American CDOs Are Not The Same As Japan's Bad Bank Loans of the 1990s

Many people seem to be confused about the role that 'mark to market' accounting has played in the current US financial services sector debacle.

No less than two former leading lights of US financial regulation, Arthur Levitt, former SEC chairman, and Lynn Turner, the agency's former chief accountant, both of whom retired in 2001, wrote an editorial in the Wall Street Journal yesterday on the subject.

As you might expect, they lecture on why anything but that method of valuation misleads investors.

Another frequent analogy is made to Japan and its famous 'lost decade' of the 1990s, due to the false valuation of nonperforming bank loans allowed to remain on bank balance sheets at face value.

Allow me to dispel these myths. Especially the second one.

Levitt and Turner are correct in their assertion that full information on the true value of an asset is essential for investors, in order that they can make informed investment decisions with respect to the value of a firm with questionable assets on its balance sheet.

However, they cannot wave away, as if with some magic wand, the real disparity between two values of a performing (or, in some cases, even a non-performing) asset: the current market price at which the asset can be sold, and the present value of the economic performance of the asset when held to maturity, or for a very long duration.

It is as wrong to penalize owners of a firm whose performing assets, though expected to be held to maturity, are currently valued at a far lower value in the market, by forcing them to value those assets at the current 'liquidation' value, as it is to overstate asset values by denying their nonperforming status.

In the case of Japan, non-performing bank loans were held on bank balance sheets at their full value, overstating the banks' values and causing severe counterparty risk. Their financial system froze up due to these inflated values of non-performing loans.

The current dilemma in the US concerns securitized, structured financial instruments backed by residential mortgage loans.

Because the securities are artificial creations composed of mixes of payments from many underlying mortgages, investors are now concerned that it is difficult, if not impossible, to discern which CDOs are relying on payments from delinquent or defaulting mortgage loans, or loans likely to default. Thus, a counterparty risk has arisen which is depressing the 'liquidation' value of such CDOs.

But nobody disputes that most of these CDOs are, in fact, still 'performing.' Their economic value, held to maturity, is quite high. Far higher than the immediate market liquidation value.

Unlike Japan's clearly non-performing bank loans, most of the mortgages underlying the outstanding CDOs are performing.

Yes, a percentage of the underlying mortgage loans will become non-performing. Some were 'alt-A' or 'subprime,' and will have a higher-than-average default rate. But it won't be 100%.

Thus, there is no comparison whatsoever between the Japanese experience with bad, non-performing bank loans in the 1990s, and American mortgage loans in this decade, most of which are, in fact, still and expected to be, performing.

Levitt and Turner are incorrect to suggest that investors may only be misled by overstated values of performing assets whose immediate, liquidation values are depressed due to factors not related to the actual performance of the assets.

Assets such as CDOs can be performing, but valued much lower for immediate sale, due to a general misperception of the true percentage of non-performance in the asset class, and the inability to distinguish which are non-performing at this time.

That doesn't mean that a firm wishing to hold performing CDOs to maturity owns an asset which is, for the purposes of that firm's use of the asset, worth only its liquidation value.