The Silo Effect
Page 25
But as time passed, BlueMountain’s tactics shifted in subtle ways. Between 2003 and 2007, the group had become best known for trading bundles of credit, bonds, and derivatives. It had other activities too, such as a side operation that traded equity derivatives. However, what it was really respected for was how it handled the CDO tranches that groups such as UBS were creating. That generated returns of about 10 percent a year, a level that was not dazzling compared to some other funds, but perfectly respectable.
However, the financial crisis of 2007 and 2008 delivered a painful jolt. As the credit markets gyrated wildly, the CDO business ground to a sudden halt, and BlueMountain suffered big losses on some of its holdings. In 2008 its flagship fund, Credit Alternatives, lost 6 percent of its value. Some investors withdrew their money. The following year, though, BlueMountain rebounded. In 2009 the wounded Credit Alternatives Fund rose by 37.4 percent,13 which bolstered client confidence and attracted new investment flows. The firm started to grow rapidly again and by late 2009 it had amassed around $5 billion in assets under management.
However, in the post-crisis world, the tactics that BlueMountain had used before 2007 no longer worked so well; or not in relation to CDO tranches. Banks such as UBS had lost so much money with CDO deals that they had little appetite to create new products again. Terms such as “securitization” and “super senior CDO” had become virtually taboo, and the markets trading them were drying up, reducing the amount of opportunities to exploit price gaps. So the BlueMountain team looked for other opportunities to exploit. First, they gobbled up some of the bonds and derivatives banks needed to sell at knock-down prices in the aftermath of the financial crisis. In 2011, for example, Crédit Agricole, the French bank, sold a $14 billion portfolio of assets to BlueMountain. The deal was an attractive one for the hedge fund. However it also benefited the French bank, since it needed to cut its balance sheet to comply with new regulatory rules.14 “It was a win-win situation,” David Rubenstein, the hedge fund’s European CEO, declared. “We transferred the risk to BlueMountain and helped Crédit Agricole’s balance sheet become more efficient from a risk-weighted perspective.”15
Then the BlueMountain traders hunted for discrepancies in other parts of the credit derivatives sector. In the summer of 2011, Feldstein and others noticed that something odd was happening to the price of the IG9 index, a specialised credit derivative product. Outside the world of specialist derivatives trading, not many investors knew much about the IG9 index, or the numerous other similar series that existed for different types of European and American corporate derivatives. But the reason that the IG9 index existed was that it enabled investors to have a simple way to place a bet on the health of corporate America, in much the same way that somebody might buy a security linked to the Dow Jones index of company stock prices, or the FTSE. Investors did not necessarily need to use the IG9 index to take that bet. They could also buy single name credit derivatives linked to debt issued by individual companies. If they just wanted to place a bet on the debt-worthiness of Macy’s say, or the American retailers in general, they could buy credit derivatives attached to those two specific companies. If they bought the IG9 index, they were taking a bet on 125 different companies—of which Macy’s was merely one (although technically there were only 121 companies in the index in 2010, since four the companies had gone into default since the launch of the IG9 in 2007). But many investors found it easier to deal with an index, rather than separate derivatives linked to different companies, since it offered a one-stop bet that appeared to be easy to trade. And it was widely presumed in the market that the price of the IG9 was roughly equivalent to the average price of all the different credit derivatives inside the index.
But by the summer of 2011, that assumption was starting to break down. Normally, the price of an index of derivatives should move broadly in line with the price of the constituent parts, or so finance theory might suggest. The value of the whole should equal the sum of the parts. But in reality that logic did not always play out. Just as the price of a complete CDO might diverge from the value of its component tranches, the price of a derivatives index sometimes diverged from the underling derivatives. And in the autumn of 2011, these discrepancies were becoming unusually large for the IG9 product.
The BlueMountain traders set to work trying to understand why this pattern had occurred. So did a few other traders in the credit markets. After conducting some detective work, the traders at BlueMountain realized that JPMorgan’s Chief Investment Office team was placing heavy bets with the IG9 index and other structured credit instruments. This reflected a subtle, but important, shift in policy that had taken place deep in the bowels of JPMorgan. Traditionally, the CIO’s office had been considered one of the most boring parts of the bank. It was charged with preserving the value of the money that JPMorgan held and supporting its treasury function. As part of that, the CIO team had started investing in structured credit back in 2006, supposedly to hedge the bank’s wider credit risk, and by 2008, the CIO team was running a full-fledged structured credit portfolio. Initially this was modest, totaling just $4 billion or so. However, during the course of 2011, the trading book exploded in size to $51 billion, as a team of traders in London started to take massive bets on the health of American and European companies. However, they did not generally do this by buying individual derivatives. Instead, they overwhelmingly purchased index products, such as positions in the IG9. And as the year drew to an end, their positions kept swelling in size. Indeed, by the first quarter of 2012, the CIO office’s holding of credit derivatives products had further swelled to $157 billion in size, with $84 billion worth of positions in American index products, and the rest in European indices.
It was not clear to Feldstein, or the others, exactly why these bets were taking place. Some market observers thought that the CIO’s office was trying to use its massive size to squeeze the market or overwhelm other traders to such a degree that prices would move in a certain direction, allowing the CIO to profit. As 2011 drew to a close, for example, the CIO group “bankrolled a $1 billion dollar credit derivatives trading bet that produced a gain of approximately $400m,” as a Senate report later observed.16 However, the uneven trading pattern also reflected some of the peculiarities of the rules governing the CIO. Since this department was only supposed to engage in safe trades, it was not permitted to invest in individual company credit derivatives on a large scale, since these were deemed risky. The CIO’s office was, however, allowed to invest in index products, since these were considered fairly safe. Indeed, when the CIO officials talked about the index trades, they often presented them as a hedge for the bank, or a way of insuring against losses. This created an uneven pattern for demand, since the CIO office was buying the index, but not its constituent parts. And that, in turn, was helping to distort prices. Once again, artificial boundaries and rules were creating a dislocated market.
For several months, the BlueMountain traders watched the prices in the market diverge with a growing sense of amazement. Outside JPMorgan, almost nobody knew about the scale of bets that the small team of London traders in the Chief Investment Office were placing. However, relatively few people inside JPMorgan knew about them either. These were not usually mentioned in any official risk reports, let alone company accounts. “Prior to [April 2012] the synthetic credit portfolio had not been mentioned by name in any JPMorgan Chase public filings,” as the Senate report pointed out.17 Some traders in the investment bank suspected that the CIO office was taking large bets, and they feared that these whale trades were distorting the IG9. However, they did not intervene. There had traditionally been a high level of rivalry between the CIO and the investment bank, and the CIO team did not take kindly to external interference. Meanwhile, the CIO office did not see any reason to tell the rest of the bank what was going on, even though the size of the whale trades started to breach internal risk limits by the start of 2012. “The [structured credit portfolio group’s] many breaches were routinely rep
orted to JPMorgan Chase and CIO management, risk personnel and traders,” the Senate report damningly noted. “The breaches did not, however, spark an in-depth review . . . or require immediate remedial action to lower risk. Instead, the breaches were largely ignored or ended by raising the relevant risk limit.”18 As with UBS, a tiny silo inside the mighty JPMorgan was running wild.
As weeks passed, the price distortions became more extreme. During the summer of 2011, BlueMountain had taken the other side of the whale trades, betting that economic logic would prevail—and the index move back in line with financial fundamentals fairly soon. But by January 2012, that had notably not occurred: the discrepancies between the price of the indexes and individual credit derivatives kept swelling, leaving BlueMountain sitting on big paper losses. Feldstein debated with his colleagues what to do. They could not believe that the discrepancies could keep growing indefinitely. But they did not quite understand what was going on, and they were worried about the scale of risk that their hedge fund was running. So, although other funds were starting to jump into the anti-whale trade too, or doubling down on their bets, Feldstein and his colleagues decided against raising their stake. Their position—and their potential losses—was already large. Instead they nervously sat tight and hoped that eventually the trade would reverse.
The next few weeks were tense. BlueMountain’s paper losses kept mounting. Meawhile, JPMorgan’s Chief Investment Office kept putting more muscle—and money—behind its whales trades. But then, in the spring, news of the trades suddenly leaked into the mainstream media.19 Suddenly this once-obscure corner of the markets started to attract a wave of scrutiny, and a set of opportunistic new investors dived in, taking the opposite side to the whale trade. That caused the price of the index to finally snap back, and as momentum built in the markets, this turned JPMorgan’s paper gain into a massive loss.
For a period, the CIO team tried to keep the scale of the damage a secret from the outside world. The traders in London changed the way that they valued the positions to minimize their reported losses. The senior managers at JPMorgan tried to brush off market speculation about the scale of the losses.20 But eventually, the damage could no longer be contained. A full-scale inquiry was launched, and the traders at the CIO who had been linked to the whale trades were all removed. Then the JPMorgan risk managers set about combing through the books. They quickly discovered that the losses were dramatically bigger than anyone had guessed. Indeed, they eventually swelled to more than $6 billion. That partly reflected the fact that the prices kept moving against JPMorgan as the scandal became widely known. But there was another problem too: when the auditors went in, they discovered that the CIO office had been valuing its holdings of credit derivatives in a different way from the investment bank—even when the different parts of the bank were holding exactly the same types of credit derivatives instruments on their books. Just as the sociologist Donald Mackenzie had described, different banking teams, or silos, were measuring complex products in different ways—even when they were supposed to be using the same models.21
Horrified, Dimon insisted that the investment bank take over the CIO department’s whale trades and then get rid of them. In the new post-crisis regulatory climate the bank could not afford to have a stain sitting on its balance sheet. So the JPMorgan investment bank officials eventually asked BlueMountain to unwind the bank’s credit derivatives positions—for a fat fee. For the JPMorgan bankers it seemed to be an easy way to quickly and quietly resolve the whole saga: since they already knew Feldstein well, they trusted his fund and knew that it was big enough to swiftly get the whale trades off the bank’s books. But for BlueMountain the deal was doubly sweet. It had already made money by betting against the whale trades. Now it was being paid again; indeed the fees that BlueMountain earned by helping JPMorgan to get rid of its embarrassing problem actually ended up being larger than the profit it earned on the anti-whale trades. It was a striking sign of how the power balance on Wall Street was shifting and how nimble hedge funds could sometimes get the upper hand over the mighty banks. And a year later BlueMountain enjoyed another symbolic coup: Jes Staley, the head of JPMorgan investment bank and a man who had been viewed a possible successor to Dimon, left the bank to join BlueMoutain—and work alongside Feldstein.
AS TIME PASSED, THE BlueMountain team moved into other fields, looking for new ways to apply their bucket-busting ideas in other corners of the market where prices were being distorted by artificial or rigid boundaries. They became increasingly interested, for example, in how institutions divide financial flows into separate asset classes, such as equities, bonds, and loans. Investors normally take these distinctions for granted and assume they are almost inevitable, if not natural. Shares, after all, have fundamentally different features from bonds, say, and as a result these different products tend to be handled by entirely different teams of traders and analysts inside banks and investment companies. But what would happen, Feldstein and Siderow wondered, if you tried to look beyond those different boundaries? If you analyzed bonds and shares as a whole, instead of using the fragmented pattern that normally shaped how investors analyzed financial assets, would investments look different? Was it possible to look across the entire “capital structure” (to use the term that bankers typically employ when they describe the totality of the financial channels that companies use to raise money)?
Some banks had tried to combine categories in that way. Back in the early years of the twenty-first century, Candace Browning, then the head of securities analysis at Merrill Lynch, announced that she wanted to break down the long-standing split between analysts who studied equity markets and those who looked at bonds. She tried to force the U.S. bank’s 500-odd analysts working in the different specialist fields of equity and fixed income to collaborate with each other. “We just worked in our various silos [and] I consciously wanted to change that,” she explained. “[I thought] if we all start communicating and sharing our resources better, not only would we be more efficient but we would give our clients a better product and employees would feel more connected.”22 Or as Yaw Debrah, an analyst who worked in the team that dealt with convertible bonds, observed: “At Merrill Lynch and probably at other large firms . . . equity [analysts] worked in equity, debt worked in debt, derivatives worked in derivatives, et cetera. People did not do much cross-asset-class research because it was very difficult to coordinate.”23
In 2005, the Merrill Lynch research team produced several pieces of analysis on companies in the American auto and cable industries that combined insights from equity and bonds specialists. They were groundbreaking. Some individual teams started collaborating on a day-to-day basis. In London, one group produced what they called a “Dequity report,” a pioneering research report that compared the price of high-yield bonds, derivatives, and equities. This was popular among hedge funds, since the initiative started at a time the leveraged buy-out business was booming in European markets, causing the price of companies’ bonds and shares to swing in some unusual ways. “Because of these [LBO] dynamics, both equity and credit people in sales, trading and research at the firm started to seek out each others’ inputs,” observed Jon Gunnar Jonsson, then a young Merrill Lynch analyst who worked on the Dequity report. “On the CDS side, we started building frameworks that could identify mispricing between the shares and bonds (or CDS) of the same company. Corporate finance theory was of little help to us, so we had to invent everything.”
But these collaboration initiatives soon foundered. The different analysts were so specialized in their areas of knowledge that it was time-consuming for them to communicate, and to translate their concepts and methods. “Many very bright equity analysts don’t know a lot about the fixed-income side. They’ve been able to do their jobs well without ever really having to get their hands dirty with the rest of the capital structure,” Michael Herzig, head of Merrill’s equity research marketing, observed. “They’ll know what the bond rating is and they’ll know how much debt ther
e is, but will they know the complexity of the underlying products?”24 The mental split went hand in hand with a physical separation. “It’s really quite hard to work with someone on an ongoing basis if they sit on a different floor from you,” said equity analyst Jonathan Arnold. “I sit next to a retail analyst on one side and have an airline person outside my door, [but] the debt guys are on a different floor.”25
The biggest obstacle to collaboration, though, was the pay structure. Analysts were only incentivized to promote trading in their particular corner of the markets, since bonuses were paid according to how each product group performed. The clients of the bank, such as pension funds, were organized into rigid silos too. In theory, the idea of collaboration sounded sensible for the Merrill Lynch analysts; in practice, few people had much incentive to make it work on a long-term basis.
However, Feldstein and Siderow thought—or hoped—that precisely because the silos at banks were so rigid, they might find some good opportunities by taking the opposite tack and adopting a diverse perspective. From its inception BlueMountain had employed researchers and portfolio managers who looked at bonds, loans, credit derivatives, and equity derivatives. However, in 2010, the hedge fund made a more concerted effort to expand into equities. They hired new equity experts and asked them to sit with the credit team and swap ideas about trading strategies and investment opportunities. “We gather everyone into a room together, and by drawing on diverse perspectives we get to a richer research process and generation of ideas,” explained Marina Lutova, a credit portfolio manager at BlueMountain. Or as David Zorub, an equity portfolio manager, who sat near her, echoed: “The idea is to generate investment opportunities that you will not necessarily see if you were just evaluating them from an equity or credit perspective. We are agnostic about which piece of the capital structure we invest in—equities, bonds, and loans. It is just a question of the fundamental view, and of relative price.”