Gretchen Morgenson has yet another article about credit default swaps in today's NYT. I don't have time to debunk all of the myths or point out all of the false statements (of which there are many) in her article. Suffice to say that at several points she confuses currency swaps with credit default swaps, which is pretty goddamn hard to do. But one statement jumped out at me, because while it's not terribly important, it's an indisputably false statement, and provides a pretty clear example of the kind of lazy and uninformed analysis that Morgenson brings to the table. She writes:
The Office of the Comptroller of the Currency reported that revenue generated by United States banks in their credit derivatives trading totaled $1.2 billion in the third quarter of 2009.No, Gretchen. The OCC does not report trading revenue by instrument. The OCC report she's referring to is the latest Quarterly Report on Bank Trading and Derivatives Activities. She got that $1.2 billion number from the chart accompanying graph 6A (on page 16), which I've reproduced here:
These numbers do not represent banks' derivatives trading revenues. Look at the title of the table: "Cash & Derivative Revenue." Right there in huge font. Notice also the note accompanying the table:
The trading revenue figures above are for cash and derivative activities.So it should be very clear to anyone who can, you know, read words, that U.S. banks did not generate $1.2bn in revenues from credit derivatives trading in Q3-2009. That Morgenson somehow missed this is simultaneously astonishing and not at all surprising. (I don't know what newspapers' policies on corrections are, but shouldn't provably false statements at least be corrected online? So that you're not continuing to knowingly mislead your readers?)
In fact, most of the trading revenue figures in the OCC report are considered unreliable. The trading revenue figures are taken from Schedule RI of banks' call reports, which directs banks to determine the appropriate risk exposure category for trading revenues the same way they make that determination in other financial reports. But banks usually don't break out revenues (or fair values) by these five risk exposure categories in other financial reports, so they're basically left to use their own internal system to make this determination. And that's where things get messy and unreliable, because every major dealer bank organizes their trading floors differently. The "credit desk" at one bank may not trade the same instruments as the credit desk at another bank. Rates desks at some banks trade derivatives and pretty much every cash instrument under the sun, while other rates desks trade only interest rate derivatives. So the way the banks break out revenues internally, by desk/division rather than by instrument, differs substantially. The end result is that the category-specific trading revenue figures in the OCC's report are basically useless.
Also useless: Gretchen Morgenson.