EN
Credibility of prognoses is of crucial importance for the processes of economic organization management. It is not only a problem of management boards of huge corporations, but also small economic entities. Prognoses are required by individual daily decision-making processes too. Authors of scientific work emphasize poor creditworthiness of prognoses. They foresaw neither the crisis on the New York Stock Exchange in 1986 nor the crisis on the financial markets in 2008. The consequences of the last crisis have negatively influenced economic systems of all countries in the world as prognosis tools which were used so far failed. The work is an attempt of showing tools solving the task of future description in cases of large and unforeseeable direction dynamics of the phenomenon, the changes of structure of the prognostic model as well as the occurrence of extreme events. On basis of empirical investigation results it was showed that a credible solution is possible. Namely an assumption on normal schedule of density function of probability of occurrence was repealed and a credible result was obtained replacing it with an assumption on fractal schedule formulated based on Mandelbrot sets.