naked capitalism

An article in the Financial Times, “Credit crisis lurches from bad numbers to bad writing,” by Arturo Cifuentes, contains a stunning revelation: the deal documents governing CDOs are in many cases so badly written that, for instance, it isn’t clear how to handle a default.

To give readers a sense of proportion, the contracts governing transactions involving large amounts of money are typically the focus of a great deal of legal attention. Negotiation of definitive agreements in M&A transactions are done on a line-by-line basis. For financings, law firms usually have standard forms that give them boilerplate for major elements of the transaction. For CDOs, which had been done in high volumes in 2003, it seems inconceivable that a good deal of contractual language wasn’t largely pre set (firms and investment banks have different stylistic preferences, but the general framework and key terms ought to fall within certain well accepted parameters).

But no, the Financial Times tells us quite the reverse:

Understandably, the rating agencies have embarked on an effort to restore their shattered reputations. But, surprisingly, the risk management profession has not received the same criticism the rating agencies have.

That is odd. The fundamental assumptions of this discipline should be, if not trashed, at least re-examined critically. After all, most recent subprime-related losses were suffered by institutions that took pride in the sophistication of their risk management systems.

To summarize: the crisis so far has been an unfortunate sequence of bad decisions related to models, default probabilities, correlations, real estate valuations, expected losses, etc. In short, a massive failure of numbers.

These bad decisions, in turn, have resulted in the collapse of many investment vehicles: more than 100 collateralised debt obligations (CDOs) and structured investment vehicles (SIVs) have already entered the murky post-event of default (EOD) state. This number will grow in the coming weeks.

Unfortunately, the legal documents that govern these transactions are so poorly written – full of ambiguities, inconsistencies, “circular references” and worse, contradictions – that many investors, trustees and respective legal advisors do not know how to interpret them.

For instance, in a number of cases it is not clear whether the assets should be sold (liquidation) or let to run off (acceleration). Moreover, even the rules to distribute the money post-EOD (who gets what) are unclear.

In essence, we have gone from bad models to bad writing. From a failure of numbers to a failure of words. Which brings us to another issue: maybe the “quants” who ran the models and interpreted the results were incompetent. But what about the structured finance lawyers who drafted these legal documents? Did they read them? More relevant perhaps, did they understand what they wrote?