Central bankers are rethinking their approach to economic forecasting after their high-profile failures to spot the most recent inflationary outburst, as officials argue for greater candour with the markets about the uncertainties they are confronting.
The European Central Bank, the Federal Reserve, the Bank of England and other official forecasters failed to see how the end of Covid-19 lockdowns and an energy shock triggered by Russia’s full-scale invasion of Ukraine could pave the way for the worst inflationary spiral in a generation.
After responding with aggressive rate rises, central banks have engaged in intensive postmortems as they unpack the reasons for their failure.
Christine Lagarde, the ECB’s president, told the Financial Times in a recent interview that the central bank needs to learn from its mistakes. “What we should have learned is that we cannot just rely only on textbook cases and pure models. We have to think with a broader horizon,” she said.
One outcome, officials say, is an increased focus on alternative “scenarios” for future economic developments, to illustrate how policy might react.
Huw Pill, the BoE’s chief economist, has told the Financial Times this might be a better way of communicating to the markets than via the BoE’s traditional “fan chart” forecast, which communicates uncertainty by ranges of probabilities.
Sarah Breeden, a BoE deputy governor, said in a December 19 speech that looking at differing scenarios is a “helpful policy tool against a backdrop of unprecedented shocks”.
For its part, the ECB now models different scenarios for inflation and produces a range of sensitivity analyses, such as examining what would happen if wages rose faster or slower than expected or if another energy supply shock struck.
The problem is that so far its early attempts have had mixed results. Even the most “severe scenario” it published in March 2022, modelling the impact of big cuts in Russian gas supplies to Europe, underestimated how high eurozone inflation would rise. It predicted inflation would average 7.1 per cent in 2022 and 2.7 per cent in 2023. Last year, prices in the bloc rose 8.4 per cent and this year they are expected to jump 5.4 per cent.
The BoE’s court of directors has commissioned a review from Ben Bernanke, former chair of the Fed, to examine the BoE’s forecasting and communications. Officials think a greater use of scenario analysis will be among the options examined under the review, due to report in 2024.
Charles Goodhart, a former BoE policymaker, said that, while central banks focusing more on scenario analysis than a central forecast would be desirable, they should produce an even number of scenarios. If they were to produce an odd number, markets would immediately fixate on the ‘central’ outlook, even if it should not be seen as carrying the central bank’s endorsement, he argued.
The lessons on forecasting go well beyond this topic. In Frankfurt, ECB staff are paying more attention to how fast changes in wholesale prices pass into household gas and electricity bills, which differs from country to country.
They focus more on refining margins when forecasting energy prices and no longer assume gas and oil prices will move in parallel, after they diverged massively last year.
An analysis published earlier this year by the ECB found wrong assumptions on energy prices accounted for three-quarters of its overall inflation forecasting errors in 2021, when its prediction for the first quarter of 2022 turned out to be 2 percentage points too low.
In another change, the ECB closely tracks hundreds of shifts in fiscal policy — such as the many government subsidies on energy and food — to get a better grip on their growing influence on inflation. In addition, its staff use a wage tracker they have developed and the results of consumer and business surveys to adjust the output of their models.
Lagarde said that, while the work that ECB staff are doing to improve their forecasting models will help, it will not fix all the problems because “I think that this broader horizon, and the empirical data that we observe and the judgment that we apply to it, is needed.” Many ECB governing council members still treat their own forecasts with a heavy dose of scepticism.
Indeed, the ECB has already decided to rely less on its forecasts and has scrapped its formal guidance on policy. Instead, when deciding its next moves on interest rates, it has committed to putting more weight on whether underlying prices, excluding energy and food, are slowing and to what degree higher borrowing costs are squeezing bank lending and economic activity.
Since missing the extent of the inflation problem immediately after the Covid-19 shock, US central bankers have been open not only about the vast uncertainty underpinning the outlook, but also the limitations they face in deciding monetary policy against such a backdrop.
Fed chair Jay Powell has stressed the need to think beyond typical models. One link that so far appears to be breaking down is the relationship between controlling inflation and the labour market.
Many economists feared getting inflation down to target would need far higher unemployment. Instead, consumer price growth has notably slowed without too significant a jump in the jobless rate from multi-decade lows. That has raised expectations that a painful recession can indeed be avoided — a starkly different outcome than in past inflation fights.
“Our economy is flexible and dynamic, and subject at times to unpredictable shocks, such as a global financial crisis or a pandemic,” said Powell at a recent event celebrating the Fed’s research and statistics division. “At those times, forecasters have to think outside the models.”
Summing up the challenge in September, Powell said: “Forecasting is very difficult. Forecasters are a humble lot — with much to be humble about.”
