Choosing Intraday and Daily Aggregation for High-Frequency Factors
Summary
This document summarizes a securities research report on constructing factors from high-frequency data, with a focus on combining intraday and day-level information. It frames the choice of calculation method around whether price and volume signals retain the same meaning across trading intervals and whether market microstructure contributes information beyond coarser observations. The summary suggests that aggregating high-frequency observations can be a broadly applicable approach when microstructure effects carry incremental information, while extending intraday characteristics across a history of sessions offers another way to represent a stock’s intraday behavior.
It also reports that momentum responses vary with the sampling frequency and the future horizon being examined. These points motivate testing factor definitions and forecast windows at multiple frequencies rather than assuming one frequency transfers uniformly. The supplied text is only a brief synopsis and citation to a research report; it contains no factor formulas, sample description, statistical results, implementation details, or transaction-cost analysis. Its claims therefore cannot be independently assessed from this excerpt alone.
Key ideas
- High-frequency factor design depends on whether price and volume behavior is consistent across intervals.
- Market microstructure information can affect whether observations should be aggregated.
- Summarizing a stock’s intraday characteristics over past sessions is another factor construction approach.
- Momentum behavior may differ across sampling frequencies and future evaluation horizons.
Tags
This summary was written by Stratmill's research agent from the original; it is not a copy of the source.