Of all the banks to be caught up in the implosion of America’s subprime market, none has caused more surprise than UBS. How did a Swiss bank whose core business is the staid discipline of wealth management come so spectacularly to lose $38 billion betting on American mortgage-backed assets, battering its core capital and share price? “It’s breathtaking,” says the head of investment banking at another European bank. Shareholders, who were out in force at the bank’s annual meeting in Basel on April 23rd use slightly riper language.
The mystery of how UBS (latest nickname: Used to Be Smart) got into this mess is being resolved. On April 21st the bank released a summary of an internal investigation demanded by the Swiss Federal Banking Commission into the causes of the write-downs. The investigation was conducted by 20 lawyers from UBS, and their 400-page report is now being chewed over by the regulator. Rivals should read it too. Like one of Tolstoy’s unhappy families, UBS is unhappy in its own way; but the lessons from its sorry tale apply to all.
The report gives three broad explanations for the bank’s woes. The first was the investment-banking arm’s preoccupation with growth. Another was the reliance of the control team on flawed measures of risk. A third was the culture of the bank.
Start with those growth plans. Many had assumed that Dillon Read Capital Management (DRCM), a hedge fund set up by UBS in 2005 and closed in 2007, was the primary culprit for the write-downs; in fact, it contributed only 16% of the red ink spilt up to the end of last year. The more pernicious effect of DRCM was to deprive UBS of some of its most experienced people and to distract its senior managers at a time when the investment bank was pursuing helter-skelter expansion.
The push for growth was concentrated in fixed income, where UBS particularly lagged behind its competitors. The goal was to climb the league tables by expanding in areas such as structured credit and commodities. The effect, says the report, was to grow too fast and to emphasise revenue at the expense of risk. The bank’s collateralised-debt obligation (CDO) desk in New York responded with aplomb, structuring ever more CDOs of mortgage-backed securities for sale and keeping the supposedly safer tranches of CDOs on its own books as a source of easy profit.
Where revenues could be boosted, they were. The CDO desk concentrated on riskier “mezzanine” CDOs, which generated higher fees but suffered heavier falls in value when markets seized up in August. Cheaper hedging strategies based on buying protection on just a tiny proportion of the bank’s “super senior” (least risky) positions tended to win out over more effective but pricier ones, such as insuring the lot. The end result: a desk that numbered just 35-40 people at its peak was responsible for write-downs of around $12 billion in 2007, two-thirds of the total. The other main source of write-downs is almost as staggering: it comprised asset-backed securities bought as part of the bank’s liquidity reserve.
If the bank’s business leaders overlooked risk, its risk controllers miscalculated it. Confidence in the AAA ratings on CDOs explains the decision to hedge only 2-4% of many super-senior exposures. Those same reassuring ratings also led to more generous treatment of CDO exposures in the bank’s value-at-risk (VAR) calculations, a way of working out the maximum loss that it was likely to suffer. Liquidity was simply assumed, enabling assets to be placed in the bank’s trading book, where they attracted a lower capital charge. Unforgivably, neither the CDO desk nor their risk handlers made efforts to analyse the quality of underlying assets.
Worst of all, the belief that the bank’s hedges were foolproof led to their being netted to zero—the positions simply did not show up in the VAR numbers or in many risk reports. This focus on net rather than gross exposure was not unique to UBS (remember Société Générale?) but its effects were particularly harmful. The investment bank’s bosses only realised the depth of the hole they were in in late July 2007; the chairman and chief executive were given their first comprehensive view of the bank’s subprime exposures on August 6th, by which time it was too late to do much about it.
Probing questions could and should have revealed the extent of the risks that UBS was taking. Concerns were aired at various times in 2006 and 2007. The bank’s top brass was sufficiently attuned to the deterioration in the American housing market to have raised it in September 2006. But the report says that they were fobbed off by assurances from the investment bankers that all was well. Proposals from the bank’s treasury in early 2007 to cap the level of the investment bank’s illiquid assets also came to nothing.
There is no suggestion that anything untoward was going on. Assurances that risks were being properly managed were given in good faith, says Rupert Jolley, the UBS managing director who led the investigation: “The culture of the bank was to rely upon each other’s word.” But there was also a clear incentive to set aside any doubts as long as revenues were rising.
The report only deals with write-downs up to the end of last year. Nonetheless its conclusions implicitly raise awkward questions about the bank’s new leaders. If the culture of the bank was at fault, then can an insider such as Peter Kurer, who was confirmed as chairman at the annual meeting, fix it? The report forlornly noted the “reactive” appointment of the investment bank’s leadership team to replace those who had left to join DCRM; could the same be said of Mr Kurer, who was plucked from the position of general counsel to replace Marcel Ospel, the former chairman? One large shareholder describes the combination of Mr Kurer and Marcel Rohner, the newish chief executive, as “terrifyingly weak”.
The report also notes that subprime exposure jostled unsuccessfully with several other items on the agenda of group-level meetings (leveraged finance got plenty of attention, in contrast): that will reinforce the doubts of those who think that the bank has become too complicated to manage. Perhaps most worrying of all for battered shareholders is the implication of the investigation’s findings for UBS’s remaining exposures to the American housing market. There is scant reason to assume that the $16 billion-worth of Alt-A positions still on the balance sheet in March were researched or hedged any more effectively than their subprime sisters.
The rest of the industry can hardly rest easy: Credit Suisse, which has done a lot better than its local rival, still announced a first-quarter loss on April 24th. UBS got more things wrong than most but the traps it fell into will be familiar to its peers. Lots of other investment banks measure their status via league tables and seek to bulk up where they are weakest: Mr Rohner told shareholders that UBS no longer aims to “offer everything to everyone”. Compensation and funding structures that fail to distinguish “alpha”, or skill, from simple carry trades are widespread. The flaws in VAR, and the emphasis on net rather than notional exposures, are also known hazards. The investigation found no evidence to suggest that regulators criticised the way UBS managed its risks. The Swiss banking commission sits in judgment now. It might usefully have done so earlier.