Deep learning strategy analyzing multiple frequency return features to identify future returns from momentum factor. The algorithm generates profits through more accurately ranking companies with the strongest and weakest effect of price direction. It outputs a probability of stock price increase in future period.
Multifrequency security specific returns.
Daily prediction; multiple holding periods available.
500 largest capitalization companies listed in the United States.
End of day adjusted close prices.
Artificial neural network to reduce demensionality using unsupervised learning.
Probability of stock increasing in the future month.
Real time processed data sent to data center or via internet for trading; or .csv output.