Accelerated ensemble Monte Carlo simulation

Traditional vanilla methods of Monte Carlo simulation can be extremely time-consuming if accurate estimation of the loss distribution is required. Kevin Thompson and Alistair McLeod show that the ensemble Monte Carlo method, introduced here, significantly outperforms unbiased Monte Carlo simulation, in terms of both accuracy and speed

Accurate assessment of the portfolio loss distribution is of great practical importance in portfolio risk measurement, active portfolio management and structured credit trading. Whether it be in assessing the likelihood of unexpected portfolio losses, the determination of economic capital1 or the valuation of tranches of structured portfolios (for example, collateralised debt obligations), the accurate evaluation of the loss distribution is critical. However, because tail events are, by