Regime Switching Asymmetric-GARCH Models for Estimating Financial Risk in the Nigerian Stock Index

• Activity Number: 165 - Statistics for Business and Financial Markets
• Type: Contributed
• Date/Time: Monday, July 31, 2017 : 10:30 AM to 12:20 PM
• Sponsor: Business and Economic Statistics Section
• Author(s): Mary Akinyemi* and Georgi Boshnakov
• Companies: University of Lagos and University of Manchester
• Keywords: Markov-switching GARCH ; Financial Risk ; Expected shortfall(ES) ; Asymmetric-GARCH ; value-at-risk (VaR)

Abstract

This paper applies various Markov switching asymmetric GARCH models in estimating value-at-risk (VaR) and its coherent complement Expected shortfall(ES) on returns of the Nigerian stock index. This was done by considering a mixture of Student's-t distributions with varying variances over different time and regimes. Single regime asymmetric GARCH models were compared with their Markov switching counterparts. We found that although the Markov switching models were able to adjust for spurious high persistence found in the single regime asymmetric GARCH models. Under relative performance and hypothesis-testing evaluations, the VaR forecasts derived from the Markov-switching GARCH models were not necessarily preferred to their single regime counterparts.