Forecasting Weekly Auction of the 91-Day Treasury Bill Rates of Ghana

Abstract

A Treasury Bill is a short term investment product offered by the Bank of Ghana on behalf of the government. An investor in Treasury Bills lends his/her idle funds to the government at a fixed interest rate for either 91, 182 or 364 days. In this study we employed univariate time series Autoregressive Integrated Moving Average (ARIMA) model to forecast government 91-Day Treasury Bill rates for the next eighteen (18) weeks in Ghana in 2014 using Box Jenkins approach. The results showed that ARIMA (2,0,1) model was appropriate for modelling the weekly auction of 91-Day Treasury Bill rates with a maximum log likelihood value of 246.42, and least AIC value of-482.83, AICc value of-482.14 and BIC value of-470.23. An ARCH-LM test and Ljung-Box test on the residuals of the models revealed that the residuals are free from heteroscedasticity and serial correlation respectively. Therefore investors can plan their future investments by considering the future behavior of the 91-Day Treasury Bill rates for next eighteen (18) weeks.