In 2002 a colleague suggested I compare the bear market then under way with previous big falls in share prices. The three largest peak-to-trough declines in UK shares over the prior 100 years were 1929-32, 1936-40 and 1972-74. I rebased the respective bull market peaks to the FTSE 100 peak in 2000 and calculated an average of subsequent performance. Smoothing this average generated our “three bears forecast”.

As the first chart shows, the three bears forecast proved remarkably accurate in pinpointing the level and timing of the low in share prices in the early 2000s as well as the tracking the recovery over the subsequent four years.

The second chart decomposes the forecast into its three historical components. The latter began to diverge significantly in late 2007, suggesting the forecast – based on the average – would break down, as it has indeed done.

There may, however, be some life left in this data-mining exercise. As the second chart shows, stocks seem to be following the pattern of the interwar years. The early 2000s bear market tracked the 1929-32 decline while recent falls so far mirror the 1936-40 bear.

The third chart extends the interwar comparison over the next five years. Share prices are now below the historical template, suggesting a near-term rally. An extended plateau is then indicated followed by a final lurch down to a bottom around the levels reached in 2003. A sustainable recovery occurs only in 2011.

The 1936-40 bear market initially reflected a US recession but was extended in duration and magnitude by the onset of the Second World War – the final move down in 1940 coincided with the Battle of Britain. The current financial crisis is momentous but it is hard to believe it represents a threat to the British economy on a par with Nazi invasion.

It would be unwise to place strong weight on mechanical historical comparisons but the suggestion of an imminent base followed by an extended sideways movement is plausible. Valuations are now low – the price to book ratio of UK non-financial stocks is at levels last reached in 1992 – but buyers are likely to be slow to return after the confidence-shattering events of recent months.

The success of the UK banking rescue plan will depend importantly on yet-to-be-announced terms and conditions. The authorities need to strike a balance between protecting taxpayers’ interests and providing some upside for equity holders – necessary to stem the downward spiral in share prices and attract new private sector capital.

A key issue is the fee to be charged by the government for guaranteeing an expected £250 billion of bank debt. As argued previously, the effectiveness of the special liquidity scheme has been reduced by its perverse fee structure. The Bank of England charges banks the spread between three-month LIBOR and the three-month rate on government borrowing. This means the scheme becomes more expensive when LIBOR rates rise – the time banks most need to use it.

The fee charged for guaranteeing debt must reflect the strength of the institution concerned but should not depend on volatile market assessments, such as credit default swaps. One possibility would be to base the charge on the amount borrowed and the credit rating of junior bank debt.

Similarly, the terms of new government-subscribed capital issues should not be unduly onerous. Demanding a Buffett-style 10% yield along with interference in dividend policy and other decisions affecting future ordinary shareholder returns might generate short-term political plaudits but could prove self-defeating.

The £100 billion expansion to £200 billion in the special liquidity scheme does not represent a new initiative, since there was no previous upper limit on banks’ access to the facility. The extension of the definition of eligible securities for the scheme to include new government-guaranteed bank debt also appears of little consequence – it is unclear why banks would wish to pay a fee (charged on top of the guarantee fee) to swap such government-backed paper for Treasury bills of the same credit rating.

Unlike the Fed and the ECB, the Bank of England has yet to allow securities rated lower than AAA to be used as collateral in its money market operations. This may partly explain the undersubscription of its £40 billion auction of three-month funds this week. It is doubtful that the long-awaited plans for a new discount window facility, to be revealed next week, will loosen collateral requirements further.

The one percentage point rate cut in Australia overnight has increased speculation that major central banks are planning to “shock” markets with large-scale and/or co-ordinated interest rate action.

When analysing past interest rate decisions in order to construct the MPC-ometer, I found the MPC had behaved unusually following the 9/11 terrorist attacks, cutting rates earlier and by more than suggested by economic and financial data. This was counterbalanced by a less dovish stance than warranted by incoming news in later months. I accommodated this behaviour in the model by including a 9/11 “dummy variable”.

Escalating financial turmoil in recent days risks damaging economic confidence in a similar way to the 9/11 attacks, suggesting the MPC will again choose to bring forward easing as insurance against a worst-case scenario.

In terms of the MPC-ometer, this can be analysed by “switching on” the 9/11 dummy variable. Using the model to predict this week’s decision based on the current level of Bank rate, the forecast then changes from a quarter- to a half-point cut. If three-month LIBOR is used instead of Bank rate, a three-quarter point move is predicted.

If the post-9/11 period is a guide, however, any shock move this week will represent a shift in the timing of MPC action rather than implying a lower ultimate level of official rates.

A week ago most economists thought it was too early to cut rates. Now there will be disappointment with a half-point reduction.

My MPC-ometer model has been suggesting an October rate cut for several weeks and recent data have confirmed the forecast. According to Reuters, only 21 out of 66 economists projected an October move in a Reuters poll last Wednesday but opinion had shifted by the end of the week, with 49 respondents expecting a cut, of which four forecast a half-point reduction.

The MPC-ometer was designed to answer the question “How will the MPC vote given incoming data and the current level of Bank rate?” Used in this way, the forecast is for either a 7-2 vote for quarter-point cut this week or – more likely – 1-5-3 (Blanchflower voting for a half-point again).

The trouble with this approach currently is that the model is unlikely to capture fully the impact of the seizure in money markets on the MPC’s thinking. One way of addressing this is to change the question to “How will the MPC vote given incoming data and the current level of three-month sterling LIBOR?” On this basis, the model predicts a unanimous vote for a cut and an evens chance of a half-point reduction. (This is based on today’s 6.27% three-month LIBOR fixing.)

Another issue is that the Committee will have an early look at the September CPI figures, which may show annual inflation moving above 5%. This would have a small effect on the above forecasts but a quarter-point cut would still be odds-on even using the model in the normal way.

While several MPC members may vote for 50bp, I suspect a majority will prefer a quarter-point cut with a follow-up move next month. Current economic risks stem less from the price of money than the complete breakdown in the plumbing of the financial system, a problem better addressed by direct official intervention, including the Bank of England intermediating money and credit flows – see previous post.