The ABCs of counterparty credit

The FT has an article on the failure of SCC, a Dublin-based credit derivatives product company or CDPC. SCC was only twenty five times leveraged, a relatively low level compared with some of the other CDPCs. But it did have a small absolute level of capital, $200M, and no rating.

What is interesting is why SCC collapsed. It wasn't that it had written CDS on underlyings that defaults. No, as with ACA, it defaulted because it could not post sufficient collateral. Let's pick up the story with the FT:

Court documents from Nomura’s attempts to liquidate the company and SCC’s successful response to secure a Chapter 11-style restructuring show that between the end of June and August 16 last year, collateral demands rose from $55m to $438m. SCC managed to put up $175m worth before running out of funds and sparking Nomura’s High Court petition to have the firm liquidated.

The ravages wrought by mark-to-market accounting are visible in the tens of billions of losses among investment banks, the collapse of the structured investment vehicle industry and in the ever-more precarious position of the bond insurers.

And yet, credit losses from actual defaults outside of the US subprime mortgage market remain minimal. SCC told the High Court that expected losses over the life of its contracts would be a fraction of the collateral it had to post.

Now of course we have no idea whether SCC is right about that claim or not, but the key point is that credit support default is a real honest to goodness game over default. The relevant question for a CDS counterparty is therefore not only whether they are sufficiently well capitalised to remain solvent under a claim: it is also whether they are sufficiently liquid to be able to adhere to the terms of their CSA. It's the volatility of the mark to market of their portfolio that matters, not only the ultimate loss. Did the prime brokers think about when they were buying protection I wonder?