Machine Learning and Trade Direction Classification : Insights Fromthe Corporate Bond Market

Leveraging the availability of a large panel of signed trade data in the corporate bond market,we explore how machine learning methods can be used to improve upon standard tradedirection classification methods in markets in general, both with and without pre-trade transparency.Using the signed data set allows us to show how both the trading and informationenvironment at the time of the trade critically affect the accuracy of existing trade classificationrules in general and also illustrate the importance of optimizing the feature set incorrectly classifying trade direction. These insights extend to the equity market