Abstract: in the planning work, only such a forecast is used, in respect of which it is possible to speak about significant efficiency, for the evaluation using the relative indicator of the excess income received as a result of the prognostic information. The existing methods for solving problems of predicting the value of stock dynamics are very time-consuming and require knowledge of multidimensional distribution laws (probability functionals) of all processes. In cases where the processes are different from Gaussian ones, analytical expressions of such functionals can usually not be found or are very approximate, which does not guarantee high accuracy of the analysis. Most often, the processes under consideration or the observed (measured) parameters are expressed in terms of one-dimensional process probability densities, and one or even both processes may be non-Gaussian.
The article offers a method for optimizing the forecast of stock dynamics as well as the principle of clarifying the dynamics of the cost of shares of an economic object, which allows reducing the risks of exchange operations with shares of these economic objects when selling and buying. Knowledge of the parameters and calculation based on the entropy criterion will allow owners of objects to more accurately represent their position in the stock market. The calculation using the proposed algorithm will provide reliable information about the possible risks and profitability of an economic object. The effectiveness of forecasts depends on the forecast object, the average forecast error, and the value of the forecast limit error. A set of statistics regarding fluctuations in the value of shares of an economic object associated with a particular economic object and its market of economic interactions will allow approximating their probability density by one of the options considered in the article.
Keywords: forecasting, promotions, economic object, value, criteria, risks, shares, stock exchange transactions