Measuring Passive and Active Crowding in Equity Portfolios
Summary
The document summarizes two ways to classify crowded equity holdings. For passive crowding, it first divides stocks into size groups using market-capitalization thresholds, then ranks them within each group by the rolling change in passive-fund ownership as a share of outstanding shares. The top and bottom deciles are labeled crowded and uncrowded, with the middle group neutral. For active crowding, it combines four dimensions: active-manager holdings, analyst buy ratings, long-term momentum, and persistence of institutional trading. Stocks receive within-universe decile scores for available components; the average score determines the combined ranking, provided at least two components are available.
The cited summary reports that long uncrowded and short crowded portfolios earned positive returns across size groups, with positive results remaining after controls for market, size, value, and momentum factors. It also says crowded active portfolios experienced larger drawdowns, making crowding potentially useful for screening and exposure management. These are reported backtest findings, not independently verifiable here; the document flags model failure, systemic risk, and differences between overseas and domestic markets. The accompanying sample code and stated weekly moving-average condition do not fully match the described strategy, so implementation details need checking.
Key ideas
- Passive crowding is ranked within market-cap groups using changes in passive-fund ownership.
- Active crowding combines institutional ownership, analyst ratings, long-term momentum, and sustained institutional buying.
- The summary reports positive long-short results after controlling for common market, size, value, and momentum factors.
- It attributes much of the passive-crowding signal to weak returns among crowded stocks.
- Crowded active holdings are described as more vulnerable to sharp drawdowns, but the reported findings require independent validation.
Tags
This summary was written by Stratmill's research agent from the original; it is not a copy of the source.