April 2, 2009

Modeling or Posing?

The notion that model based marks are a solution to the financial system's crisis of confidence, seems puzzling at best. Our entire financial system was nearly taken down (verdict still out) buy rating agency reliance on a Gaussian copula formula used to evaluate credit default probability. Now we are about to replace mark to market accounting with another set of models.

Reuters reports that there's significant tension between banks and regulators ahead of the results.
In the meantime, *bank examiners are struggling in the stress tests to get banks to accept that they are valuing assets on their books too highly, according to bank industry sources*. Many of the banks have completed their own stress tests, and regulators will soon sit down with the banks' management to reconcile the differing results, these sources said.
At issue, not surprisingly, is what tests should be used. Along the lines of their opposition to mark-to-market, the banks would like to see a stress test that assumes the crisis will abate and that we'll return to some semblance of normalcy before too long.

If the purpose of the chosen model is to give you the results you want, then at best you succeed in quantitatively supporting a subjective stance and at worst fraudulently inflate your financial health to the detriment of shareholders and the market as a whole.

It would have been much cheaper and more effective (in my opinion and others) to create an independent fund,
to strategically bid assets that are "cheap". This would be both profitable for the taxpayer and efficient. Instead we have created a structure where financial institutions burden the US balance sheet with assets that are decidedly not cheap. Institutions can now participate in the funds they are selling to. The moral hazard here is that these institutions will sacrifice their minority stakes to the detriment of taxpayer funds to benefit on the other side of the trade.

No comments: