Time Series Modeling and Forecasting of Monthly Average Exchange Rate of Nigerian- Naira and United States-dollar
Love Cherukei Nnoka *
Department of Statistics, Captain Elechi Amadi Polytechnic Rumuola, Port Harcourt, Nigeria.
Stephan Nlerum
Department of Physics, Rivers State University, Port Harcourt, Nigeria.
*Author to whom correspondence should be addressed.
Abstract
This work considered the time series modeling and forecasting of monthly average exchange rate of the Nigeria -Naria (NGN) and United States -Dollar (USD). The data for this work was obtained from the Central Bank of Nigeria spanning from 2003 to 2024. The time plot of the monthly average exchange rate of both currencies indicated an upward rise in the Dollar and a relatively reduction in the Naira.
To achieve this aim, a powerful timeseries forecasting model known as autoregressive integrated moving average (ARIMA) was employed.
The Exchange rate data was non stationary at level but achieved stationarity after first difference using augmented dickey-fuller (ADF) unit root test via E-views 12.
The strategies for model specification or identification as recommended by Box and Jenkins (1976) were adhered to in this work and ARIMA (8,1,2) was selected as the most parsimonious model among the other five potential or tentative estimated models. This ARIMA (8,1,2) satisfied the residual diagnostic test because of its invertibility and covariance stationary behaviors, this is to say that it had relatively small Akaike’s Information Criterion (AIC) and Standard Error of Regression (SER), higher number of Statistically Significant Coefficients and relatively high adjusted R-Squared.
It is hoped that the forecast graph in Fig. 8 of this work would help the Government, Business Operators and Policy makers to invest wisely, plan their budgets, make inform decisions regarding monetary and fiscal policies of the country. The wandering away of the Original Exchange Rate Data from the Forecasted Series stood amazing and a big lesson for Nigerians to strengthen their currency at all cost.
Keywords: Time series, exchange rate, ARIMA model