Abstract
PRML, a novel candlestick pattern recognition model using machine learning methods, is proposed to improve stock trading decisions. Four popular machine learning methods and 11 different features types are applied to all possible combinations of daily patterns to start the pattern recognition schedule. Different time windows from one to ten days are used to detect the prediction effect at different periods. An investment strategy is constructed according to the identified candlestick patterns and suitable time window. We deploy PRML for the forecast of all Chinese market stocks from Jan 1, 2000 until Oct 30, 2020. Among them, the data from Jan 1, 2000 to Dec 31, 2014 is used as the training data set, and the data set from Jan 1, 2015 to Oct 30, 2020 is used to verify the forecasting effect. Empirical results show that the two-day candlestick patterns after filtering have the best prediction effect when forecasting one day ahead; these patterns obtain an average annual return, an annual Sharpe ratio, and an information ratio as high as 36.73%, 0.81, and 2.37, respectively. After screening, three-day candlestick patterns also present a beneficial effect when forecasting one day ahead in that these patterns show stable characteristics. Two other popular machine learning methods, multilayer perceptron network and long short-term memory neural networks, are applied to the pattern recognition framework to evaluate the dependency of the prediction model. A transaction cost of 0.2% is considered on the two-day patterns predicting one day ahead, thus confirming the profitability. Empirical results show that applying different machine learning methods to two-day and three-day patterns for one-day-ahead forecasts can be profitable.
Highlights
Analyzing and forecasting the stock market is notoriously tricky due to the high degree of noise [1] and semi-strong form of market efficiency [2], which is generally accepted
Krauss et al (2017) implemented and analyzed the one-day effectiveness of deep neural networks (DNNs), gradient-boosted-trees (GBTs), and random forests (RFs) on all stocks of the S&P 500 from 1992 to 2015, and the trading signals were generated based on the forecast probability
Long short-term memory neural networks (LSTMs) are one of the most common forms of recurrent neural networks (RNNs), which are a type of deep neural network architecture
Summary
Analyzing and forecasting the stock market is notoriously tricky due to the high degree of noise [1] and semi-strong form of market efficiency [2], which is generally accepted. Duvingage et al tested the intraday predictive power of Japanese candlesticks at the 5-minute interval on the 30 constituents of the DJIA index and concluded that candlestick trading strategies do not improve investment performance [15] These conflicting conclusions about candlestick patterns prompt us to investigate further. Krauss et al (2017) implemented and analyzed the one-day effectiveness of deep neural networks (DNNs), gradient-boosted-trees (GBTs), and random forests (RFs) on all stocks of the S&P 500 from 1992 to 2015, and the trading signals were generated based on the forecast probability. These techniques sort all stocks over the cross-section k probability in descending order.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
