Using Intraday Information to Build Daily Predictive Factors
Summary
The post argues that factor research can use information from later minutes within the same trading day when constructing a daily-frequency factor, provided the factor does not use data from the following day. It proposes examining relationships between current-minute indicators and later-minute observations, then aggregating the resulting signal to daily frequency. The author links this idea to the correlation between a factor and subsequent returns used in information-coefficient evaluation, and says that several top-performing factors in their work used minute-level leads.
The key caveat is that whether a lead is legitimate depends on the intended observation and execution times. A signal formed using later minutes of a day is only available after those minutes have occurred; it cannot support a trade placed earlier that same day. Aggregating such information into a daily factor may be suitable for a later decision point, but a valid backtest must align signal availability, return measurement, and trade execution to prevent look-ahead bias. The post provides a research suggestion, not detailed validation or evidence of out-of-sample performance.
Key ideas
- The author proposes using within-day minute data to study relationships with later minutes.
- Signals based on later observations are only tradable after those observations become available.
- Daily aggregation does not by itself prevent look-ahead bias in a backtest.
- Factor evaluation should align information timing with the return horizon and execution time.
Tags
This summary was written by Stratmill's research agent from the original; it is not a copy of the source.