CONSENSUS FORECASTS OF INFLATION
During the course of this year, the forecasts for next year’s inflation had to be revised upwards several times, to a large extent due to new shocks, such as the war in Ukraine, and their impact on commodity and in particular energy prices. The fact that inflation became very broad-based also played a role. This year’s experience implies that inflation forecasts for 2023 are surrounded by an unusually high degree of uncertainty.
FREQUENCY DISTRIBUTION OF INFLATION FORECASTS FOR THE USThis point is illustrated by the distribution of inflation forecasts for the fourth quarter of next year. Compared with the distribution for the fourth quarter of this year, there is a shift to the left -forecasters expect lower inflation- but there is nevertheless a high degree of disagreement amongst the forecasters (charts 1 and 2).
The Federal Reserve and the ECB also face this uncertainty in their own projections, which explains why they have switched to a data-dependent policy.
FREQUENCY DISTRIBUTION OF INFLATION FORECASTS FOR THE EURO AREAIt is indeed difficult if not impossible to estimate with a sufficient degree of confidence when and at which level of official interest rates, inflation will have sufficiently converged to target. The former question concerns the speed of disinflation, the latter how much cumulative tightening will be necessary to slow down demand growth to cool inflation pressures.
Disinflation can occur with an unchanged monetary policy. At the risk of oversimplifying, imbalances between demand and supply narrow because demand slows down due to high inflation and supply increases because companies see attractive opportunities due to higher output prices. We can call this process endogenous disinflation. However, such a process would take a lot of time.
Disinflation can also occur when an inflation-targeting central bank reacts to elevated inflation by tightening its monetary policy. For the spending and production decisions of households and businesses, this represents an exogenous shock, so we can call this exogenous disinflation.1
Both types of disinflation are intertwined. Rate hikes reinforce the endogenous dynamics, but the pace of the latter will influence the extent of monetary tightening. Indeed, if it is too slow, it would raise concern that households and companies would expect inflation to stay well above the central bank’s target, i.e. inflation expectations would become unanchored. This could influence demands for wage increases and price setting decisions of firms.
This explains why the Federal Reserve and the ECB have decided to frontload their rate hikes. This way they demonstrate they’re serious about bringing inflation down to target. It reduces the risk of being surprised to the upside in terms of inflation developments because a faster cumulative tightening will slow down demand more quickly compared to a more gradual approach.
A second reason is that it should influence inflation expectations. A recent ECB survey concludes that “the significant rise in inflation since 2020 appears to have influenced the role of expected inflation in expected selling prices, with 58% of firms now reporting it as “very important” (up from 30% in 2020).”2
This raises the question whether firms’ inflation expectations are forward or backward looking. In case of the former, they would incorporate the expected negative impact of monetary tightening on the outlook for aggregate demand, pricing power, etc. In case of the latter, recent inflation developments play a key role in forming expectations about the future, so inflation first would need to drop before seeing a decline in expectations.
Recent research of the IMF sheds light on this question. Using a unique dataset on Chilean firms, the authors find “that firms use changes in input prices observed in the transactions with their suppliers -i.e. supply chain inflation- to form expectations about aggregate -i.e., CPI- inflation3 .” Firms use price information to which they are directly exposed -i.e. local information- to form an opinion of future, aggregate inflation. Similar results were obtained in earlier research on French companies4 . This could mean that disinflation could be a slow process considering that a majority of euro area manufacturing companies are still confronted with rising input prices5 .
William De Vijlder
