation, they underprice what will happen until inflation gets really high, then overprice it. By which point, growth may already be heading south.
This can complicate policy-making. The chart below shows how U.S. consumers have consistently over-estimated inflation over the past decade, although the negative spread in recent months suggests that may be changing.
In his rather more incendiary paper last month, Fed economist Jeremy Rudd states that the only basis for the view that expectations influence future inflation is essentially because expectations are embedded into forecasting models at all.
He argues there is nothing more than circumstantial evidence for a link between long-term inflation expectations and inflation's long-run trend, and "no evidence at all" about what might be required to keep that trend in place.
But until something better fills that void, policymakers and financial markets can only work with the tools they have.
"It is without doubt the biggest question of the day, and rather than a clear idea of how it works, all we have is market pricing and forecasts," said SocGen's Juckes.
Editing by Gareth Jones