Informed Trading Intensity
72 Pages Posted: 15 Jun 2021 Last revised: 6 Feb 2022
Date Written: January 28, 2022
Abstract
We train a machine learning method on three classes of informed trades---activist trades, insider trades, and short sales---to develop a new measure of informed trading, the Informed Trading Intensity ("ITI"). ITI trained on one class of informed trades detects other classes of informed trades, pointing to commonalities in how informed investors trade. ITI measures increase before earnings, M\&A, and news announcements, and have implications for return reversal and asset pricing. ITI is effective because it captures nonlinearities and interactions between informed trading, volume, and volatility. Overall, learning from informed trading data can generate an effective informed trading measure.
Keywords: Informed trading, machine learning, adverse selection, stock returns, intraday data
JEL Classification: G10, G12, G14
Suggested Citation: Suggested Citation