The Performance of Gaussian and non Gaussian dynamic models in assessing market risk: The Implications for risk management. Issue 13 (3rd July 2019)

• Record Type: Journal Article
• Authors: Nawaz, Faisal; Shahzad, Faisal; Ur Rehman, Ijaz; Shujahat, Muhammad; Hyder, Shabir; Al Barghouthi, Sameer
• Is Part Of: Communications in statistics. Volume 48:Issue 13(2019)
• Page Start: 3360
• Page End: 3376
• Publication Date: 2019-07-03
• Subjects: Gaussian models -- non Gaussian dynamic models -- value-at-risk models -- financial risk management -- market risk; Mathematical statistics -- Periodicals; Mathematics; Statistics; 519.2
• DOI: 10.1080/03610926.2018.1476707
• Languages: English
• ISSNs: 0361-0926

Abstract

Abstract: The developing markets are more volatile, unstable illiquid, and more prone to the external shocks. The non Gaussian VaR model gives more accurate risk models than Gaussian VaR models. Hence, the purpose of this study is to test if and how non Gaussian VaR models are comparatively better fit for risk modeling of developing markets than the Gaussian VaR models. The study measures the market risk for the daily closing price of Karachi Stock Exchange 100 index over the period of 2004–2016. The evaluation of VaR models suggests that non Gaussian dynamic model outperformed the Gaussian VaR models in developing markets.