Stock market trading rule based on pattern recognition and technical analysis: Forecasting the DJIA index with intraday data

Autores UPV
Revista Expert Systems with Applications


This work presents empirical evidence which confronts the classical Efficient Market Hypothesis, which states that it is not possible to beat the market by developing a strategy based on a historical price series. We propose a risk-adjusted profitable trading rule based on technical analysis and the use of a new def- inition of the flag pattern. This rule defines when to buy or sell, the profit pursued in each operation, and the maximum bearable loss. In order to untie the results from randomness, we used a database com- prised of 91,307 intraday observations from the US Dow Jones index. We parameterized the trading rule by generating 96 different configurations and reported the results of the whole sample over 3 subperiods. In order to widen its validity we also replicated the analysis on two leading European indexes: the German DAX and the British FTSE. The returns provided by the proposed trading rule are higher for the European than for the US index, which highlights the greater inefficiency of the European markets.