My MPC-ometer model predicted large cuts in Bank rate in November and December but not on the scale delivered. There was a similar divergence in 2001, after the 11 September US terrorist attacks, when the MPC appeared to “bring forward” reductions ahead of an expected deterioration in economic data.

To capture this behaviour in 2001, I included a “dummy variable” in the model, which had the effect of lowering the predicted level of Bank rate in October and November 2001 and raising it over the subsequent four months. Using an identical timing shift variable starting in November 2008 improves the model’s recent forecasting performance.

The 2001 parallel suggests that, having cut rates by more than suggested by incoming data in November and December, the MPC will require compelling evidence of further economic deterioration and / or diminishing inflationary pressures to warrant continuing to ease aggressively in early 2009.

For January, the version of the model including the dummy variable predicts a 25 bp cut. After its late 2008 pyrotechnics the MPC is unlikely to move by as little as 25 bp so the choice is between no change and a 50 bp reduction. I favour the former, on the view that the Committee will regard falls in the exchange rate and interbank / policy rate spreads over the last month as substituting for a further official cut. (The sterling effective index on Friday was 8% below its level at the start of the December meeting, while the three-month LIBOR / overnight indexed swap rate spread was 80 bp lower at this morning's fix.) An additional argument for delaying any further move until February is that fourth-quarter GDP figures and results of the Inflation Report forecast round will then be available.

Interestingly, the Sunday Times Shadow MPC, which often calls the decision correctly, voted 6-3 for no change at its latest meeting.