Browsing by Author "Majige, Selemani"

Now showing 1 - 1 of 1
  • Thumbnail Image
    Item
    Forecasting the money supply and bank deposits in Tanzania using time series techniques
    (University of Dar es Salaam, 2005) Majige, Selemani
    The study of forecasting the money supply and bank deposits in Tanzania explored issues associated with modeling, identification, estimation, diagnostic checking as well as forecasting of univariate time series of demand deposits, narrow money supply, broad money supply, and extended broad money supply using the time series techniques. The objective of this study was to assess the behaviour of money supply and bank deposits in Tanzania and to build a time series model, which could enable to forecast over the next few years using monthly data. In this study Box-Jenkins methodology was used to identify time series models for monthly data of demand deposits, narrow money supply, broad money supply and extended broad money supply in Tanzania. Unfavourable characteristics were found in the data just like in other economic time series. This necessitated a transformation of original series data to achieve stationarity of series, such as natural logarithms and first and seasonal differencing as suggested by Nelson (1973) and Stephen (1998). In the identification stage, the series were plotted and the sample autocorrelations were examined. The following models were identified and used to estimate the associated parameters: SARIMA (1,1,1) × (0,0,2)12 for demand deposits, SARIMA (2,1,2) × (0,1,2)12 for narrow money supply, SARIMA (0,1,1,) × (0,1,1)12 for broad money supply and SARIMA (0,1,0) × (0,1,2)12 for extended broad money supply. The test statistics such as Q-statistics (Portmanteau Test) or Box-Pierce test, Durbin Warson statistics, Akaike information criterion, and mean square error were applied to verify the adequacy of the identified models. The models were then used to forecast monthly values of the respective series. Since the forecast values were reasonable, the identified models can be considered as adequate for univariate short-term forecasting for the time series considered.