Skip to main content

The CDX And The Whale

  • Author:
  • Updated:

The OCC report on bank derivative activities is rarely what you would call a laugh riot but I enjoyed that the 2Q2012 one released today gives the London Whale a belated sad trombone:

Commercial banks and savings associations reported trading revenue of $2.0 billion in the second quarter of 2012, 69 percent lower than the first quarter of 2012, and 73 percent lower than in the second quarter of 2011, the Office of the Comptroller of the Currency reported today in the OCC's Quarterly Report on Bank Trading and Derivatives Activities.

“Trading revenues were weak in the second quarter,” said Martin Pfinsgraff, Deputy Comptroller for Credit and Market Risk. “While both normal seasonal weakness and reduced client demand played a role, it was clearly the highly-publicized losses at JPMorgan Chase that caused the sharp drop in trading revenues.” Mr. Pfinsgraff noted that JPMorgan Chase reported a $3.7 billion loss from credit trading activities, causing the bank to report an aggregate $420 million trading loss for the quarter.

How big a deal Whaledemort is depends on your denominator: compared to JPMorgan's assets, or even its revenues, he's a drop in the ocean, but his misadventures in credit derivatives did wipe out two-thirds of all derivative trading revenues among all US banks. And he's a good enough excuse to talk about a random assortment of other credit-derivative-trading things from the last few days. First is a neat Bloomberg article (appears to be terminal-only now) about CDX NA HY 19:

Trading in credit-default swaps has slowed so much that Markit Group Ltd. for the first time ran out of active contracts to create its latest benchmark tied to U.S. junk bonds as Federal Reserve stimulus measures sap demand for insurance against companies missing debt payments. ... The gauge, updated every six months to maintain consistent contract maturities and to switch out companies no longer meeting index rules, is adding swaps tied to CIT Group Inc., Calpine Corp. and a Charter Communications Inc. unit that aren’t among the active contracts reported by the Depository Trust & Clearing Corp.

The hope appears to be that their index inclusion - plus, to be fair, their recent bankruptcy emergence - will drag single-name swaps into being for those names. But the necessity of this maneuver suggests that credit indexes are more popular and single-name CDS is less popular than it used to be, which seems to be the case, at least relatively speaking:

The net amount of protection bought or sold on individual issuers worldwide has dropped 8.3 percent in the past year to $1.1 trillion, according to DTCC data compiled by Bloomberg. Wagers on the two most recent junk credit swaps indexes declined 3.9 percent to $34.6 billion during the same period.

And that doesn't tell the whole story because you can increasingly get index protection away from CDX; Bloomberg has also pointed out that "Exchange-traded funds are poised to overtake credit derivatives by year-end as a way to speculate on junk bonds."1

What is driving this? Well you can partition the space with hypotheses:

  • People don't want to get long single names via CDS. Why? This seems straightforward: there's lots of physical inventory. If you want to be long a company's credit, just buy the bonds. In the past you might have preferred to write CDS for funding reasons (you don't have to stump up the cash), but now that everyone wants derivatives to be better collateralized you kind of do, and anyway with rates really low the cash isn't that hard to stump up.
  • People don't want to get short single names via CDS. Why? Bloomberg's answer is "because nobody thinks any companies will default any more because of the Fed," which, I guess? That seems like a price rather than quantity issue. But, sure; traditionally one reason you might buy CDS on single names because you own the credit and want to hedge your credit risk, and perhaps you are now so emboldened by the lack of defaults, and/or so averse to paying for protection out of your miserly yields, that you're foregoing CDS you would once have bought.
  • People want to get short all the bonds via CDX or ETFs. Why? Well, because you think the macro outlook is bad, the Fed won't save us all, and there will be more junk-bond defaults than are priced in and you don't want to do the credit work on particular names. Or, because you do want to do the credit work on single names, and then buy their bonds, but you want to remain market neutral so you hedge by shorting the broad market.
  • People want to get long all the bonds via CDX or ETFs. Why? Well, because you think the macro outlook is good, the Fed will save us all, and there will be fewer junk-bond defaults than (the very few that) are priced in and you don't want to do the credit work on particular names. Or because you do want to do the credit work on single names, and short them individually, but you want to remain market neutral.2

So all of these are plausible macro stories. But here's something else:

In a world of shrinking bond yields, many mutual funds have found a way to make themselves look better. Their secret: Invest in riskier bonds but continue to measure their performance against benchmarks composed of safer investments.

At least 187 bond funds, including those run by big asset managers such as Allianz SE's Pacific Investment Management Co., Oppenheimer Funds Inc. and Putnam Investments, are handily beating the benchmarks against which they compare themselves, according to an analysis by investment-research firm Morningstar Inc. for The Wall Street Journal.

Many of them have done so by investing in high-yield corporate bonds, mortgage-backed securities and emerging-market debt, among other securities, according to Morningstar.

Of course, all active fund managers try to beat the benchmark. But concerns can arise when fund managers invest in bonds or other instruments in a proportion that is far off the benchmark. That makes it all the more important, experts say, for investors to keep tabs on their funds' holdings.

I don't have CAPM hard-wired in my brain, so my first reaction is that outperforming your benchmark is not "looking better" but "doing better." That said, though, I also will believe you when you sing me a song about uncompensated risk and Sharpe ratios. I will particularly believe that song when you point out, as the Journal does, that average bond-fund above-benchmark performance was 61bps this year, and was negative (49bps) last year, sort of hinting that this outperformance is pretty pure beta and destined to go horribly wrong at some point.

But more broadly. One of my pet obsessions is the interaction between, on the one hand, the need for someone to pick investments, and, on the other, the apparent moral imperative of indexing. There's some imaginable world where what I want in a bond fund manager is a guy or gal who does lots of his or her own credit analysis, picks good bonds to invest in, and makes a lot of money. That world is, sort of, the 1980s. There's another world, the world we live in, where I want a bond manager who modestly outperforms a benchmark while remaining closely correlated to it so that I can fit his or her fund into a mathematically constructed portfolio with certain defined correlations and exposures. In that world, outperformance with tracking error may be worse than underperformance with strong correlation; knowing what macro variables I'm tracking is more important than idiosyncratic alpha.3 And when my bond manager says "I got you some extra return using emerging market bonds, hope you like it," I get mad at him.

But there has to be a tipping point in indexing; if everyone else in the market is stock-picking, or here bond-picking, then you will on average outperform them by indexing, but if everyone else in the market is indexing then you're screwed. If bond managers are mostly making their own judgments about what bonds look "investment grade," and then choosing the yieldiest ones, then you get prices that reflect the perspectives and research of people with a stake in getting prices right. If bond managers are mostly buying whatever's in the index, then you get prices that reflect size and credit ratings.

Deep down I don't really read any moral judgments into Whaledemort; he made some bets that went right, then he made a bet that went wrong, and that's that and JPMorgan will get over it. But I've half-ironicallyentertained the notion that the great sin of Whaledemort was that JPMorgan, the biggest and bestest bank in the land, was putting hundreds of billions of dollars of its balance sheet into credit indexing rather than lending money to companies that it thought were good credits. Not because that's starving the world of credit,4 but because it's starving the world of judgment: instead of JPMorgan credit officers deciding where JPMorgan should take credit risks, the Markit index of large liquid CDS names made that decision for them. If you believe that, then Whaledemort's loss is a fitting comeuppance for JPMorgan - and, apparently, for the whole derivatives market.

OCC Reports Second Quarter Trading Revenue of $2.0 Billion [OCC]
Swaps Traders Scrape Bottom in Fed’s New World: Credit Markets [Bloomberg Terminal]
ETFs Overtaking Swaps for Junk-Bond Speculation: Credit Markets [Bloomberg]
Funds Leap Beyond Their Benchmarks [WSJ]

1.Incidentally, this article points out that there are 13.1mm shares of JNK currently out on borrow, meaning a $500mm notional short interest, which ... is ... just a bit less than the $26 billion in net notional of CDX.NA.HY.18 outstanding. So as a way to get short junk credit that seems insufficient. But on the long side, sure, you can get your broad-based high-yield credit exposure by buying JNK from your local discount broker instead of messing around with ISDAs to write CDX.

2.But this is less compelling than in the previous bullet because it would create offsetting short single-name positions, either physical, which is harder to do than being long physical, or via CDS, which is not reflected in the more steeply declining single-name CDS.

3.Using "alpha" in its colloquial sense of "beta."

4.Because I have an unexamined belief that all long credit exposure is fungible, and writing credit derivatives creates credit in the same way that buying bonds does? I have little confidence in this belief though.