A horse. A zebra. Or maybe it’s a fish?

Is it the Bank’s position that the data changed whilst its economists were tucked-up in bed?

Until a few days ago, I had never heard the expression: “When you hear hoofbeats, think of horses not zebras”. A visiting US professor used it in conversation with me. Then I heard it again, last night. This time spoken by Patterson, an FBI agent (sort of).

The expression originated in the 1940s in the University of Maryland medical school. Dr Theodore Woodward wanted his students to think, first, of the most likely explanation for a set of symptoms, not the most outlandish.

Seems sensible, so long as they don’t take it too far. The challenge for young doctors is to be ready to diagnose the most likely condition and not think, first, of the rare diseases they have read about. But the challenge for the experienced doctor is to be alert to the rare possibility and not focus solely on the diseases they see year-in, year-out. Some estimates suggest that it can be an average of nine years for a rare condition to be diagnosed – something I know about from my own experience.

So, too, in business and finance. The challenge for those in risk management roles is to be alert to all the possibilities, not just the most likely, in time to take suitable measures, be they warnings, preventions or remedies.

This morning, it is widely reported that the Bank of England’s chief economist, Andy Haldane, has likened the failure to predict the 2008/09 financial crisis to the famously inaccurate forecast by the BBC weatherman, Michael Fish, ahead of the UK’s great storm of 1987. Mr Haldane has a point, but it may not be as good as he thinks.

Before each event – the storm and the crash – there must have been a period of time during which the evidence was consistent with a normal outcome as well as the rare one which actually occurred. But, at some point, the evidence must have narrowed down the options. When the wind whipped up on 16 October 1987, most people were fast asleep and it was too late to batten down the hatches. Was the same true before the financial crash in 2008? Is it the Bank of England’s position that the evidence changed during a period of time so brief that economists were all tucked up in bed?

It is reassuring that the most likely outcome, in probabilistic terms, is usually the least sinister. A headache not a brain tumour. A thunderstorm, not a hurricane. But sometimes, that clippety-clop sound really is a zebra.