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 (7 page)

For several weeks, Masters made endless phone calls to London from New York as she and Donaldson—together with a phalanx of lawyers—worked out legal terms for the deal. In most sectors of finance, well-established rules governed deals. But the credit derivatives concept was so new that they had to be crafted on the fly; Masters was making history as she went. Nobody quite knew what the fees paid to the EBRD should
be or even quite what to call this “product.” Eventually though, the deal was done, and it was dubbed a “credit default swap,” which, as the business spread, was shortened to CDS.

Hancock and the team were jubilant. The Exxon deal showed that a substantial credit derivatives contract could be made to work. And it was not the only one. As Masters was cutting her deal with the EBRD, Robert Reoch, a British banker who worked on Winters’s team in London, cut a deal with Citibank asset management that transferred the risk attached to Belgian, Italian, and Swedish government bonds. However, it was one thing for a bank to arrange a few isolated deals; it was quite another to turn those deals into a full-blown business as lucrative as Hancock was aiming for. So, as 1995 got under way, the group went into overdrive to attack the key obstacles to making the concept fly on a large scale.

One of those was convincing their internal commercial lending team, as well as regulators, that the concept was sound. This would open up a wide pool of loans on the bank’s books to use for making swaps. They also had to figure out a way to “industrialize” CDS deals, so that they could do many more of them much faster. One other key thing the team wanted to investigate was if, by removing the risk of its loans in this way, the bank would be allowed to lower its capital reserves. That was one of the most important potential payoffs, as a good deal of reserve cash would then be freed up for use in making profits. That was the dream: credit derivatives would allow J.P. Morgan—and in due course all other banks, too—to exquisitely fine-tune risk burdens, releasing banks from age-old constraints and freeing up vast amounts of capital, turbo-charging not only banking but the economy as a whole.

Behind the scenes, Masters and Demchak started visiting US regulators. Two main institutions held responsibility for monitoring this activity: the US Federal Reserve, in New York and in Washington, and the Officer of the Comptroller of the Currency (OCC), in Washington. Neither of them had spent much time studying the idea of credit derivatives. After all, the concept had not even been invented when the Basel Accord, or any American financial rules, had been written. But that didn’t mean they wouldn’t weigh in with strict views if J.P. Morgan started doing lots of deals. Masters and Demchak posed the crucial question: if banks used
credit derivatives to shift their default risk, would the regulators let them cut their capital reserves?

The signs looked promising. At the start of the decade, regulators had grappled with the savings and loan disaster and had learned firsthand the dangers of banks concentrating loan risk, one of the big mistakes the S&Ls had made. Many regulators were consequently thrilled to learn that tools could be developed to
disperse
risk. Indeed, one senior American regulator was so enamored of the idea that when he heard about what Masters was doing, he telephoned her to learn more. “When I read the details, it seemed to me this was one of the best innovations I had ever seen. It was just a wonderful idea!” he later said.

Sure enough, in August 1996, the Fed issued a statement suggesting that banks would be allowed to reduce capital reserves by using credit derivatives. Masters, Demchak, and Hancock were thrilled.

Even as that battle was being waged, Demchak and others on the team threw themselves into an internal lobbying campaign to sell the concept to their colleagues, particularly the loan officers. In some respects, that proved harder than dealing with the regulators. During the late 1980s and early 1990s, the wild antics of the swaps team had provoked unease among the older members of the bank’s commercial lending arm. Those more traditional bankers now reacted with even more horror when Demchak declared he wanted to overhaul the way that commercial lending was done.

To bolster his case, Demchak unveiled a flood of supportive data. It was drawn from a crucial new twist on the original VaR idea that Weatherstone had developed. During the early 1990s, the quantitative experts inside J.P. Morgan had refined their tools. The first version of VaR was concerned primarily with measuring market risk. However, by the mid-1990s, J.P. Morgan analysts were using similar ideas to analyze the risk of untraded loans, too, and then compare those numbers to the risks attached to bonds and other assets. The basic idea was to look at all the assets that a bank held on its books and work out which parts of the bank were producing good returns, relative to the risks—and which were not. The implications of this analysis looked alarming. By 1995 it seemed four fifths of the bank’s capital was tied up in activities that typically earned
less than 10 basis points of return for the bank each year, meaning just 0.1 cent for each dollar used. Areas of the business that generated much higher profits, such as derivatives, were often short of capital. The bank’s capital simply wasn’t being used as profitably as it should be, let alone in a manner that would enable the bank to hit a self-imposed target of a 20 percent annual return on equity. “We have to change the way we do business!” Hancock declared over and over again, convinced that if the bank didn’t make this change, it faced a slow death.

The harder Demchak and Hancock pressed the case for reform, though, the more recalcitrant some of the commercial loan officers became. They had spent their careers evaluating loan risks, and they still highly valued the merits of relying on relationships with firms and reputations, in addition to the math Demchak was relying on to produce his data. Demchak’s fervor didn’t help. Though he was most often easygoing, he found it hard to suffer fools. The more they refuted his arguments, the more biting Demchak’s comments became. “These guys are
dinosaurs
!” he wailed to his team in fury. The loan officers retaliated by dubbing Demchak “The Prince of Darkness,” because he seemed so intent on launching otherworldly schemes. Hancock was considered the king of these dark plans.

For months, they were at an impasse. Finally, in July 1997, an unexpected twist broke the deadlock. In the middle of that year, a financial crisis erupted in Asia, and the commercial lending group suffered painful losses on loans across the region. That raised the pressure on the bank’s new CEO, Douglas A. “Sandy” Warner, who had replaced Weatherstone in 1995, to take some decisive action to improve the bank’s profits. It was becoming increasingly clear—just as Hancock and Demchak had long complained—that the bank’s level of profitability was lagging behind its rivals’, and these losses were a body blow. Warner decided to throw his weight behind the value of credit derivatives business.

He gave Hancock responsibility for managing not just the fixed-income business, but the commercial lending department, too. That was a radical move for staid J.P. Morgan, since almost no other Western bank had ever tried to combine lending, bonds, and derivatives into a single group before. Then Hancock handed responsibility for managing the
loan book to Demchak. The Prince of Darkness was now in charge and able to remold the bank’s credit risk in line with all his dreams.

However, one more obstacle still stood in the way of the team unleashing its revolution, and it was a daunting one. They still had to find a way to process a high volume of deals rapidly; to
industrialize
the CDS trade, transforming it from a cottage industry into a mass-production business.

The crucial, last piece of the puzzle that fell into place went by the strange name “BISTRO” (although bankers would later give the idea an even stranger tag, “synthetic collateralized debt obligations”). This brainchild emerged from months of heated debate and experimentation. By the mid-1990s, Hancock’s group had two views of how to make credit derivatives work large-scale. In London, Bill Winters was inclined to try to create what bankers call a “liquid market” in credit derivatives. That would entail finding a way to make credit derivatives as easy for clients and investors to buy and sell as stocks, and might even require setting up an exchange. “Pile ’em high and sell ’em cheap!” Tim Frost, one of the young traders in Winters’s group, sometimes quipped. He was utterly convinced that a mass-market approach could be found for derivatives. He also had the “can do” spirit to drive that dream.

Demchak’s team in New York, however, preferred to focus on a different idea for turbo-charging the market. That was, essentially, to bundle lots of deals together, pooling all of their risk, and then to create derivatives carved out of that whole pool, rather than swapping loan by loan.

The key question for the team was, How could you bundle a whole host of loans together, extended to lots of different companies with different credit histories and business prospects, in a way that investors could feel confident about the level of risk they were taking on in buying a slice of the total? If investors couldn’t analyze the specific risk of default for all loans in the pool, one by one, as the EBRD could do in insuring Exxon’s default risk, how could they be assured? Their solution to that puzzle came by linking derivatives technology to a technique known as
“securitization.” Experiments along those lines had already been bubbling for a couple of years at Swiss Bank Corporation, NatWest, and Chase Manhattan. The BISTRO plan essentially stole some of those ideas—but repackaged them with such crucial new twists that it transformed the field.

The securitization concept that lay at the heart of this breakthrough was even older than the swaps idea. It had first cropped up in a significant manner back in the 1960s and 1970s, when some banks started selling off mortgage loans to outside investors in an effort to diversify their portfolios. Those who bought them could make a good profit in the mortgage business without needing all the infrastructure required to originate loans. It was a win-win.

Making these sales was time-consuming, however, as investors wanted to scrutinize the details of the loans to guard against the risk of default. So bankers came up with the idea of bundling up large quantities of loans in packages. That would spread the risk of any problematic loans out over the whole bundle, so that if any borrowers with mortgages in the bundle did default, that loss would be covered by the profits made on the rest of the loans.

Over time, bankers realized that they could use the cash flow from the mortgage payments being made on that bundle of loans to make a tidy extra profit. If they issued securities, such as bonds, they could back those securities with the cash flow from the mortgage payments—for example, making the regular payments to bondholders from that cash—and overall, they’d make money from both the mortgage payments and the sales of the securities. These became known as mortgage-backed securities—hence the term
securitization
—and the business boomed.

Then, those trading in these securities got the crucial idea that Demchak’s team now jumped on. They realized that they could divide the securities they sold into several “tranches,” each with a different level of risk, and of return, for the investor. The highest risk and highest return were called “junior,” the middle level were called “mezzanine,” and the lowest risk and lowest return were called “senior.” The idea was that if defaults on any mortgages in the bundle did occur, those losses would be
charged against the junior-level securities first. If losses were so high that they weren’t covered by the junior securities, then the mezzanine-level investors would take the additional hit.

Those who invested in the senior-level securities would almost surely never suffer losses, because the chances of so many defaults happening at the same time were extremely slim. Just as in a flooding house, where the water floods the cellar first while the roof stays safe, losses from defaults would flood the junior and mezzanine levels—erasing those bondholders’ expected earnings—and the senior bonds would always be safe, except, that is, in the event of a truly major cataclysm. Due to their higher risk of loss, the junior and mezzanine notes paid proportionately higher returns, while, due to their extremely low risk, the senior notes paid quite low returns.

The key attraction of this bundling, multitiered approach was that investors could choose the level of risk they wanted. And since investors were usually willing to pay a little more for the sheer convenience of someone else tailoring their level of risk, banks could often sell the complete set of notes for more than the total value of the mortgage loans. The concept was akin to a pizzeria that takes an $8 pizza, cuts it into eight pieces, and sells each piece for $1.25. Customers will sometimes pay more to buy just the amount and flavor they want, whether of pizza or of risk.

In the 1980s, bankers took the idea that had been used to “slice and dice” mortgages and applied it to corporate bonds and loans. Demchak and his team, though, then took this a step further and applied it to credit derivatives. The idea was that, instead of grabbing a portfolio of different mortgages and selling investors a stake in it, they would instead sell a bundle of credit derivatives (CDS) contracts that insured somebody else against the risk of default. In financial terms this was equivalent to taking thirty different home insurance contracts, bundling them together, and persuading a bigger consortium of outside investors to underwrite the risk that losses might affect those thirty homes.

As with the early mortgage bond deals, though, the investors who were underwriting that “insurance”—or that pool of CDS contracts—could choose their level of risk. Investors who wanted to roll the dice
could agree to pay out the first wave of claims that might hit if, say, a few contracts went bad. Investors who wanted more safety might underwrite the mezzanine—or in-between—level of risk. That was similar to paying up on a collective insurance scheme when losses got bigger than, say, $5,000 but not higher than $100,000. The “safest” part of the scheme was the senior tier, where investors would be forced to pay the cost of defaults only after the claims—or losses—had become so widespread that all the other investors had been wiped out. There were thus different tranches of risk.