In a story today, Fortune CNN Money takes the Federal Reserve to task for extending the deadline for banks to sell non-Volcker compliant CLO paper. The story itself is a polemic about the dangers that CLOs pose to the banking system. To back up this argument, the commentator, Cyrus Sanati, compares the potential losses of CLOs to those suffered by CDOs, which “imploded so spectacularly during the 2008 meltdown.”
There are several factual errors in the story. For instance, the story says, “But with Volcker set to go into force (officially) in July 2015, banks would not only have to stop creating CLOs now, they would also need to quickly sell off investments they have accumulated.”
However, while the rule would undoubtedly make it more challenging for banks to originate CLOs and, potentially, own CLO liabilities, there are regulator-approved fixes, such as structuring CLOs that are unable to hold bonds.
Also, the article says that “[f]or the last 94 weeks, there has been a positive inflow of cash into the CLO market.”
This figure is clearly a reference to retail inflows into U.S. loan mutual funds, not CLOs.
Those errata are, however, beside the point. The crux of what’s misleading about the Fortune piece is what amounts to this red herring:
“CLOs didn’t implode as badly as their CDO counterparts, so investors have gravitated to them over the last two years as a way to achieve outsize returns with no measurable increase in risk.”
The statement is technically correct, as far as it goes. In fact, CDOs had demonstrably worse performance in the downturn than CLOs. But saying that this constitutes different degrees of implosion is a statement that is grossly unfair. Here’s why:
According to S&P Ratings, the cumulative default rate on U.S. CDO obligations originally rated AAA (since 1996) is 14%; those include deals that bundled sub-prime mortgage backed securities. The comparable measure for U.S. cash-flow CLO liabilities originally rated AAA, by contrast, is zero.
What’s more, of the over 6,100 ratings issued by S&P on over 1,100 U.S. CLO transactions – including investment-grade and speculative-grade instruments – 25 tranches have defaulted, and had their rating lowered to D as a result.
Based on this, S&P calculated a 0.41% default rate, or just over four tranches for every 1,000 it has rated, across the entire debt stack. In either light, the CLO default experience hardly brings to mind an implosion.
Thus, the Fortune piece – by comparing CLO default rates with those of CDOs – is, in effect, calling an ant hill a lesser peak than Mount Everest. Again, this is true in the narrowest sense of the word, but specious with regard to describing the historical default experience of CLO liabilities. – Steve Miller