Intervention Analysis of Daily Yuan-Naira Exchange Rates
Ette Harrison Etuk & Alapuye Gbolu Eleki
Department of Mathematics, Rivers State University of Science and Technology, Port Harcourt
Department of Statistics, Port Harcourt Polytechnic, Port Harcourt, Nigeria
Email: ettetuk@yahoo.com
ABSTRACT
Time series of daily Chinese Yuan – Nigerian Naira exchange rates from 1 May 2016 to 28 October 2016 shows a very slight negative trend until 20 June 2016 after which there is an abrupt increase and then a fairly level trend. This calls for intervention. The pre-intervention data displays a negative slope and it is non-stationary. Its first difference has an overall horizontal trend and is adjudged as stationary. Following its autocorrelation structure an MA (2) is fitted to these first differences of the pre-intervention data. The residuals of this ARIMA(0,1,2) model follow a Gaussian distribution with mean zero and this is an indication of model adequacy. Differences between the forecasts of this model and the observations in the post-intervention period are used to estimate the intervention transfer function. The coefficients of this function are significant and the intervention forecasts closely agree with the corresponding observations. Therefore the intervention effect is significant. The fitted intervention model is hoped to be a basis for managing this situation and possibly helping to proffer a due solution.
Keywords: Yuan, Naira, exchange rates, intervention analysis
