Abstrakt:
Recognition of specific graphical formations such as Flag, Head and Shoulders, Wedge or Triangle in time series of stock prices is an important component of discipline known as technical analysis. These chart patterns are mainly detected by expert visual confirmation long after their occurrence. The aim of presented contribution is description of new revisited automatic detection and confirmation algorithm for Head and Shoulders Pattern (H\&SP) with neckline. The detection algorithm is based on statistical estimation of unknown parameter in nonlinear regression model and coefficient of determination evaluation. Confirmation of detection and buying (selling) signals are determinate from stock prices and neckline position in specially chosen trading days. Whole procedure is tested on selected time sections of Dow Jones Industrial Averages, where H&SP is expected to occur. Results of analysis and final discussion, presented in Case Study, show that detection and confirmation of H\&SP signalize the fundamental change in trend of stock market prices.