Active-Buying Factors from Tick Data for Stock Selection
Summary
This research summary describes stock-selection factors derived from tick-level trades. Trades are aggregated to minute intervals to estimate active buy and active sell amounts, which are used to form active-buy share and active-buy intensity measures. It also discusses daily versions that can be calculated from daily active-buy amounts and ratios when intraday processing is impractical.
The summary reports that some factors retained monthly stock-selection power after orthogonalization against conventional low-frequency factors. It gives positive information-coefficient, information-ratio, win-rate, and long-short return statistics for selected intraday measures, and says selected daily variants were weaker. The factors reportedly performed better within the CSI 300 universe than across the full market. The summary also notes correlation with prior returns and, for some measures, size, turnover, and volatility, while correlation with profitability measures was lower. These are reported historical research results; the underlying paper is not included, and the summary warns of market, liquidity, and policy risks.
Key ideas
- Minute-level active buy and sell amounts from tick data underpin active-buy share and intensity factors.
- Some intraday factors reportedly retained monthly selection power after orthogonalization against conventional factors.
- Daily factor variants were presented as less demanding to calculate but weaker in the reported tests.
- The factors showed stronger reported discrimination within the CSI 300 universe than across the full market.
- Prior returns and some size, turnover, and volatility measures were correlated with the factors.
Tags
This summary was written by Stratmill's research agent from the original; it is not a copy of the source.