Option pricing in the model with stochastic volatility driven by Ornstein-Uhlenbeck process. Simulation (Q340795)

On of the promising direction of enhancement of the classical Blach-Scholes model is a construction and a research of diffusion models with volatility of risky asset governed by a stochastic process. In the presented paper a discrete-time approximation of paths of an Ornstein-Uhlenbeck process as a mean for estimation of the price of European call option in the model of financial market with stochastic volatility is considered. The Euler-Maruyama approximation scheme is implemented. The estimates for the option price for predetrmined sets of parameters are determined. The rate of convergence of the price and an average volatility when discretization intervals tighten are determined. For the case where the exact value of the price is derived the discretization precision is analysed. In the Introduction the classical Black-Scholes model is discussed. Also, the stochastic volatility modification of this model is considered. In Section 2 the definition of the model under consideration and the discretization scheme are presented. The concept of the diffusion model with stochastic volatility governed by Ornstein-Uhlenbeck process is presented. Here, the pair of stochastic differential equations \[ d S_t = \mu S_t dt + \sigma(Y_y) S_t dB_t, \eqno{(1)} \] \[ d Y_t = - \alpha Y_y dt + k d W_t, \eqno{(2)} \] where \(\{S_t: 0 \leq t \leq T\}\) is the geometric Brownian motion and \(\{B_t, W_t: 0 \leq t \leq T\}\) are Wiener processes, are considered. Also, the initial conditions of the processes (1)-(2) are given. The explicit representation of the unique solution of the Langevin equation (2), or the so-called Ornstein-Uhlenbeck process, is given. With an aim to proceed the risk-neutral setting characterised by the minimal martingale measure \(Q,\) the equations (1)-(2) are rewritten in the form \[ \begin{aligned} d S_t & = r S_t dt + \sigma(Y_t) S_t B_t^{Q},\\ d Y_t & = - \alpha Y_t dt + k d Z_t^{Q}, \end{aligned} \eqno{(3)} \] where \(B_t^Q\) are implemented Wiener processes. The Euler-Maruyama discrete-time of the solution of the Langevin equation (2) is applied to construct a Markov ghain \(\{ Y^{(n)}\}.\) In Section 3 the prices of the European call option are compared for discrete-time and continuous volatility processes. In Theorem 3.1 the order \({\mathcal O}(m^{-{\gamma \over 2}})\) of the expectation of the error, which is arising as a result of applying of the Euler-Maruyama scheme, is obtained. In Section 4 numerical results of the simulation are provided. These results illustrate how the price of the option changes with the decrease of the time step in the discrete model. In Section 5 the precision of discrete-time for the case of deterministic volatility is demonstrated. The option prices obtained for the Euler scheme with the true prices of the European call option for different sets of parameters for the case of deterministic time-dependent volatility are compared. In Appendix A the Euler-Maruyama scheme for numerical approximation of stochastic differential equations is described.