Estimation of tail Risk measures in Tehran Stock Exchange Using Generalized Multi-Dimensional Autoregressive Ranking Approach (DMS-GAS)

Subject Areas : Financial Knowledge of Securities Analysis

Seyed Ali Mousavi Sarhadi ¹ , Hosein Izadi ² , mojghan safa ³ , Mohammadreza pour Fakharan ⁴

1 - Student of Financial Management, Qom branch, Islamic azad university, Qom, Iran
2 - Department of Management and accounting, Islamshahr branch, Islamic azad university, Islamshahr, Iran
3 - Department of accounting, Qom branch, Islamic azad university, Qom, Iran
4 - Department of ACcounting, Qom Branch, Islamic Azad University, Qom, Iran

Received: 2022-11-24 Accepted : 2024-06-18 Published : 2024-06-29

Keywords: tail Risk measure, Tehran Stock Exchange, Generalized Autoregressive Ranking Approach (GAS), Time Horizons, Maximum Overlap Discrete Wavelet Transform (MODWT),

Abstract :

The main purpose of this study is to investigate the tail risk measures (VaR and ES) in Tehran Stock Exchange using the dynamic multi-scale generalized autoregressive ranking approach (DMS- GAS-1F). In this regard, using the daily data of the total index of Tehran Stock Exchange in the period 2011/03/26 - 2022/03/19 and the maximal overlap discrete wavelet transform (MODWT)algorithm, the short-run, medium-run and Long-run components of time series returns are extracted. Then, using the approach of generalized autoregressive ranking models (GAS), the tail risk measures at different time horizons are dynamically estimated and finally using the inverted wavelet transform, the final results of estimating the risk criteria based on the proposed model (DMS- GAS-1F) is provided. The results of backtests show that the proposed model has performed better in out-of-sample forecasting of tail risk measures than competing and traditional models in this field, including GARCH models and rolling window models. In addition, the results show that the use of the Maximum Overlap Discrete Wavelet Transform (MODWT) algorithm to extract information components at different time horizons has increased the predictive efficiency of the model.