THE EX POST ANALYSIS OF THE FORECAST ACCURACY OF LIQUID FUEL SALES FOR DAILY DATA
THE EX POST ANALYSIS OF THE FORECAST ACCURACY OF LIQUID FUEL SALES FOR DAILY DATA

Summary/Abstract: The study presents the results of the daily retail sales forecast volume of liquid fuel in one petrol station. Time series econometric models, describing two types of the periodic fluctuations of a week cycle and a 12-month cycle, are used for the purpose of the forecast construction. Moreover, zero-one (dummy) variables concerning the appearance of holidays as well as pre-holiday and post-holiday periods are used as one of the “explanatory” variables. The ex post analysis of forecast error accuracy in total and in disaggregation into weekdays and months is conducted.

• Details
• Contents