Building Equity Factors from High-Frequency Market Data
Summary
This research surveys how several forms of Chinese Level 2 market data can support equity signals: minute bars, order-book snapshots and queues, and transaction-level records. It describes factors based on intraday return shape, downside variation, late-session trading, price-volume relationships, order-flow changes, large trades, and inferred informed selling. Combining resting orders with executed trades is presented as a way to capture both prospective and realized buying pressure.
The study reports positive cross-sectional ranking statistics for the proposed factors and tests their use in index-enhancement portfolios. It incorporates high-frequency signals either into stock-return forecasts or into exclusions from the short side, reporting higher annualized excess returns for portfolios linked to several Chinese equity indexes. The evidence is summarized rather than fully documented here; details about implementation, transaction costs, data access, and robustness are not supplied. Results from historical tests do not establish that the factors will retain predictive power or remain profitable after trading costs.
Key ideas
- Level 2 data includes minute bars, order-book information, order queues, and trade records.
- Intraday factors measure price behavior, trading activity, and investor order-flow signals.
- Order submissions and completed trades can be combined to estimate broader buying pressure.
- High-frequency factors can enter return forecasts or help screen stocks from the short side.
- Reported index-enhancement results are historical and do not establish live performance after costs.
Tags
This summary was written by Stratmill's research agent from the original; it is not a copy of the source.