2013-02-21 – London
Earlier this month, the Bank of England released the latest figures of forecasts in its inflation report, which some may see as overly optimistic.
The Bank’s current modelling platform, Compass (Central Organising Model for Projections and Stochastic Simulations), was brought in to replace the existing smaller model, BEQM. The new model was brought in following criticism that the Bank’s previous inflation forecasts were inaccurate. Inflation has consistently exceeded the Bank’s target rates.
Compass was to provide better and more reliable forecasts than its predecessor. Given the shock to the financial system following the credit crisis a new model was required that had the capability to deal with the shock-waves.
The new model steered the Bank’s minimum two-year inflation forecast to 1.8% in February. It seems that the new forecasting model, Compass, is wide of the mark. More worryingly, apparently the model just doesn’t work. Allegedly, users have simply been entering in the numbers.
The Bank of England refuses to release details of the system’s composition and how it works. Until it does, speculation regarding the merit and accuracy of Compass will continue. And of course, expert judgement remains critical to any inflation forecasting done by a computer model.
For the full February 2013 Inflation Report, an electronic version is available on the Bank of England website here.
Sir Mervyn King on inflation target
“You might be tempted to think that an above-target inflation forecast justifies a tighter monetary policy, and certainly ensuring that inflation returns to target in the medium term is our primary responsibility and objective, but the MPC’s remit is to deliver price stability in the medium term in a way that avoids undesirable volatility in output in the short run. The prospect of a further prolonged period of above target inflation must therefore be considered alongside the weakness of the real economy.”
