WHY ARE CENTRAL BANK FORECASTS SO WRONG? / THE FINANCIAL TIMES

Why are central bank forecasts so wrong?

Mistakes in forecasting the severity of price pressures lead to soul-searching among rate setters

Chris Giles in London

© FT montage/Reuters

The Bank of England is holding a “Festival of Mistakes” this week, celebrating lessons learnt from financial disasters of the distant past. 

Some would argue that they, and their counterparts at other central banks, should focus on more recent errors.

Advanced economies are experiencing the most acute — and most enduring — outbreak of inflation for a generation. 

Yet almost all rate-setters failed to spot the degree to which price pressures would ratchet up — and stick around, despite record amounts of monetary and fiscal stimulus.

Most of the Federal Reserve’s rate-setters failed to foresee that inflation would ever rise, and then overestimated the speed of its decline. 

Economists at the BoE and the European Central Bank underestimated the scale and persistence of inflation. 

Across the world, poor forecasts have contributed to central bankers failing to do their main job: maintaining price stability.

The failure to spot inflation has not only left central bankers risking financial instability by having to raise rates far faster than usual but threatened the credibility of institutions that rely on trust to steer the economy towards sustainable growth.

Stephen King, senior economic adviser to HSBC, blames their collective failure on rate-setters relying too much on their own capacity to control the public’s expectations of what will happen to prices in the future.

In normal times, the rules that govern companies’ decisions on pricing and workers’ demands for wage rises can be influenced by central bankers’ own inflation targets of around 2 per cent.

But what forecasts failed to show was that those rules only hold when inflation is broadly stable. 

Once price pressures soar — and stay high — people begin to believe that “the central bank is now talking nonsense”. 

Scepticism abounds, and recent inflation readings come to matter more than central banks’ insistence that their policies can quell price pressures.

The reputational hit from their failure to foresee inflation has left central banks having to raise rates aggressively — by up to 75 basis points in single policy votes — to convince investors and the public of their commitment to low inflation.

In some quarters, it has also led to soul-searching.

The ECB issued a mea culpa for its underestimate of inflation and has promised to focus more on underlying inflation rather than forecasting models in future. 

The IMF has also talked openly about its forecasting “misjudgments” — although this candour did not appear in any of its flagship reports.

The BoE has been more combative, arguing that its mistakes owed little to errors in how it devises its forecasts and were instead the result of big shocks such as the war in Ukraine, which it could not have predicted.

While other organisations have been less defensive, economists elsewhere caution that the public should focus less on whether or not projections turn out to be correct.

That, they argue, is impossible and the public should focus more on whether the projections say something insightful about the economy at this point in time.

Richard Hughes, who heads the UK’s independent fiscal watchdog, the Office for Budget Responsibility, acknowledged the failure to spot the build-up of price pressures was — along with underestimating the decline in productivity growth since the global financial crisis — one of “two big macro forecasting errors” made in recent decades.

However, forecasts remained, he said, “the best understanding of the future, conditional on your knowledge of the present”.

He highlighted the similarity between these predictions and financial market pricing, which also moves as the facts change. 

“[Markets are] ‘reacting to news’, while [we] ‘got it wrong’,” Hughes said.

Alexandra Dimitrijevic, global head of research and development at Standard & Poor’s, the credit rating agency, said the purpose of forecasts was not to get the numbers right to the last decimal point but “to look at the narrative, the direction and the risks”. 

“By definition a forecast is never right. 

The question is whether it is useful,” added Dimitrijevic.

Clare Lombardelli, the new chief economist at the OECD, noted that dire predictions of a bleak winter across Europe were based on assumptions for the weather that, by luck, was warmer than normal — meaning gas storage, and therefore economic growth, held up.

Daniel Leigh, who heads the team that produces the IMF’s World Economic Outlook — which includes projections for each of the fund’s 192 member countries — said that failing to predict major economic trends did not mean forecasters were clueless.

Even if they turned out to be incorrect, officials and ministers still found the fund’s projections useful, he said, because they gave a sense of scale and explained the likely ripple effects of global trends.

“The priority is to give decision makers a sense of what to expect, but also the risks, so that they can take the steps needed,” said Leigh.

However, others are less sympathetic.

Mohamed El-Erian, president of Queens’ College, Cambridge, and an adviser to Allianz, has said the Federal Reserve’s original forecast that high inflation would be “transitory” was “one of the worst calls in decades”.

He has argued that if the Fed had more closely examined evidence from businesses and the implications of its own actions, then it would have spotted the severity of the rise in inflation earlier.

Lombardelli acknowledged that central banks faced an especially tricky task, as they must not only produce forecasts for inflation but set policy to influence price pressures two years from now.

“What do you assume about the effects of policy if you’re the policymaker?” she said.

That challenge is especially tough in an environment such as the present, when the after-effects of the pandemic and Russia’s invasion of Ukraine are hard to predict.

El-Erian believes meeting it would prove difficult, particularly for the Fed.

Central banks, especially in the US, had made major “one-sided” forecasting errors without acknowledging them. 

Those errors could be blamed on models “failing to keep up with significant structural change in the economy”, being too late to look at “micro data”, and groupthink.