MNI INTERVIEW: BOE Models Can Ditch 2% Inflation Expectations
The BOE could respond to Ben Bernanke's criticism of its forecasting by using survey inflation expectations, a former head of modelling at the Bank tells MNI.
Bank of England models can be reworked in line with recommendations by former Fed chief Ben Bernanke to remove an assumption that inflation expectations always return to the 2% target, but it would be better to produce staff forecasts without input from policymakers, a former head of BOE modelling told MNI.
While Bernanke was sharply critical of the assumption in the BOE’s central model, Compass, in his recent review of its forecasting, Francesca Monti pointed out that the BOE’s quarterly economic forecasts are not just the product of models but also serve as a communication device for the Monetary Policy Committee, whose members use their judgement to provide input throughout the forecasting process.
“Imagine that the Bank of England said, 'No, actually, we're not reaching the 2% inflation target within a reasonable time frame. Inflation is going to stay at 3% … for a long time. They can't really do that. The MPC’s forecast is not just a forecast, it’s also a communication tool,” said Monti, a professor at UC Louvain and a former head of modelling in the Monetary Analysis Directorate at the BOE.
JUDGEMENT FREE
“I genuinely don't know if, letting Compass run without any judgements, it would go back to 2% in three years, maybe not. It will certainly go back to 2% in the long run, that's for sure. But I don' t know if it would go because there's been a number of very persistent shocks,” she added in an interview, adding that it would be better to have a separate forecast produced purely by staff and to be uninfluenced by the MPC. (See MNI POLICY: Time Needed For BOE To Make Bernanke Changes)
The technical work of refining Compass though is very doable, she said, adding that modellers could use survey data and allow for wedges between long-run inflation expectations and the inflation target when substituting for the rational expectations assumption criticised by Bernanke.
“There definitely are ways to model inflation expectations, and expectations in general, in a more realistic way. And I think the literature has taken really huge steps, both in terms of understanding what are the empirical features that we should try to match and also what type of models would work,” Monti said.
Her own research with former senior BOE economist Riccardo Masolo found that uncertainty over monetary policy-making creates “beliefs that are distorted .. [and] also in steady state, gives rise to a wedge between the inflation target … long-run expectations and trend inflation,” Monti said.
The BOE has its own inflation expectations surveys, she noted.
“I do think that using data on expectations would help discipline the model and account more for the reality of what the expectations are,” she added.
STAFF RETENTION
Another thread in the Bernanke Review was how some of the BOE’s most technically qualified staff spend relatively little time working on the forecast round, but Monti said the Bank struggles to retain researchers. (See MNI INTERVIEW: Clear Case For BOE Alternative Scenarios - Bean)
“Research is really penalised at the Bank. And so that's why researchers end up mostly leaving,” she said, adding that they rapidly arrive at a point where the next step up at the Bank is into management.
“Retaining researchers would help improve the forecast process, and would also help in responding more nimbly to issues that might arise. When a crisis hits, having in-house people that have strong modelling and econometrics skills is crucial for understanding quickly the drivers of the current crisis, their transmission and the best response to them,” she said.