A new higher-order weak approximation scheme for stochastic differential equations and the Runge-Kutta method (Q964684)

The authors construct a new algorithm for the weak approximation of stochastic differential equations. A new higher-order approximation scheme is constructed for a broad class of stochastic differential equations. For the realization of this scheme, an ordinary differential equation (ODE)-valued random variable is constructed whose average approximates the given stochastic differential equation. It is constructed by using the notion of free Lie algebras. From this random variable, an ODE itself can be drawn at one time. The approximating random variable can be approximated by the Runge-Kutta method for ODEs. In a numerical experiment, this is applied to the problem of pricing Asian options under the Heston stochastic volatility model. The algorithm is compared with some other methods and is significantly faster.