Time series analysis on the effect of infrastructural development on economic development
Abstract
Time series analysis and forecasting has become a major tool in many applications in measuring economic development. The aim of this paper examines the trend at which GDP grow over a period of the study, models Auto Regressive Integrated Moving Average (ARIMA) and forecasts GDP receives from infrastructural development in Uganda. Secondary data with quarterly GDP release from 2008q1 to 2021q4 from UBOS was used. Data was analyzed using STATA statistical software. The data obtained was realized to be a trend with upward slope of GDP received from construction sector with the lowest and highest value of 371 billion shillings and 2353 billion shillings respectively. The data was detected to be non-stationary by a gradual slow decay with p-value of (P> |t| = 0.450) and first different unit root test was done hence the dataset became stationary at P> |t| = 0.000 where d was confirmed to be 1, thereafter Autocorrelation function (q) was determined as 1 and lastly partial Autocorrelation function was used to determine p (p = 4). ARIMA (4,1,1) was considered as the best model with significance C, AR and MA to be ¾, sigma SQ was 116.9167 and log likelihood was -324.266. White noise was also detected by Portmanteau test with P > |t| = 0.9848.
In conclusion from the time series analysis, GDP forecast was made a realized that the level of GDP will increase sharply in the next years of 2021q4 to 2024q4. This will then follow with a gradual increase in the country’s general GDP hence exhibiting a trend.