9 August 2006

Estimation of the alpha coefficient in production scale

CD Financial Technology Ltd (CDFT) has performed a production scale procedure for the estimation of the alpha coefficient of the credit-risk-exposed trading book derivatives positions of a leading financial institution. The estimation case involved over 5 000 counterparties and nearly 100 000 contracts. Both the evolution of credit events of the counterparties and the market events were included in the full path Monte Carlo simulation analysis. Counterparties were treated as separate risk factors in simulations with no pooling applied. The execution of the immense calculation task was made possible by CDFT s recently launched R/V Platform.

The advanced approach of Basel II guidelines on the minimum capital requirements of credit-risk-exposed derivatives contracts included in the trading book is based on the procedure known as estimation and validation of the alpha coefficient. In the procedure the quantitative significance of joint effects of market and credit events on the risk of the portfolio is sought after by the estimation of the alpha coefficient. The coefficient is obtained by computing over 12 month time horizon the EPE (Expected Positive Exposure) from pure credit event simulations and the 99.9 per cent confidence level exposure from full Monte Carlo simulations involving simultaneous sampling of both market risk and credit risk events, and by dividing the latter exposure by the former.

The main challenges of the alpha estimation procedure are the potentially very large number of risk factors included in the simulations, the need for path simulation technique implicit in the setup of the problem and the potentially rather large number of relatively complex derivatives contracts included in the respective trading book. Computationally the alpha estimation is possibly the single most demanding area of the entire Basel II guidelines.

For further details please contact sales@cdgroup.fi.