مینو بخش محمدلو

We consider the hedging problem in a jump-diffusion market with correlated assets. For this purpose, we employ the locally risk-minimizing approach and obtain the hedging portfolio as a solution of a multidimensional system of linear equations. ‎This system shows that in a continuous market, independence and correlation assumptions of assets lead to the same locally risk-minimizing portfolio. ‎ In‎‎‎ addition, we investigate the sensitivity of the risk with respect to the variation of correlation parameters, this enables us to select the more profitable portfolio. The results show that the risk increases, with increasing the correlation parameters. This means that to reduce risk it is necessary to invest in low correlated assets.

The multidimensional exponential Levy equations are used to describe many stochastic phenomena such as market fluctuations. Unfortunately in practice an exact solution does not exist for these equations. This motivates us to propose a numerical solution for n-dimensional exponential Levy equations by block pulse functions. We compute the jump integral of each block pulse function and present a Poisson operational matrix. Then we reduce our equation to a linear lower triangular system by constant, Wiener and Poisson operational matrices. Finally using the forward substitution method, we obtain an approximate answer with the convergence rate of O(h). Moreover, we illustrate the accuracy of the proposed method with a 95% confidence interval by some numerical examples.