Hybrid Concepts of Long-Term Estimates for Value at Risk

There is a growing demand for models which enable to measure and assess the risk in long-term horizons (sometimes more than 2 years). The practical demand for such models is required by the institutions which manage the investments and retirement funds. In the paper the theoretical aspects of risk assessment methodology with the use of <i>Value at Risk (VaR)</i> were presented. In this method in order to estimate the long-term <i>VaR</i> limits the hybrid model which is the optimum mixture of random walk and mean reversion was used. The application of the presented methodology was exemplified by the estimation of long-term predictions for <i>VaR</i> limits for stock prices.