1: \begin{abstract}

2: In usual stochastic volatility models, the process driving the

3: volatility of the asset price evolves according to an autonomous

4: one-dimensional stochastic differential equation. We assume that the

5: coefficients of this equation are smooth. Using Itô's formula, we

6: get rid, in the asset price dynamics, of the stochastic integral

7: with respect to the Brownian motion driving this SDE. Taking

8: advantage of this structure, we propose

9: \begin{itemize}

10: \item[-] a scheme, based on the Milstein discretization of this SDE, with order

11: one of weak trajectorial convergence for the asset price,

12: \item[-] a scheme, based on the Ninomiya-Victoir discretization of this SDE,

13: with order two of weak convergence for the asset price.

14: \end{itemize}

15: We also propose a specific scheme with improved convergence properties

16: when the volatility of the asset price is driven by an Ornstein-Uhlenbeck

17: process. We confirm the theoretical rates of convergence by numerical

18: experiments and show that our schemes are well adapted to the multilevel

19: Monte Carlo method introduced by Giles \citeyear{Giles1,Giles2}.

20: \end{abstract}