The Silo Effect

by Gillian Tett

  Why? Straumann put much of the blame on the bank’s top managers. These officials were too complacent and failed to ask the right questions about what was happening at the bank. But there was another problem too: silos. What the Straumann and UBS reports also showed was that the bank, like Sony, was riddled with structural silos. Warring departments failed to cooperate with each other. Crucial pieces of information were not passed across the bank. The managers at the top of UBS were not aware of what was happening in the grass roots, because the climate was so defensive that each department hugged data to itself. Worse still, the leaders operated in such a bubble, or intellectual silo, they did not ask the right questions of their staff.

  But in one respect, the story of UBS is more alarming than Sony. At the Japanese electronics group, the presence of silos had crushed innovation, and prevented it from seeing opportunities. But at UBS, the silos had prevented the bankers from seeing risks. In that sense, then, it is a cautionary tale, since it reveals a pattern that is endemic in so much of the rest of the financial world—and in many parts of the nonfinancial world too.

  THE STORY OF UBS is a particularly painful one for the Swiss government since the bank is a powerful symbol of the nation; the financial equivalent of cuckoo clocks, luxury watches, or chocolate bars. Its main headquarters are found on the historic Bahnhofstrasse in the heart of Zurich, just a short stroll from the beautiful lake that carries the city’s name, tucked into a vista of towering mountains. Unlike most Wall Street banks, the group does not operate out of any flashy skyscraper. Instead its building is made of dull gray granite that blends discreetly into the streetcar-lined street, outshone by nearby luxury watch stores and designer clothes shops. Even the marble in the lobby is discreet. The only eye-catching detail on the sober facade is a flash of scarlet from the UBS logo. It is the same shade as the Swiss flag.

  The bank was created, in its modern form, in 1998, when Switzerland’s second and third largest banks—then known as Swiss Bank Corporation and Union Bank of Switzerland—combined to create a banking behemoth.23 (In 1998, the bank had $590 billion of assets inside the investment and corporate bank and approximately $910 billion in its asset management unit, mostly in the private bank. That made it one of the largest in the world.)24 Its roots go deep into Swiss history. Swiss Bank Corporation and UBS themselves were cobbled together from numerous mergers of Swiss companies, along with a few famous American and British names: Phillips & Drew (the asset management business of Chase Manhattan Bank),25 Dillon Read (another U.S. financial group),26 and SG Warburg (the British merchant bank).

  Initially, when the modern bank first emerged, UBS seemed a stodgy, domestic entity. During the 1990s, the two banks had tried to build up overseas operations. But Union Bank had been badly burned by an investment in an American hedge fund, Long-Term Capital Management, and in the aftermath of those losses, some of the UBS managers were wary of venturing overseas too aggressively again. The two men who combined UBS and SBC—Marcel Ospel and Mathis Cabiallavetta—had different ideas. They could see that the global markets were becoming more tightly entwined and other American and European banks were grabbing new businesses, and they were determined that UBS should follow the trend.27 So, around the turn of the century, they created an expansion plan that aimed to build a business not just in Zurich, but in London and New York as well. They started in a time-honored fashion, by making a flurry of new hires.28 Between 2001 and 2004 they spent an estimated $700 million recruiting a clutch of big Wall Street names such as the former Donaldson, Lufkin & Jenrette investment banker Kenneth Moelis, Olivier Sarkozy,29 Ben Lorello,30 Blair Effron,31 and Jeff McDermott.32 The most notable of these, however, was John Costas, a flamboyant former bond trader, who was appointed head of UBS’s investment banking division in 2001.33

  Then, Costas on board, the UBS management started to look for ways to expand their bank. Compared to most other banks, UBS was unusually cash rich, since its private bank was one of the biggest in the world and pulled in a vast quantity of savings each year from ultra-wealthy clients. This financial pile potentially gave the bank a formidable arsenal. Ospel and Costas were convinced that if they could find some way to deploy this money in high-earning business, the bank would be able to rival—or eclipse—even the mightiest Wall Street names, such as Goldman Sachs, Morgan Stanley, or Credit Suisse. “This is a once in a lifetime opportunity,” Costas declared in 2002, predicting that UBS was about to become one of the biggest five banks in the U.S. markets. Ospel was even more ambitious. He declared that UBS could become one of the top three investment banks in the world.34

  Ospel and other senior managers started hunting for new businesses they could enter and spoke with management consultants such as Ernst & Young, and Oliver Mercer Wyman. The advice from the consultants was clear: if UBS wanted to grow itself quickly, and challenge the mightiest Wall Street names, it had to jump into the corner of finance known as securitization, particularly the subsection linked to mortgage bonds. This sector is distinctly specialist, if not technical.35 What it essentially does is turn loans, such as mortgages, into bonds—which are then traded between banks and other investors. These so-called securitized bonds are then often turned into new packages of securities, using derivatives too, which adds an extra layer of complexity.

  Until that point, the bankers at UBS had not considered the bank to be an expert in securitization. On the contrary, the UBS bankers had focused on the parts of the financial markets that are better known to ordinary people, such as extending loans, taking customer deposits, or trading stocks, shares, and currencies. But while the UBS bankers were not market leaders in securitization, they could see that this business line was producing juicy profits for the big Wall Street banks. So Ospel and Costas drew up plans for the Swiss giant to jump into the securitization business, confident it would deliver big returns.

  IN THE AUTUMN OF 2005, regulators from Switzerland traveled from Bern to New York to conduct their annual inspection of UBS’s American operations. It was supposed to be a routine inspection. In previous years, the key rival to UBS in Switzerland—Credit Suisse—had developed a large footprint in America, as it had earlier acquired First Boston, an American bank. But unlike Credit Suisse First Boston, UBS was not considered to have a particularly exciting American operation: its business there seemed sober, if not dull. But when the regulators started inspecting the UBS books, they noticed something striking. A few months earlier, UBS had created a department in the New York office dedicated to trading something called “collateralized debt obligations” or CDOs. This was a particularly specialized field in the business of securitization. Essentially, it revolved around the craft of taking bundles of different loans and bonds, and turning these into complex new financial products. One way to visualize this process is with the image of how a butcher makes sausages. Instead of simply grabbing a carcass and selling steaks, a butcher will sometimes take numerous different joints, chop them up, and mix them according to somebody’s taste, and then sell it inside new casings. The process of creating CDOs echoes this, in financial terms. The banks start by amassing loans they have made to customers (companies or consumers), break these down into different pieces of lending risk, mix them up, and sell them to new customers in new cases called CDOs. Like sausages, these can be blended to different customer tastes, to contain more or less risk, and higher or lower returns.

  If a casual observer had wandered into the UBS offices in 2005 and seen the desk making those CDOs, they might have thought it was a tiny, if not irrelevant, business. By that date UBS had a global workforce of 82,000 and vast departments in America that traded equities and currencies. Indeed, it had so many traders in America that it was in the process of building a cavernous trading floor out in Stamford, Connecticut, which was billed as one of the biggest such arenas for bank traders in the world. The CDO desk, by contrast, had only a few dozen employees. It was run by Jim Stehli, a veteran trader, who worked out of a UBS office in central Manhattan, near
Radio City Music Hall.36 Most people who worked in the UBS global network did not even know the CDO desk existed at all.37

  But when the Swiss regulators looked at UBS’s American books, they spotted that the tiny CDO desk was generating an extraordinary amount of activity. According to the official accounts, the bank had amassed some $16.6 billion worth of mortgage bonds, largely via that CDO desk, in the course of a mere nine months. “UBS presented the results [to us] of an internal study summarising the overall exposure of UBS investment Bank to the US real estate market,” the Swiss regulators later explained. “The study was very comprehensive and included both direct ($16.6bn) and indirect ($7.1bn) exposure (e.g., to construction firms).”38

  The Swiss regulators tried to find out why this seemingly tiny CDO desk was generating so much activity. Was this business really safe? Was the CDO desk under control? The risk managers in New York insisted it was, and cited two reasons for that. The first was that the CDO desk was only supposed to handle assets that were very safe, or securities that had been given a AAA tag from credit rating agencies. Second, the bank was only holding these CDOs in a temporary manner, so was not really exposed much risk.

  The reason lay in how the bankers visualized the CDO business; or explained it to outsiders and themselves. Before the 1970s, when a bank such as UBS conducted its business, it tended to make loans or buy assets and then hang on to these. In financial terms, these assets stayed on the bank’s books. However, since the 1970s, as the securitization business took hold, banks have tended to sell a large part of their loans to other investors, to spread their risks. The CDO business took this to a new level. According to the theory, what banks such as UBS were doing was acquiring loans (or “originating,” in financial jargon) and repackaging them into new cases, to be sold to other, outside investors. Thus if the bank held loans or CDOs on its book, via that New York desk, it only expected to keep them for one to four months. Indeed, at UBS (like most banks) the desk was called a “warehouse,” to distinguish it from other parts of the bank that were buying assets for investment reasons. “We are in the moving business, not storage business,” Robert Wolf, president of UBS Investment Bank, liked to say.39 Or as the Swiss regulators observed in a report that they wrote in late 2005: “The bank has always presented itself . . . as an organisation that consistently followed an ‘originate to distribute’ approach. . . . Under this approach the exposure arising from securitisations is only held on the bank’s own books for a short time and then immediately passed on to others.”

  Taken together, that meant there was no reason for the regulators to worry about the CDO warehouse. Or so the UBS bankers argued. The $16.6 billion of mortgage assets sitting on the bank’s books were unlikely to ever default, since most of them had a AAA tag. They had received that designation partly because the assets inside the CDO were thought to be “uncorrelated” in banking jargon, meaning that even if one of two households defaulted on their mortgages, that was unlikely to cause widespread defaults—or so the theory went. However, many of these instruments were thought to be even safer than normal AAA securities, because they were a subset of CDO called “super senior” tranches. They had that label because the structure of the CDO implied that in the unlikely event that lots of mortgages did go into default and reduce the value of those super senior CDOs, it would be other investors, not anyone holding super senior tranches, that would take the hit.

  In reality, if an anthropologist had peered into those books, he or she might have spotted some problems with these assumptions. For one thing, the loans that lay deep inside the CDOs were not actually ultra-safe at all. Many of these had been extended to the risky mortgage borrowers known as subprime creditors. When the banks assembled those loans into CDOs, they used complex financial techniques that appeared to transfer the risk of default to other investors. As a result, the credit rating agencies had given the CDO securities a AAA stamp. Many of these were deemed to be super senior, which were supposed to be even safer than normal AAA assets. But very few people outside the rating agencies or specialist CDO desks had any idea how the banks seemed to work this alchemy. Nobody knew whether those AAA CDOs were really safe.

  There was a second oddity. Whenever people such as Wolf talked about the CDO business, they described it as a “moving” business. It was all about selling CDOs to other investors. But in reality, the bankers did not have much incentive to sell all the CDOs. When the bankers created these products, they typically created several parts, called “tranches” (after the French word for slice). Outside investors were eager to buy the CDO tranches that appeared to pay high returns. However, they had little appetite for buying the safest chunk, or the super senior tranches, since these produced extremely low returns. So the super senior tranches tended to end up unsold on the bank’s books, like the unwanted bones from a butcher’s carcass. At first the bankers were worried about this. But then the traders working on the CDO desk spotted something important. When the CDOs were stored on the bank’s books, they produced a small return, which the traders could book as a “profit.” This return was very low, in percentage terms, worth just 0.1 percent a year or so. But 0.1 percent of billions of dollars was enough to provide an attractive revenue stream for the CDO desk. Since UBS, like other banks, operated with an “eat what you kill” pay structure, which paid bankers a bonus according to profits produced by their team, that gave the CDO desk a strong incentive to hold as many CDOs as they could. At other banks, there might have been a ceiling on this activity, because it was expensive to borrow money to buy loans. But the UBS bankers did not have any problems accessing cheap cash, since the private bank supplied them with as much money as required, at a price that seemed almost free. Indeed, the incentive to expand was so strong that by early 2006 Stehli’s team was not just holding onto the super senior CDOs that it was creating. It was also buying additional super senior CDOs from other banks.40 There was a sharp gap between rhetoric and reality of CDOs.

  From time to time, some of the bankers inside UBS would crack jokes about these contradictions. The traders on the CDO desk knew that they were exploiting some of the peculiarities of the system. Because the business was supposed to be a client activity, UBS did not feel the need to post big reserves against the assets sitting in this warehouse. Those assets were earning reasonable returns. However, the traders had no incentive to point this discrepancy out to their bosses, or the regulators. And to most of the bankers and regulators, the oddities in this pattern were not immediately clear, partly due to a sense of tunnel vision. Like the Kabyle village that Pierre Bourdieu studied in Algeria, twenty-first-century investment banks operated with a deeply entrenched classification system. AAA assets were considered completely different from BBB assets. Bank activities that served clients were viewed differently from “proprietary businesses,” or those where the bank took calculated gambles with its own funds. The latter was considered very risky. The former, known as client business, was not. In reality, these categories often blurred. Client businesses could be dangerous. AAA assets might not be ultra-safe. But once activities and items were defined into particular mental categories, financiers would not usually reclassify them. And what made this classification system even more rigid and powerful was that accountants and risk managers used it to measure and manage banks’ assets. So did regulators, When the Swiss financial supervisors decided how much reserve capital a bank needed to hold against the danger of losses, they did this by first dividing the bank’s assets into different categories.

  On one level, the bankers knew this was odd; the classification system contained numerous contradictions. But generally the classification system in Western banks, like that in the Kabyle village, existed at “the borders of conscious and unconscious thought,” as Bourdieu observed. Or as anthropologist Karen Ho, who studied Wall Street banks in the late 1990s, suggests, the bankers were behaving in accordance with patterns of learned behavior that seemed natural to them because of their habitus.41 In the Wall Str
eet habitus it seems natural to treat “client” and “proprietary” businesses separately. It also seems natural for the individual teams of traders to do everything to maximize their profits. That is how they get paid. Nobody had much incentive to challenge the contradictions in the status quo, or the classification system. A silo mentality ruled. After all, as Bourdieu observed: “The most powerful forms of ideological effect are those which need no more than a complicitous silence.”

  So, after looking into the bank’s books, the regulators issued their verdict. “Securitisation was not deemed to be a key risk due to the apparently purely client orientated nature of the business,” the Swiss regulators stated.42 “These [$16.6 billion] figures were not seen as a major concern by either the bank or the supervisory authorities.”

  BY THE SPRING OF 2007, some two years after UBS created its CDO desk, the senior managers in the bank’s granite headquarters building in Zurich were getting worried about their company. That was not because anybody was paying much attention to the CDO warehouse in New York. On the contrary, most of the senior officials who worked at the elegant Bahnhofstrasse office still had only the scantest idea what the CDO desk was doing. But the UBS leaders knew that whenever a financial system enjoys a period of heady expansion, banks—and bankers—tend to do stupid things. And by the spring of 2007, global markets seemed to be in the grip of a wild party. There was so much money swirling around the system that the cost of borrowing money had collapsed, and financiers were competing with each other to invest in all manner of risky ventures. Loans to subprime mortgage borrowers, or American households with a bad credit score, were booming. So were loans to finance leveraged buyouts, venture capital, and other risky corporate ventures.

  This pattern made the UBS bankers very nervous, particularly given that they liked to think of themselves as a risk-averse, prudent group. They were keenly aware that back in 1998 Union Bank had suffered embarrassing losses from its ill-judged investments in the Long-Term Captial Management hedge fund. The UBS managers were determined to avoid making similar mistakes. So the top managers held a series of meetings to discuss the risks facing the bank, to rein these in before they caused any damage.