Computation methods - Smoking adjoints: fast Monte Carlo Greeks

Monte Carlo calculation of price sensitivities for hedging is often very time- consuming. Michael Giles and Paul Glasserman develop an adjoint method to accelerate the calculation. The method is particularly effective in estimating sensitivities to a large number of inputs, such as initial rates on a forward curve or points on a volatility surface. The authors apply the method to the Libor market model and show that it is much faster than previous methods

The efficient calculation of price sensitivities continues to be among the greatest practical challenges facing users of Monte Carlo methods in the derivatives industry. Computing Greeks is essential to hedging and risk management, but typically requires substantially more computing time than pricing a derivative. This article shows how an adjoint formulation can be used to accelerate the calculation of the Greeks. This method is particularly well suited to applications requiring sensitivities