EN
This article presents one of the most promising methods to predict the sudden changes in the stock market, namely the theory of log-periodic oscillations. The authors, in addition to the above theoretical basis method referring to the theory of complex systems and critical phenomena, show empirical evidence of the effectiveness of this approach in predicting both the bust in the stock market, as well as in the resources market. The examples (based on the analysis and forecasts) shown in the research confirm that self-similar log-periodicity with a parameter contraction lambda approx. 2 is able to properly describe the dynamics of the stock exchange on different time scales. What is more, the prediction, indicating a reversal of the uptrend in the stock market in September and October in 2009, was shown.