PREDICT THE UNEXPECTED: THEORETICAL FOUNDATIONS FOR THE FORECASTING OF THE STOCK MARKET FIASCO USING ENTROPY OF INFORMATION SPACE

The paper contains a theoretical study and algorithm for decision the problem of forecasting the stock market fiasco caused by non-financial and other factors. The market fiasco is considered as a casual event, which is non-periodical and sudden one, and appears due to the action of many unforeseen factors. Methods of technical and fundamental analysis are useless for this problem, therefore, authors are proposing the use system analysis methodology. The authors describe a method for decision the problem by numerical calculation the entropy of the information space of search queries. The decrease in Renyi’s entropy, associated with an increase the number of search queries, containing key words of the subject area, indicates the possibility of the stock market fiasco in the near future. The paper describes an algorithm for the dynamic calculation of the Renyi’s entropy, which allows to predict rare events that are reflected in the statistics in insufficient amount to build econometric and statistical models.