It wasn?t the butler, it was the accountant that murdered the markets. On Monday the 9th I was invited to Paris to speak to a group that advises the International Accounting Standards Board on accounting standards for financial institutions, on a new accounting standard that may take away some of the self-inflicted damage in a financial crisis.
There are two principal methods for valuing an asset. The first is its current market valuation, or an estimate of what that might be. The second is the present value of the future cash flows from that asset. In normal times these two alternative approaches to value lead to similar numbers; but sometimes like today, liquidity risks, and other causes of market dysfunctionality, causes these two alternative approaches to diverge, occasionally sharply so. The choice of valuation methods in such cases should be based on the relative maturity of the intermediary?s funding and I have called this approach: ?mark-to-funding? accounting.
In order to illustrate the issue most simply, assume that there are two financial institutions, A and B. Both hold a single identical asset, whose market value has fallen well below the present value of future expected cash flows. Institution A has financed this asset on the basis of a five year bond so that the durations of asset and liability are exactly matched; whereas institution B has financed the same asset on the basis of a one month liability which needs to be rolled over every month. If the assets in both institutions are valued equally at the latest market price, both will appear to be insolvent. This is unfair to institution A, which has no need to sell the asset, and can ride out the liquidity crisis, because its funding is unaffected. To require institution A to mark-to-market its assets and then in response to the apparent insolvency of mark-to-market valuations to try and raise capital by selling assets, pushing down asset prices further in a loss-spiral, is an artificial source of crisis when there are already enough real ones.
If, on the other hand, the assets in both institutions are valued at the present value of expected cash flows, both may appear to be sound, but this too is incorrect, since the liquidity crisis means that institution B may either not be able to roll over its funding needs at all, or only at a much higher rate of interest. Institution B really is insolvent. Clearly there is a major difference in solvency, depending on different funding positions, and accounting methods ought to reflect this.
Apart from those who would deny that market prices can ever move away from the fundamentals of expected future cash flows, the above analysis should not be controversial. The problem that many see with my ?mark-to-funding? proposal is practical. In practice banks and other financial institutions have many assets financed by many different forms of funding. It is not possible normally to say that a particular liability finances a particular asset; instead all liabilities go into a common pot to finance all assets. On this view, although the objectives of mark-to-funding may be praiseworthy, it may not be practical.
We think that this objection can be met. There are cases where particular assets and liabilities can be directly related. Moreover present values of cash flows and market prices diverge significantly occasionally. In practice our proposal would allow an institution, when such a divergence did occur, to ?carve out? the assets to which this applied, and to select which liabilities it chose from its portfolio to support those assets. Suppose that the chosen liabilities had as long a duration as the assets, then the valuation would be present value of cash flows (PV); if half as long, then the valuation would be half PV and half market price.
In a sense what this proposal does is to allow any institution in a liquidity crisis to carve up its own internal ?bad bank? mechanism to isolate hard-to-sell assets but only so long as it has sufficient long term funding to do that. The need for such a mechanism should only be occasional, but a benefit of doing this is two fold: first it will reduce the selling stampede in a crisis that can become self-fulfilling and secondly it provides an additional incentive for banks to seek out additional longer term funding in normal times.
Such a carve out could be operated flexibly with no limit on the use of assets or liabilities assigned to it, which could vary over time. Such a mechanism might have been very beneficial to such banks, financial institutions and the market as a whole, where institutions had longer dated liabilities, but were instead forced by their accounting standards to behave as if they did not.
-The author is chairman of London-based Intelligence Capital, governor of the London School of Economics and Emeritus Professor of Gresham College in the UK