Building a Buy-Sentiment Factor from Aggressive and Passive Orders

Summary

The document describes a buy-sentiment factor built by distinguishing aggressive purchases, which execute against sell orders already on the book, from passive purchases, which wait as limit orders for incoming sells. It argues that finer-grained market data can reveal price and volume variation and help identify trading behavior, then uses the relationship between these two types of buying to form a stock-selection signal.

The reported evidence is a mean RankIC of 0.0724 and annualized excess returns over the CSI 500 for three versions of the factor: the raw signal, one neutralized for reversal, and one neutralized for reversal and market capitalization. The backtest covers January 2010 through February 2017. These figures are reported without details on data construction, portfolio rules, transaction costs, or statistical significance, so they do not establish that the factor would perform similarly in other periods or after implementation costs.

Key ideas

• Aggressive buying executes against resting sell orders, while passive buying waits for sellers to trade against a limit order.
• The factor uses the relationship between aggressive and passive buying to measure buy sentiment.
• The document argues that finer-frequency data can improve both price-and-volume analysis and identification of trading behavior.
• It reports a mean RankIC and excess-return results for raw and neutralized factor variants over a historical backtest period.
• The reported results do not include implementation costs or enough methodology to assess their robustness.

Tags

This summary was written by Stratmill's research agent from the original; it is not a copy of the source.