Comparison of Parametric and Semi-Nonparametric Methods for Investment Risk Assessment in the Stock Market: A Case Study of the Main Indices of the Tehran Stock Exchange

Articles in Press

Document Type : Original Article

Author

Assistant Professor of Economics, Department of Economics and Social Sciences, Nahavand Higher Education Complex, Bu-Ali Sina University, Hamedan, Iran,

10.22075/jem.2025.38232.2018

Abstract

This article aims to assess the risk of the main indices of the Tehran Stock Exchange by introducing a semi-nonparametric approach. The performance of this approach is evaluated against various parametric models using backtesting techniques and different risk measures, including Value at Risk, Expected Shortfall, and Median Shortfall. Moreover, the Value at Risk at 99% and 97.5% confidence levels is employed as an alternative to the Value at Risk at the 99.5% level and the expected shortfall at the 97.85% level. The empirical results, obtained for six key indices of the Tehran Stock Exchange-namely TEPIX, the First Market index, the Second Market index, the Industry index, the Mali index, and the Farabourse index, using daily data from 14/8/2007 to 3/6/2025, indicate the superior performance of parametric approaches such as the Student's t and skewed Student's t distributions over the Gaussian and the semi-nonparametric SNP approaches, as well as the Gram-Charlier expansion, in out-of-sample prediction of risk measures, providing a more conservative risk assessment. The outperformance of parametric approaches in evaluating the key indices of the Tehran Stock Exchange can be attributed to the fact that these indices exhibit smooth, non-wavy tails without extreme jumps at their tails-features that are typically captured by semi-nonparametric models.

Keywords

Journal of Econometric Modelling

Articles in Press, Accepted Manuscript
Available Online from 03 November 2025

Files

Share

How to cite

Statistics