SPECIFICATION OF THE FOREGOING SEASONAL PRODUCT OF THE TRADE NETWORK

At the present stage, forecasting is an integral part of every business activity. Therefore, there is a need to explore new economic forecasts that are necessary to determine further development and use of resources for the most likely and effective plan variants. In the conditions of scientific and technological progress of the improvement of the economic system, the forecast becomes one of the decisive factors in the formation of the strategy and tactics of development for current and future enterprises. The article deals with the specific conditions relating to the forecasting of the seasonal product of the trading network. In particular, the forecasting models with the seasonal component have been analyzed in order to improve the results for enterprises. As prediction as one of the most important parts of the management process, without which it is not possible to develop not only tactical, but also strategic plans for enterprise development. As a result, the author identified an effective forecasting model on which the volume and inventory generated and the corresponding funds would depend on. In the first place, the successful activity of any enterprise engaged in the sale of goods in the field of sales depends on the implementation of such a management function as forecasting. But modern forecasting models do not always show good results. Therefore, it's worth analyzing modern approaches and proposing a new calculation option, or improving the already existing model, this will allow the company to lose less money and begin to effectively prepare the composition for possible loads, which will increase the efficiency of the logistics department. Therefore, the challenge is to improve the forecast for the period. Periodic fluctuations observed in time series are called seasonality. Before seasonal variation it is necessary to calculate the seasonal season. In most cases, the period is known from the context of the task, but if the period is not known before, then it can be found using the autocorrelation function. Autocorrelation function is a signal characteristic that helps to find repetitive parts of the signal and finds the carrier frequency of the signal, hidden due to imposed noise and oscillations at other frequencies. The author provided all possible seasons, which allowed to construct clear coefficients, according to which it is possible to construct a clear picture that will be as close as possible to actual sales.