Evaluating the efficiency of GARCH models in estimating the conditional risk value in Tehran Stock Exchange Accounting data approach

Predictions are extremely important for a better decision-making. Uncertainty in decision making makes investors always seek to assess and estimate risk to minimize potential losses. Conditional risk value (CvaR) is considered as a comprehensive measure of risk that has been considered a useful tool in recent years. Due to the characteristics of capital market data, not all models will be able to make accurate predictions, and among the multitude of models, only the models which make predictions can correctly explain this market. In this study, according to the existing theoretical foundations and using Delphi model and analysis and review of experts, first the accounting variables in the financial statements are effective in predicting the conditional risk value, then the data of accepted companies are used. In Tehran Stock Exchange during 2012- 2018, we evaluated the capability of GARCH and Markov index models in predicting conditional risk value as a criterion for predicting coherent risk. The results showed that the estimates made with the GARCH model (1, 1) are closer to reality with the distribution of T-Student.