Fool's Gold: How the Bold Dream of a Small Tribe at J.P. Morgan Was Corrupted by Wall Street Greed and Unleashed a Catastrophe (9 page)

Capital reserves could be cut only if banks could prove that default risk on the super-senior portion of the deal was truly negligible, and if the notes being issued by a BISTRO-style structure had a AAA stamp from a “nationally recognized credit rating agency.” Those were strict terms, but J.P. Morgan was meeting them. The implications were huge.

Banks had typically been forced to hold $800 million in reserves for every $10 billion in corporate loans on their books. Now that could be just $160 million. The CDS concept had pulled off a dance around the Basel rules. The feat was so clever that some bankers started to joke that “BISTRO” really stood for “BIS Total Rip Off,” referring to the Bank for International Settlements (BIS), which had overseen the Basel Accord.

For a period, Demchak’s team stopped transferring super-senior risk from J.P. Morgan’s books. But then Bill Demchak got uneasy. The super-senior risk was accumulating to a staggering figure, because when the bank arranged CDS transactions for clients, it typically put the super-senior risk in the deal on its own balance sheet. In theory, there was no reason to worry about that. After all, Hancock, Demchak, and Masters had repeatedly told the regulators that the super-senior risk was safe. But by 1999, the total pipeline of super-senior risk had swelled toward $100 billion. Something about that mountain of risk started to offend Demchak’s common sense. “If you have got sixty, one hundred billion, or however many billions of something on your balance sheet, that is a
very
big number,” he remarked to his team. “I don’t think you should ignore a big number, no matter what it is.”

Time and again, Demchak had battled with the “dinosaurs” in the commercial lending department, waving the risk assessment models before them as proof that the bank was mismanaging its risk. Yet even as he had evangelized about these models, he had never been tempted to think for a moment that models were anything more than a
guide.
They were exceedingly useful, if not essential, for navigating in the world of modern finance. But they were not infallible, no matter how well crafted they were. Models were only as good as the data that was fed into them and the assumptions that underpinned their mathematics.

Demchak was highly cognizant that the modeling of risks involved in BISTRO-style deals had its limits. One of the trickiest problems revolved around the issue of “correlation,” or the degree to which defaults in any given basket of loans might be interconnected. Trying to predict correlation is a little like working out how many apples in a bag might go rotten. If you watch what happens to hundreds of different disconnected apples over several weeks, you might guess the chance that one apple
might go rotten—or not. But what if they are sitting in a bag together? If one apple goes moldy, will that make the others rot, too? If so, how many and how fast? Similar doubts dogged the corporate world. J.P. Morgan statisticians knew that company defaults are connected. If a car company goes into default, say, its suppliers may go bust, too. Conversely, if a big retailer collapses, other retail groups may actually benefit. Correlations can go both ways, and working out how they might develop among any basket of companies is fiendishly complex. So what the statisticians did, essentially, was to study the past correlations in corporate default and equity prices and program the models to assume the same pattern in the present.

This assumption wasn’t deemed particularly risky, as corporate defaults were rare, at least in the pool of companies that J.P. Morgan was dealing with. When Moody’s had done its own modeling of the basket of companies in the first BISTRO deal, for example, it had predicted that just 0.82 percent of the companies would default each year. If those defaults were uncorrelated, or just slightly correlated, then the chance of defaults occurring on 10 percent of the pool—the amount that would have been required to eat up the $700 million of capital raised to cover losses—was minuscule. That was why J.P. Morgan could declare super-senior risk so safe and why Moody’s had rated so many of the BISTRO notes triple-A.

The fact was, though, that the assumption about correlation levels was just human guesswork. And Demchak and his colleagues knew perfectly well that if the correlation rate ever turned out to be higher than the statisticians had presumed, serious losses might result. What if, for example, a situation transpired in which if a few companies did default, numerous others would, too? The number of defaults that might set off such a chain reaction was a vexing unknown; maybe no chain reaction would result from a few defaults, but if ten happened—say, among big economic players—the rot might spread, destroying the entire portfolio.

Demchak had never seen that happen, and the odds seemed extremely long, but even if there were just a minute chance of such a scenario, Demchak didn’t want to be sitting on a pile of assets as big as $100 billion that could conceivably go bust. It just did not feel prudent. So he
decided to play it safe and told his team they needed to look for ways to cut their super-senior liabilities again, irrespective of what the regulators were requiring.

Taking that stance cost the bank a fair amount of money, because it had to pay AIG or others fees to insure the risk, and those fees steadily rose as the decade wore on. In the first such deals that J.P. Morgan had cut with AIG, the fee had been just 0.02 cent for every dollar of risk insured each year, or in banking terms 2 basis points. By 1999, the price was nearer 11 basis points. But Demchak was determined that the team be prudent.

Around the same time, the team stumbled on a second potentially more worrisome problem with the BISTRO concept. As the innovation cycle turned and earnings from CDS deals declined, Bill Demchak asked his team to explore new ideas for the BISTRO concept, either by somehow modifying the structure or by putting new kinds of loans or other assets into the mix. Mortgage loans were one type they decided to experiment with.

Terri Duhon took charge of the endeavor. In 1998, Demchak had asked her to run the so-called exotics book, which handled a large volume of CDS. Only ten years earlier, Duhon had been a high school student in Louisiana. When she told her relatives she was going to work in a bank, they assumed she would be a teller. Now she was managing tens of
billions
of dollars. She was trained as a mathematician, and she thrived on adrenaline—in her spare time, she rode Harley-Davidsons—yet even so, she found the thought of being in charge of all those zeroes awe-inspiring, if not a little scary. “It was just an extraordinary, intense experience,” she would later recall.

A year after Duhon took on the post, she got the word that Bayerische Landesbank, a large German bank, wanted to use the BISTRO structure to remove the risk from $14 billion of US mortgage loans it had extended. She debated with her team whether to accept the assignment, because working with home loans wasn’t a natural move for J.P. Morgan. The bank had never done serious business in offering mortgages, and on
the few occasions when it had tried trading mortgage-backed bonds, its efforts had backfired. In the early 1990s, for example, the bank had taken the rare step of hiring an outside team to trade mortgage bonds. The team had suffered such huge losses that it had ultimately been shut down. The senior J.P. Morgan management considered the experience as a salutary lesson on how difficult it was to judge mortgage risk. Duhon knew, though, that some of the bank’s rivals were starting to conduct CDS deals with mortgage risk. So the team decided to accept the assignment.

As soon as Duhon talked with some quantitative analysts, she encountered a problem. When J.P. Morgan had offered the first BISTRO notes in late 1997, the bank had had access to extensive data about all the loans it was repackaging. So had the investors, as the bank had deliberately named all of the 307 companies whose loans were included in the deal. In addition, many of these companies had been in business for decades, so extensive data was available on how they had performed over many business cycles. That gave J.P. Morgan’s statisticians—and investors—great confidence about predicting the likelihood of defaults.

The mortgage world was a good deal different. For one thing, mortgages were generally dumped into pools of debt that were entirely
anonymous,
since when banks sold bundles of mortgage loans to outside investors, they almost never revealed the names and credit histories of the borrowers. Investors had to rely on data from the lender itself about the default risks of the borrowers or the judgments of ratings agencies. Worse, when Duhon went looking for data to track mortgage defaults over several business cycles, she discovered it was in short supply.

In the second half of the twentieth century, while America’s corporate world had suffered several booms and recessions, the housing market had followed a steady path of growth. Some specific
regions
had suffered downturns. Prices in the Texas property market, for example, fell during the savings and loan debacle of the late 1980s. Yet in the period since the Second World War, there had never been a nationwide house price slump. The last time housing prices had fallen en masse, in fact, was way back in the 1930s, during the Great Depression.

The lack of data made Duhon nervous. When bankers assembled models to predict defaults, they wanted data on what normally happened
in both booms
and
busts. Without that, it was impossible to know whether defaults tended to be correlated or not, in what circumstances they were isolated to particular urban centers or regions, and when they might spread nationwide. Duhon could see no way to get such information for mortgages. That meant she would either have to rely on data from just one region and extrapolate it across America or make even more assumptions than normal about how defaults were correlated. She discussed what to do with the mathematically gifted Krishna Varikooty and the other quantitative experts.

Varikooty was renowned on the team for taking a sober approach towards risk. He was a stickler for detail who loved to get things right, and that stubborn scrupulousness sometimes infuriated his colleagues, who were urgent to make deals. But Demchak always defended Varikooty. “Once, people shouted at Krishna and made him upset, and Demchak just went ballistic,” one of his teammates later recalled. Varikooty’s judgment on the mortgage debt was clear: he could not see a way to track the potential correlation of defaults with any level of confidence. Without that, he declared, no precise estimate of the risks of default in a bundle overall could be made. If defaults on mortgages were
un
correlated, then the BISTRO structure should be safe for mortgage risk, but if they were highly correlated, it might be catastrophically dangerous. Nobody could know.

Duhon and her colleagues were reluctant to simply turn down Bayerische Landesbank’s request. The client was intensely keen to go ahead, even after the uncertainty in the modeling was explained, and so Duhon came up with the best estimates she could to structure the deal. She used the S&L data from Texas as a proxy to imagine what might happen if a disaster ever occurred to the US mortgage market as a whole. And to cope with the uncertainties, the team stipulated that a bigger-than-normal funding cushion be raised, which made the deal less lucrative for J.P. Morgan. The team also hedged its risk. That was the only prudent thing to do, though, and Duhon couldn’t see doing many more, if any, such deals. Mortgage risk was just too uncharted. “We just could not get comfortable,” Masters later said.

In subsequent months, Duhon heard on the grapevine that other
banks
were
starting to do CDS deals with mortgage debt, and she wondered how the other banks had coped with the data uncertainties that so worried her and Varikooty. Had they found a better way to track the correlation issue? Did they have more experience with dealing with mortgages? She had no way of finding out. Because the CDS market was unregulated, the details of deals weren’t available, and she had no good intel sources at the other banks. Like most of those working on Demchak’s team, she had spent her entire career at J.P. Morgan.

The team did only one more BISTRO deal with mortgage debt, a few months later, worth $10 billion. Then it dropped the line of development altogether. Years later, Duhon was stunned when she learned of the staggering volume of mortgage-based CDS deals the rest of the banking world had gone on to do.

Five years after the Boca Raton event, the J.P. Morgan derivatives team hosted a conference for its clients, at Wall Street’s Cipriani ballroom. The venue was one of the most prestigious and elegant in the financial district, a cavernous room with a classical domed ceiling, glittering chandeliers, and a marble floor the size of several tennis courts. Before the event, some of the J.P. Morgan team had worried whether they could attract enough clients to fill the grand space. Blythe Masters was determined, however, to bring in a good crowd. Soon after joining Demchak’s team, she had persuaded the group to start pooling details about all their potential customers, and she had created a list jokingly referred to as “MOAD”—as in the “Mother of All Databases”—which she and her colleagues now tirelessly worked. So many clients responded that the event was standing room only.

By October 1999, the official volume of credit derivatives deals in the market was estimated at $229 billion, six times the level just two years earlier. J.P. Morgan alone accounted for almost half of the total. The bank had not just kick-started the business; it virtually
was
the market. The J.P. Morgan clients, investors, and rivals packed into the Cipriani ballroom that day were eager to find out what the young derivatives Turks were planning next.

As the audience filed into the Cipriani, they were given a hefty, seventy-three-page tome decorated with a blurred picture of orange and white squares on a black background, not unlike a piece of artwork by Mark Rothko. Next to that was the title
The J.P. Morgan Guide to Credit Derivatives,
and inside the book, which Demchak’s team sometimes referred to, with only a faint sense of irony, as the “bible,” pages of dense text set out in exhaustive detail exactly how credit derivatives worked and the rationale behind their creation. “Until recently, credit remained the major component of business risk for which no tailored risk-management products existed,” the book solemnly declared. But now, it added, there was banking salvation. “By separating specific aspects of credit risk from other risks,” it continued, “credit derivatives allow even the most illiquid credit exposures to be transferred to the most efficient holders of that risk.”