Direct Search Algorithm for Numerical Solution of GARCH(1,1) Problem

Problems of estimation and forecasting of volatility are of considerable interest in various applications of economy and finance. In this paper, we propose an open program code (by C ++ language), which implements a robust heuristic algorithm for the numerical evaluation of the parameters of the linear model GARCH (1,1), with provision for its machine-dependent aspects. The program can be easily implemented by any programming language, does not require third-party mathematical procedures, so as selection of the initial values of the parameters for the model volatility. A numerical example is presented.