Here's a pretty interesting Bloomberg News story about proprietary trading at Citigroup. Here's a schematic summary:
- The Volcker Rule bars banks from "proprietary trading" in credit.1
- But it allows proprietary trading in rates products such as Treasury and agency bonds.2
- So Citi set up a prop desk to trade agency bonds, managing over $1 billion of Citi's money.
- It's run by a woman named Anna Raytcheva, who lost billions of dollars trading agency bonds during the financial crisis.3
Obviously, some people are scandalized because people are scandalized by everything related to the Volcker Rule. And because the Volcker Rule is light on coherence. For instance, why does the Volcker Rule allow prop trading in rates? Well:
Lawmakers sought the flexibility to finance government spending and didn’t see the trading as particularly risky, said Barney Frank, who as a Massachusetts congressman helped draft the 2010 Dodd-Frank Act that mandated the Volcker Rule.
“To the extent the instruments being traded are completely secure, some of the rationale for the rule disappears,” Frank, a Democrat, said in a phone interview.
This is a nice clear category mistake: An instrument can be completely "secure" (that is: certain to be paid back), but you can still lose a lot of money trading it, because day-to-day changes in market prices are not determined solely by the likelihood that you'll get paid back. They're determined by other things too. Like, I don't know, interest rates. I mean, that's what rates trading is. It's trying to make money off the fact that instrument prices change in response to changes in interest rates. If you can make money off of that, then it stands to reason that you can lose money too.
So, I mean, I guess the Volcker Rule missed that.4 Sorry?
What else can we learn from this story? Well, for one thing, it reinforces the ways in which, for a bank, accounting is destiny. JPMorgan once owned a bunch of credit derivatives, and in 2012 they lost about $6 billion of value. This reduced JPMorgan's profits by $6 billion, because those derivatives were marked to market in its income statement. This led to no end of recriminations. There was a Senate hearing.
Bank of America once owned a bunch of agency mortgage securities, and in 2013 they lost about $6 billion of value. This reduced Bank of America's profits by about $0, because the securities were accounted for as "available for sale" and the losses were due to interest-rate moves. The recriminations were limited to a couple of news articles. There was no Senate hearing.
More recently, banks have been moving similar agency portfolios to "held to maturity" portfolios, so that they can pay even less attention to swings in their market prices.
And now there's this desk at Citi, which "uses a variety of strategies to beat returns available from simply buying mortgage bonds," which presumably involves active trading and a mark-to-market book. So if its investments lose value, that will flow through Citi's income statement, and people will freak out, as they did with JPMorgan. People are harrumphing already:
“I would have expected Citi to take a different path,” said Rossi, who left in 2009 and is now an executive-in-residence at the University of Maryland’s Robert H. Smith School of Business in College Park. The team’s existence is “surprising, and a little disappointing to be quite honest.”
I mean okay sure whatever. But Citi owns $130 billion of Treasuries and agencies in its available-for-sale book.5 If interest rates go up, those securities will lose value. So might Anna Raytcheva's $1 billion worth of agency bonds. But hers will lose, you know, much less value.6 Her desk is a "prop desk," and so risky and evil, but if you cared only about economics you might worry more about Citi's $130 billion portfolio of similar securities.
The other nice lesson of this story is that there's some sort of conservation law in which, no matter what regulators do, banks always have a constant quantity of fun. If you write a rule against fun, and that rule has exceptions, then all the fun will happen in the exceptions. If you tell banks that they can't have any more fun trading credit, they will ramp up the fun they have in trading rates. Because, if you can't have fun, what's the point in being a bank?
This oversimplifies in several respects:
- The Volcker Rule goes into effect in July 2015, though everyone kind of operates in its shadow now.
- It bars proprietary trading in lots of asset classes; credit seems like the most relevant for my limited purposes here.
- It actually allows prop trading in municipal bonds, which I guess are a credit product.
2 Or, rather, its prohibition on proprietary trading "does not apply to the purchase or sale by a banking entity of a financial instrument that is:"
(1) An obligation of, or issued or guaranteed by, the United States;
(2) An obligation, participation, or other instrument of, or issued or guaranteed by, an agency of the United States, the Government National Mortgage Association, the Federal National Mortgage Association, the Federal Home Loan Mortgage Corporation, a Federal Home Loan Bank, the Federal Agricultural Mortgage Corporation or a Farm Credit System institution chartered under and subject to the provisions of the Farm Credit Act of 1971 (12 U.S.C. 2001 et seq.);
Or some other things, including municipal bonds.
3 It's tempting to be like "hahaha Citi," but I don't think that's fair. A good chunk of the people who could plausibly be put in charge of a trading desk in 2014 lost a lot of money in 2008, so you can't hold it against any of them individually.
4 Technically there's an exception to the exception: You can't trade even Treasuries or agencies if doing so would "Result, directly or indirectly, in a material exposure by the banking entity to a high-risk asset or a high-risk trading strategy," that is, an asset or strategy that would "significantly increase the likelihood that the banking entity would incur a substantial financial loss or would pose a threat to the financial stability of the United States." It will be fun to see the Fed and other regulators interpret this rule. Perhaps in the case of this desk at Citi!
5 See page 215 of the 10-K. I'm adding "U.S. government-sponsored agency guaranteed" ($42.5bn amortized cost) plus ("Total U.S. Treasury and federal agency securities" ($87.2 billion amortized cost).
6 Probably. I mean surely her positions are a bit more exotic than the AFS book. But are they 130 times more exotic?
To contact the writer of this article: Matt Levine at email@example.com.
To contact the editor responsible for this article: Tobin Harshaw at firstname.lastname@example.org.