By Patricia Jackson of the Bank’s Regulatory Policy Division.
As part of their efforts to improve their risk control, the major banks are developing new statistically based tests to measure some of the risks they face. Although they are re-examining the risks in traditional lending and borrowing activities, progress has so far been greatest in the measurement of the position risk in securities and derivatives trading books. This article reviews developments in both areas, and compares the two main types of test being developed for trading books - value at risk models and ‘stress tests’. It also looks at the way that the value at risk models are influencing the development of international capital standards.
The main recent developments have been:
- Banks have used statistical techniques to look at the risks in different parts of their trading books for some time, but a number are now using more sophisticated, value at risk (VAR) models and ‘stress tests’ to look at the risks in the whole trading book. Large securities houses are developing these approaches in a similar way.
- In ‘traditional banking business’ (mainly lending and its deposit funding), the most significant change has been in the management of the embedded interest rate risk. Most large banks now manage this in their trading books, enabling it to be hedged actively.
- The growing sophistication of some banks’ measurement of their overall trading risks has led the Basle Committee on Banking Supervision to consider allowing them to use their internal VAR models to determine the capital required to back their trading positions.