Forget computers: What we want now is common sense

One plus from the recession has been that inflation has all but disappeared, and is everywhere well below the projected rate which is built into pension fund assumptions. A second bonus is that employees have shown remarkable restraint when it comes to pay. There have been a few disputes and strikes for more money, but these have been comfortably outweighed by the numbers of people prepared to job-share and go on short time working and sabbaticals, to help their employers cut the wage bill in these difficult times.

Two of the big three drivers of pensions costs are inflation and the rate of increase in earnings — the third is the life expectancy of the members. And without wishing to dwell on an uncomfortable truth, recessions and the accompanying increase in poverty don't do much to increase longevity. So with at least two of the three drivers of liability being flat or negative, common sense says that this year there should be no significant increase in pension fund liabilities.

But that is not what we find. Yesterday Aon, one of the biggest pension consultants in the country published a study which highlighted a further ballooning in the liabilities for the 8000 or so UK companies which still offer some form of defined benefit pensions schemes. The total deficit of the top 200 schemes rose in August alone from £73 billion to £78 billion. Total liabilities of all 8000 schemes rose to more than £1 trillion for the first time. Though the deficit across the board is only a small fraction of this amount, nevertheless the consultants said this worsening of the liabilities and deficits would put further pressure on companies and mean that even more of them would surely close their schemes altogether.

But this is where we need to pause. One thing we should have learned from the credit crunch is not to have blind faith in economic and financial models. When those models come up with answers which fly in the face of common sense, then we should trust common sense not the computer. If bankers and investors had been more willing to do that they might not have believed that a bag of toxic mortgages could be turned into a risk-free copper-bottomed investment and we could have been spared a great deal of pain. The reason of course is that computer models make predictions, but the answers they come up with are determined by the information and assumptions which are fed in. If the maths is wrong, the answer will be wrong and actions based on that answer may well be disastrous.

It is time this thinking was applied to pension funds. The entire industry is in thrall to the valuation models produced by consulting actuaries which seek to measure assets' liabilities and therefore deficits. These models which measure liabilities are, like all models, dependent on their underlying assumptions.

Indeed to prove the point, any actuary could show you how to slice millions off a deficit — or add millions to it for that matter — by making only the tiniest changes to the assumptions used for earnings growth and inflation.

It is these models which are saying that pension liabilities are going through the roof. They say that because in calculating these liabilities the models assume a rate at which they will increase over time, known as the discount rate.

This formula is in turn derived from rates in the market, and in particular from the yields on corporate bonds.

These are behaving very oddly at the moment, partly because of the economic uncertainty, partly because of the interventions by the authorities. This odd behaviour affects the discount rate and that drives the calculation which says pension fund liabilities are going through the roof even when common sense says they can't be.

One of the insights from Lord Keynes — the Thirties economist whose thinking has become fashionable again — was that it is better to be roughly right than precisely wrong. But that is not how it works in the pension world. No one seems to care that , once again, if the maths is wrong the answer will be wrong and policy based on that answer will again be disastrous.