Testing a Smart-Money Factor Across Size and Turnover Groups

Summary

This note describes a two-dimensional factor analysis of a “smart money” signal in A-shares, grouping stocks by market capitalization and turnover. The author reports that the factor appeared useful only among medium- and large-cap stocks with lower turnover. In the highest-turnover groups, reported Sharpe ratios were negative across combinations, while the smallest-cap group showed no effective signal. The note interprets this as a possible shift toward institutional activity in more liquid core assets, especially planned trading near the close.

The author presents these conclusions as tentative and does not provide detailed portfolio definitions, factor construction, sample controls, or full performance statistics in the text. The overall factor appeared nearly ineffective without segmentation, motivating the two-way analysis. The implementation process involved querying minute data one trading day at a time to manage memory, then calculating the factor and required test data. The findings are specific to the stated 2024–2026 test period and should not be treated as proof of a persistent effect.

Key ideas

• The note evaluates a smart-money factor by market capitalization and turnover groups.
• The reported signal was concentrated in medium- and large-cap stocks with lower turnover.
• The author reports negative Sharpe ratios in high-turnover groups and no effective signal among the smallest stocks.
• The overall factor appeared nearly ineffective before the stock universe was segmented.
• The conclusions are tentative and lack detailed methodology and robustness evidence in the text.

Tags

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