Abstract: This work contains a review of GARCH models families, where BEKK-MGARCH model is presented and applied to 4 years data of Rwanda exchange market. 1632 daily observations average of buying and selling Rwanda exchange rate in the period from Friday 12/03/2010 to Saturday 26/07/2014, to model the time varying variance-covariance in USD/Rwf, and Ksh/Rwf exchange rate series in Rwanda exchange market. Time varying variance-covariance have been constructed with the models (BEKK-GARCH) significant at lag 4 using AIC, BIC criteria and QMLE method have been used to estimate parameters. R, Eviews and RATS software have been used for data analysis. Results shows that exchange rate market responds more to shocks arriving from other exchange rate markets (usd/rwf to Ksh/Rwf and vis versa) than it does to its own shocks (usd/rwf), it confirm our hypothesis of existential of cross-market spillover effects in the studied areas. We have worked with two series (USD/RWF and KSH/RWF), while Rwanda Exchange market have also many other major currencies, we propose to future researchers to incorporate many others majors currencies like EURO/RWF, USH/RWF, as they may have an important impact on time varying variance covariance model.
Keywords: Exchange rate, BEKK-GARCH model, Volatilities, Forecasting.
Title: Modeling Time-Varying Variance-Covariance for Exchange Rate Using Multivariate GARCH Model “Case Study of Bivariate BEKK-GARCH Model of Us Dollars and Kenyan Shillings Vis a Vis Rwandan Francs”
Author: Nsabimana Innocent, Joseph Kyalo Mung’atu
International Journal of Thesis Projects and Dissertations (IJTPD)
Research Publish Journals
