International Journal of Engineering, Science and Mathematics

  • Year: 2017
  • Volume: 6
  • Issue: 8

Multivariate GARCH models for stock indices volatility on Nairobi Securities Exchange

  • Author:
  • Ngassima Fanny Laure1, Joseph K. Mung'atu2, MbeleBidima Martin2
  • Total Page Count: 16
  • DOI:
  • Page Number: 359 to 374

1Pan African University Institute of Basic Sciences, Technology and Innovation (Pauisti) Nairobi, Kenya

2Department of Mathematics (Finance)-Jomo Kenyatta University of Agriculture and Technology (Jkuat), Nairobi, Kenya

Online published on 19 April, 2019.

Abstract

In the financial sector, volatility is one of the important aspects that need special attention as far as risk management is concerned. A multivariate GARCH model is presented together with its univariate specifications. This paper reviews the substantial literature on specifications, estimation, and evaluation of the MGARCH models. The quasi maximum likelihood technique is expanded to allow for estimation of GARCH-type models and is applied to the MGARCH models. Therefore, empirical results suggest that the best multivariate GARCH model is revealed to be the DCC model which dominates the others with respect to the likelihood values.

Keywords

Multivariate GARCH, Quasi maximum likelihood, Univariate GARCH