At first sight, fair value accounting has an obvious validity ­ but then so did current cost accounting.

As deputy treasurer for Glaxo from 1992 to 1995, I argued strongly that ‘marking to market’ was the only way to measure the investment performance of our cash portfolio ­ more than £3bn of it. Measuring these returns through the ‘lower of cost or market’ gives a fool’s view of the daily results of actively traded investments. Fair value seems at first sight to be the modern interpretation of this performance-based principle.

But the recent difficulties in bond and cash markets should make us all question fair value. In the early 1990s, Glaxo was caught holding more than $1bn of collateralised mortgage obligations. CMOs were a bit like today’s collateralised debt obligations ­ collections of primary debt packaged by banks to suit ‘secondary market investor demand’. Many were packaged to appear like cash but to pay a higher return through ownership of riskier debts.

Their sale depended on investors’ willingness to seek higher returns and misunderstand risk. When everyone suddenly realised that the emperor had no clothes, buyers disappeared and the market collapsed.

The lesson is that neither CMOs nor CDOs were truly market-based.
We can be confident that the government bond markets and the FTSE 100 share market have high liquidity at all times. But if we need to calculate fair value for anything else, at best we are merely using a financial model using suspect assumptions. And if this is true of frequently traded investment securities, then what price would you give for the ‘fair value’ of a long-term business asset where management or even another audit firm provides the model?

We are basing much of our accounting assessments on numbers that may have no validity when markets are volatile or illiquid. Yet this is when we need them most if we are to use fair value meaningfully.

Yes, let us ‘mark to market’ where active trading is carried out and there is a real market, but let us not try to determine fair value when its assessment depends on potentially illiquid, rarely traded assets.

So how did Glaxo get out of its problem that mark to market produced volatile investment accounting results? By placing more than £3bn into bank deposits that did not need to be marked to market. When the company wanted to rearrange its finances after buying Wellcome, it simply asked for the deposits back early ­ and surprise, surprise, the banks nearly all obligingly returned the money!

There is always a way.

Mike Leyland is a former deputy treasurer of GlaxoSmithKline