Market risk

Most of UBS's market risk arises from the Investment Bank's trading activities. Treasury (part of Corporate Center) assumes foreign exchange and interest rate risk in connection with its balance sheet and capital management responsibilities, while the wealth and asset management operations take limited market risk in support of client business.

Trading conditions remained extremely difficult in first quarter 2008, with further contagion from the US mortgage markets and fears of a global slowdown affecting other asset-backed securities, equities and corporate credit markets. The situation was exacerbated by an acute lack of liquidity in the interbank lending market, despite moves by central banks to alleviate the situation, including reductions in some interest rates. Doubts about the financial strength of monoline insurers put downward pressure on US municipal and asset-backed securities.

As indicated in the fourth quarter 2007 report and UBS's Annual Report 2007 (in Risk, Treasury and Capital Management 2007), UBS has changed its approach to internal risk control for illiquid US residential mortgage-related exposures - US sub-prime and Alt-A residential mortgage-backed securities (RMBSs), super senior RMBS collateralized debt obligations (CDOs), and the US reference-linked note program, and related hedges. For further details on these positions, please see the "Risk concentrations" section of this report. Value at Risk (VaR) is neither an adequate measure of the risks in such illiquid positions nor an appropriate risk control tool. These risks are therefore now excluded from VaR limits and are controlled primarily by volume-based limits that reduce as positions are worked down, supplemented by targeted stress scenarios. The regulatory capital treatment has also changed, from trading book to banking book. These positions were previously the dominant contributors to interest rate VaR, including credit spread.

Largely as a result of this change, Investment Bank average, 10-day, 99% confidence VaR decreased significantly in first quarter 2008 to CHF 306 million, from CHF 665 million in the preceding period. Quarter-end VaR was also substantially lower at CHF 299 million compared with CHF 614 million at the previous quarter-end. Excluding the US residential mortgage-related, super senior RMBS CDOs positions from Investment Bank 10-day VaR for fourth quarter 2007, average VaR would have been approximately CHF 370 million lower and year-end VaR approximately CHF 260 million lower than shown in the tables below. 1-day VaR would have also been lower - by CHF 33 million on average and CHF 12 million at year-end. On a like-for-like basis, Investment Bank VaR for first quarter 2008 is therefore broadly unchanged from fourth quarter 2007 on average and lower at quarter-end.

During first quarter, updates to the historical time series increased VaR, but this effect was broadly offset by active risk reduction.

Interest rate VaR includes not only exposure to changes in the level and shape of yield curves, but also exposure to credit spreads. With the exclusion of the illiquid US mortgage-related, super senior RMBS CDOs positions, the relative contribution of credit spread exposure to interest rate VaR has been reduced. Additionally, credit spread exposure from corporate debt and some classes of asset-backed securities has been actively managed down during the quarter. Directional interest rate exposure is now a more dominant component and changes in US interest rate positions have been the main driver of variations in interest rate VaR over the quarter. The interaction of these factors resulted in a slight increase in interest rate VaR quarter on quarter after taking account of the exclusion of the US residential mortgage-related positions in first quarter 2008.

Average equities VaR remained stable from the previous quarter. The temporary increase in VaR in late December 2007 was reversed in January 2008 and period-end VaR was therefore significantly lower than at the previous quarter-end.

As in previous periods, VaR for UBS as a whole followed a similar pattern to Investment Bank VaR.

"Backtesting" compares 1-day VaR calculated on positions at the close of each business day with the revenues arising on those positions on the following business day. These "backtesting revenues" exclude non-trading revenues, such as fees and commissions, and estimated revenues from intraday trading. When backtesting revenues are negative and greater than the previous day's VaR, a "backtesting exception" occurs.

The US residential mortgage-related portfolios reclassified to banking book for regulatory capital are excluded from VaR and backtesting revenues from 1 January 2008 but are included in "all revenues" for the same period. It is therefore more meaningful to show the analysis of backtesting revenues split between first quarter and the prior nine-month period, as illustrated in the histograms below. Histograms at the bottom of page 29, comparing daily backtesting revenues with the corresponding VaR for days when the backtesting revenues are negative are also shown on this basis. The histogram located in the top right hand corner of the opposite page shows all daily revenues from businesses with trading activities, including US residential mortgage-related portfolios, and covers the 12 months to 31 March 2008.

UBS experienced a further 11 backtesting exceptions in first quarter 2008. Markets remained stressed and differential movements between asset classes that had been well correlated until mid-2007 continued to highlight basis risks. Increasing the granularity of risk representation in risk measures in order to improve the performance of the VaR model and the identification of basis risks, is an important component of UBS's efforts to enhance the market risk management and control framework. For further details, please refer to UBS's Annual Report 2007, pages 36 to 37 of Risk, Treasury and Capital Management 2007.

As an essential complement to VaR, UBS runs macro stress scenarios bringing together various combinations of market moves to reflect the most common types of potential stress events, and more targeted stress tests for concentrated exposures and vulnerable portfolios. Market risk exposure to emerging markets is also controlled by individual country and global limits based on emerging market stress scenarios.