Financial Times (Comments page), 30th August 2007

One frequent claim used to calm fears in the recent market turmoil has been that "corporate balance sheets are in good shape". Judging from the published accounts of companies the claim is justified; judging from national accounts it is not. This marked divergence depends on the different conventions used. In broad terms, national accounts assume that assets are worth their cost of production, adjusted for inflation, after a deduction for depreciation, while corporate accounts seek to allow for market prices whenever possible.

The difference in the apparent weight of corporate debt as shown in different sources has become much stronger in recent years for three reasons.

First has been an increasing emphasis on "marking to market" reflecting the market value of an asset. Second has been the increasing importance of financial companies, most of whose assets are financial rather than tangible. Third has been the growth - even in balance sheets of non-financial companies of financial assets, which have risen in the US from being equal to 30 per cent of tangible assets and net worth in 1952 to more than 80 per cent today.

For the UK, the data published by the Office for National Statistics show that private non-financial companies have high and rapidly rising leverage. Net debt amounted to 20 per cent of the value of assets at replacement cost in 1989 but more than 50 per cent at the end of 2006. No such adverse change is revealed in the accounts of companies.

A similar problem appears in the US national accounts data, which show that non-financial companies have been paying out more than 100 per cent of their profits in dividends and buy-backs. Up to 1984, a combination of profits' retentions and new issues allowed a steady rise in net worth in every quarter. Since 1984, buy-backs have exceeded retentions, so that the net reductions in equity have averaged more than 3 per cent a year. With investment in plant and equipment being $4,000bn per annum greater than the allowance for depreciation, the national account data published by the Bureau of Economic Analysis point to rapidly rising leverage. In order for the balance sheet data published by the Federal Reserve to conform with the accounts published by companies, the former are adjusted by the addition of "statistical discrepancies", which currently run at $800bn a year.

It could be argued that the convention on which national accounts are drawn up should be altered. If the efficient market hypothesis is correct, the stock market provides a perfect guide to the value of corporate assets and the current convention is misleading. This approach, however, produces results that would generally be seen as absurd. Valuing assets at their stock market value would have frequently produced swings in gross domestic product over one year of 20 per cent plus and minus, with total corporate profits often being heavily negative.

Nonetheless, it seems to me to be correct for accountants to require companies to mark to market their assets and liabilities. Even if the results are ephemeral in aggregate they will provide a much better guide to the relative positions of companies, comparing one with another. But if companies use mark-to-market values - and these have become increasingly important - we should recognise that the aggregate profit and balance sheets they publish will give a misleading impression if compared with similar data from earlier years.

But we still need a stable criterion for comparing today's leverage with that of earlier years and this can be done by comparing debt with output. Profits are only 30 per cent of output (the sum of broadly defined profits and labour income), so the latter is not nearly so sensitive to changes in definitions as the former, particularly if the data are drawn from national accounts. Corporate output rises in a relatively stable way year by year and, because profit margins are strongly mean-reverting, the ratio of debt to output provides an excellent measure of the capacity of companies to service that debt, independent of their current profit levels. Compared with output, US corporate leverage is high and rising. Although mildly below its previous peak in 2002, it is well above its long-term and even higher than its post-1990 average levels.

It is likely that the growth of debt off-balance-sheet makes the underlying situation even worse. When, for example, airlines sell their aeroplanes to leasing companies and then hire them back, the debt moves from the non-financial to the financial sector. I do not know how much of the growth of financial debt comes from such activities, but financial debt has grown without interruption from 3 per cent of GDP in 1952 to 110 per cent today in the US.

It is therefore unwise to assume that corporate balance sheets are in good shape. The realisation that they are not will probably become widespread as asset prices fall.