The paper investigates an issue of the accuracy of energy consumption forecasts for power supply companies, which are an intermediary between the wholesale energy market and the end user. Therefore, the purchased amount of energy should be equal to the predicted one with the least error. In case of any deviations, there is a problem of unsold amounts of energy or its shortage, which leads to financial losses, as the companies have to purchase or sell excessively the purchased energy at the price that is unprofitable for them in the wholesale market. Planning and monitoring the process of energy distribution are essential to achieving accurate energy consumption forecasting, which is an important component of all measures to optimize energy costs. The application of the statistical methods underlying the mathematical forecast models, which describe the process under study, will allow reducing the time for making managerial decisions and increasing the accuracy of energy consumption forecasts, thereby affecting the cost reduction in the power supply company. © 2018, Econjournals. All rights reserved.