Abstract—This paper proposes the VaR computation of non-Gaussian stochastic model. The problem of the VaR evaluation comes from the fact that it is not easy to estimate volatility. We present non-Gaussian models using stochastic volatility model where volatility is governed by logarithmic Ornstein-Uhlenbeck process. First, we show how to estimate the volatility with The Kalman filter procedure, and second we show how to extend the classical VaR techniques in an operational way using the stochastic process taking into account the asymmetric and leptokurtic distributions.

Index Terms—VaR, stochastic volatility, Kalman filter, Ornstein-Uhlenbeck process, non-Gaussian distributions

Cite: Hanen Ould Ali and Faouzi Jilani, "VaR Computation of Non-Gaussian Stochastic Model," Journal of Advanced Management Science, Vol. 2, No. 1, pp. 61-64, March 2014. doi: 10.12720/joams.2.1.61-64