Since JPMorgan Chase announced a $2 billion loss on derivatives on 10 May 2012, the blogosphere has been abuzz with tales of the so-called London Whale, also known as Voldemort. While disclosure from JPMorgan itself remains limited, a number of sources in the hedge fund community have laid out their theories on what happened. They wasted no time taking action: As media reports of the London Whale’s positions leaked out over the past several weeks, an increasing number of hedge funds were taking the other side of his trades, forcing losses onto JPMorgan’s books.
How did this happen? Why? Let’s take a closer look.
The London Whale is a trader by the name of Bruno Iksil, who earned his nickname by betting huge sums of money in the derivatives markets from his home base in London. As one hedge fund manager recently quipped, “Bernanke is to Treasuries what the London Whale is to derivatives.” But to point out that the London Whale jumped the shark with these trades would be an understatement (he is now expected to leave the bank).
Let’s digest what we know thus far about the trade. Iksil had three discrete components to his trading strategy, which I will call packages A, B, and C as follows:
(A) JPM purchased credit default swaps (CDS) on high-yield bonds in which JPM would make money if high-yield bonds went down.
(B) JPM wrote substantial amounts of CDX.NA.IG.9, which is a basket of CDS on investment-grade bonds from 121 different issuers.
(C) JPM bought CDS on investment-grade bonds.
Package A (long CDS on high-yield bonds) could in fact act as a hedge on a loan book, but it is unknown how directly or indirectly related the derivatives are to its actual loan book. It is likely that the basis risk is high given that high-yield bonds are not a mainstay of JPMorgan’s commercial loan book. In any event, these trades went against Iksil as the economy appeared to show some improvements.
Recognizing that package A was performing poorly, Iksil created package B. Package B was possibly intended to act as a hedge on package A, but it is closer to the truth to call it a discrete proprietary trade. Writing CDS on investment-grade bonds is intended to take advantage of the improving economy that Iksil thought he recognized. Whether or not he was right about an improving economy is a separate issue, as the trading of CDS (or any security, for that matter) can move counter to an investor’s hopes in the interim. While it is theoretically possible that JP Morgan was trying to hedge some future obligations (letters of credit, future loan growth, etc.) for which the firm is naturally short, JPMorgan Chase CEO Jamie Dimon’s exit from this position suggests otherwise.
It appears that once Iksil realized that the market was moving against package B, he sought to make up for it with trades on package C, rather than unwinding package B. It is unknown how much basis risk is introduced by package C, relative to either package B or to JPMorgan’s actual loan book. Nevertheless, all of this is like a dog chasing its tail, and that shouldn’t happen with true hedges.
In its simplest form, a hedge directly corresponds to some underlying asset. For instance, if I own 100 shares of stock in Apple Inc. (AAPL) and want to hedge my downside risk, I can buy a put at a strike price of, say, $500. If my shares in AAPL fall below $500, I can sell the puts and collect the difference. However, if I had bought a put on Google Inc. (GOOG) to hedge my ownership of AAPL shares, then I might run into the risk that the stock prices of GOOG and AAPL may not move in tandem. It’s the same thing with JPM’s more exotic trades in credit indexes that are supposed to correspond to their loan book in some way. This phenomenon is known as basis risk.
If the trading sequence outlined here is accurate, then Iksil compounded the basis risk with each subsequent tranche of the trade. But it doesn’t stop there. Because these trades use unique instruments, not only do they have basis risk, but also each instrument has its own liquidity risks. And this is where it gets interesting: Iksil’s trading in package B (writing CDX.NA.IG.9 index contracts) was so large that his trading activity created a material gap between the price of the index and the sum of the prices on the underlying CDS. In fact, the aggregate notional value of this index (CDX.NA.IG.9 ) grew by more than 50% in just three months — hence Iksil’s nickname, the London Whale. At one point in the second quarter, buying CDS on the whole index was cheaper than buying CDS individually on all 121 companies it comprises ($121,000 vs. $131,000 on the same $10 million worth of bonds). Iksil’s trading pushed around the hedge funds on the opposite end of these trades. Rumor has it that the fund managers affected were so infuriated by the adverse movement, that the London Whale became the object of their ire. Once they deduced the size of his positions, they determined that he would have inadequate liquidity to exit his position. In short, they smelled blood . . . and pounced.
Looking at the pay-off table, we get a glimpse into the risk-reward matrix facing Iksil on Package B:
Does this remind you of the debacle at MF Global? How about Amaranth Advisors? What about Long-Term Capital Management? As illustrated in the table above, there is a reward opportunity for getting the underlying thesis right. So, if this is a prop trade that went wrong . . . so what! But if this is supposed to be a hedge, it’s a little more troubling. At this point, it is still possible that Iksil’s thesis that credit losses will improve due to an improving economy could turn out to be true. However, that reward lies at the end of some indefinite timeline, during which a trader must have the wherewithal to survive adverse changes in liquidity.
Dimon has called these trading losses a failed hedge. But how exactly does one “hedge” a book of commercial loans — JPMorgan’s traditional banking business — by writing protection on other companies’ paper, thereby gaining loss exposure to these other loans? I understand the arguments for common factor exposure and historical correlations. Still, economic history is replete with examples of cause and effect being different than correlation. Calling the trades complex, poorly monitored, and poorly understood is one thing. Calling the trades for what they actually are is quite another. If the information we now have is accurate, there were no hedges in package B; it was prop trading. And package C was pretty clearly used to paper over losses on package B. As we have noted here at CFA Institute, there is nothing wrong with prop trading per se, but it should be done inside a broker dealer, thereby immunizing the bank and taxpayers from losses.
If the information we have is not accurate, then I look forward to a full and fair disclosure of the trades and JPMorgan’s activities around these trades.
As we now know from Dimon’s press conference, even the vaunted JP Morgan, with its “fortress balance sheet,” could not stand the adverse changes in liquidity.
Maybe it’s time to challenge the notion of a fortress balance sheet, but I’ll save that for another post.