[Journal:] CBN Journal of Applied Statistics [ISSN:] 2476-8472 [Publisher:] The Central Bank of Nigeria [Place:] Abuja [Volume:] 4 [Year:] 2013 [Issue:] 1 [Pages:] 89-116
This paper examines exchange-rate volatility with GARCH models using monthly exchange-rate return series from 1985:1 to 2011:7 for Naira/US dollar return and from 2004:1 to 2011:7 for Naira/British Pounds and Naira/Euro returns. The study compare estimates of variants of GARCH models with break in respect of the US dollar rates with exogenously determined break points. Our results reveal presence of volatility in the three currencies and equally indicate that most of the asymmetric models rejected the existence of a leverage effect except for models with volatility break. Evaluating the models through standard information criteria, volatility persistence and the log likelihood statistic, showed that results improved with estimation of volatility models with breaks as against those of GARCH models without volatility breaks and that the introduction of volatility breaks reduces the level of persistence in most of the models. The study recommends the incorporation of significant events in GARCH models in volatility estimation of key asset prices.
Exchange rate Volatility GARCH Unit roots Stationarity Persistence Volatility breaks Time series