Please rotate your device to landscape mode to view the charts.

Informed Trading Intensity

Journal: Journal of Finance

Date: 20240401

Author: Bogousslavsky, Vincent; Fos, Vyacheslav; Muravyev, Dmitriy

Abstract:
We train a machine learning method on a class of informed trades to develop a new measure of informed trading, informed trading intensity (ITI). ITI increases before earnings, mergers and acquisitions, and news announcements, and has implications for return reversal and asset pricing. ITI is effective because it captures nonlinearities and interactions between informed trading, volume, and volatility. This data-driven approach can shed light on the economics of informed trading, including impatient informed trading, commonality in informed trading, and models of informed trading. Overall, learning from informed trading data can generate an effective informed trading measure.

Link: Google Scholar

Background and Context

Research Problem

Informed trading is difficult to identify empirically because investors' information sets are not directly observable and informed investors usually hide behind uninformed order flow.

Methodology

The authors develop a machine learning method trained on Schedule 13D trades to create a new measure called Informed Trading Intensity (ITI) that captures nonlinearities and interactions between informed trading, volume, and volatility.

Data Sources

The study analyzes 1,593 Schedule 13D filings between 1994-2018, opportunistic insider trades from 1993-2012, and short-selling data from 2006-2019.

ITI Performance in Detecting Schedule 13D Trading Days

  • ITI alone explains 9.86% of variation in Schedule 13D trading days
  • Common liquidity measures explain only 4.61% of variation
  • Combining ITI with common variables achieves 10.73% explanatory power

ITI Increases Before Major Information Events

  • ITI increases significantly 2 days before earnings announcements
  • Peak increase occurs on announcement day
  • ITI remains elevated for several days after announcements

Patient vs Impatient Informed Trading Detection

  • Impatient traders (last 20 days) trade 37% of daily volume on 49% of days
  • Patient traders (first 40 days) trade 20% of daily volume on 30% of days
  • Shows clear distinction between patient and impatient trading patterns

Portfolio Returns Based on ITI Sorting

  • ITI(13D) generates highest monthly alpha of 0.52%
  • ITI(impatient) produces strong returns of 0.54% monthly
  • ITI(short) shows minimal return predictability

Contribution and Implications

  • Provides a new data-driven approach to measure informed trading that outperforms traditional methods
  • Demonstrates important distinction between patient and impatient informed trading
  • Shows that informed trading can be detected across different types of traders (activists, insiders, short sellers)
  • Offers practical applications for predicting stock returns and monitoring trading activity

Data Sources

  • Detection Performance Chart: Based on Table II regression results
  • Information Events Chart: Based on Figure 4 Panel A
  • Patient vs Impatient Trading Chart: Based on descriptive statistics in Section III.A
  • Portfolio Returns Chart: Based on Table XI Panel B
  • Variable Importance Chart: Based on Table III and Figure IA.3