Mardle, D. W.2016-11-142016-11-141993https://hdl.handle.net/10182/7580This study holds as is its basic thesis that a well constructed econometric model containing an error-mechanism will provide superior forecasts. Such a model is developed within the methodology espoused by David Hendry and the London School of Economics. In addition, a method of constructing this model from data collected at two different frequencies is demonstrated. The model is used to predict known values representing the monthly observation of domestic electricity sales from a single supply authority within New Zealand. The performance of this model is compared with the results obtained from three other commonly used forecasting models i.e. exponential smoothing, ARIMA and VAR using the criteria of root mean square prediction error, Theil’s inequality coefficient and the decompositions therof. According to these criteria the exponential smoothing model produces the best forecasting results.enhttps://researcharchive.lincoln.ac.nz/pages/rightseconometric modellingtime series analysisARIMAVARElectricity Corporation of New Zealanderror correction mechanismforecastingdomestic electricity sectorelectricity forecastingelectricity consumptionThesisDigital thesis can be viewed by current staff and students of Lincoln University only. If you are the author of this item, please contact us if you wish to discuss making the full text publicly available.ANZSRC::140205 Environment and Resource EconomicsANZSRC::140303 Economic Models and ForecastingQ112852825