The Bank of England has launched a review of how it makes and uses economic forecasts after coming under fire from politicians for repeatedly failing to predict the rise and persistence of UK inflation.

In a letter on Wednesday, David Roberts told the House of Commons Treasury committee that the central bank’s governing body, which he chairs, had in May decided to commission a broad external review of its “forecasting and related processes during times of significant uncertainty”.

The announcement of the review came as traders increased bets on further interest rate rises, and mortgage lenders rushed to reprice loans, on the back of BoE governor Andrew Bailey’s admission that it would take “a lot longer than we expected” for inflation to come down from its current level of 8.7 per cent.

Markets are betting that interest rates, now at 4.5 per cent, will peak close to 5.75 per cent later this year, following economic data that suggests the UK’s inflation problem will be harder to tackle than elsewhere.

The cross-party group of MPs had called for an overhaul of forecasting processes after members of the Monetary Policy Committee, which sets the base rate, told them that the BoE’s own modelling was not producing accurate results and that the MPC had reduced its role when setting rates.

In a letter earlier this week, committee chair Harriett Baldwin this week urged the BoE’s court of directors to assess “the current effectiveness of the Bank’s forecasting platform”, the transparency of the forecasting process and whether it was sufficiently open to external challenge. Roberts described her call as “timely”.

While the BoE has yet to set the terms of the review or decide who will lead it, it is likely to draw on international expertise and give it a broader scope than the Treasury committee had suggested — looking both at technical ways to improve the BoE’s modelling and at how the MPC uses the forecasts when it sets policy and communicates its decisions.

Huw Pill, BoE chief economist, told the committee last month that it was “almost inevitable” that models based on the past 30 years would go wrong in the face of big new shocks. He added that the MPC was “trying to understand why we have made those errors . . . and then make an assessment as to whether that behaviour will continue into the future”.

Broader questions, however, include whether the MPC should still centre its communication on a central forecast — presented as the committee’s collective judgment — or whether it would be better to publish scenarios, helping MPC members to explain their vote when the committee was split.