Local Hölder exponents approach for the prediction of russian stock market critical points

The paper is devoted to elaboration a new specific indicator based on the local Hölder exponents. This indicator has been used for forecasting of critical points of financial time series. The suggested approach is based on the hypothesis which claims that before market critical points the dynamics of financial time series is radically changed, namely time series became more smoothly. The approach has been tested on the stylized and real market data. It has been shown that it is possible to forecast such critical points of financial time series as large up and down movements and trend changes. On this base a new trading strategy has been elaborated and tested.