Open AccessASSESSING THE FORECASTING PERFORMANCE OF ARIMA AND ARIMAX MODELS OF RESIDENTIAL PRICES IN ABUJA NIGERIA
Author(s) -
Musa Abubakar Alkali
Publication year - 2019
Publication title -
asia proceedings of social sciences
Language(s) - English
Resource type - Journals
eISSN - 2663-6638
pISSN - 2663-662X
DOI - 10.31580/apss.v4i1.528
Subject(s) - autoregressive integrated moving average , bedroom , quarter (canadian coin) , sample (material) , econometrics , box–jenkins , economics , statistics , geography , mathematics , time series , chemistry , archaeology , chromatography
This paper compared the out of sample forecasting ability of two Box-Jenkins ARIMA family models: ARIMAX and ARIMA. The forecasting models were tested to forecast real estate residential price in Abuja, Nigeria with quarterly data of average sales of residential price from the first quarter of year 2000 to the last quarter of year 2017. The result shows that the ARIMAX forecasting models, with macroeconomic factors as exogenous variables such as the household income, interest rate, gross domestic products, exchange rate and crude oil price and their lags, provide the best out of sample forecasting models for 2 bedroom, 3 bedroom, 4 bedroom and 5 bedroom, than ARIMA models. Generally, both ARIMA and ARIMAX models are good for short term forecasting modelling.