Building Intraday High-Frequency Factors for Stock Selection

Summary

This Chinese-language article describes a workflow for turning intraday trading data into daily stock-selection factors and using those factors in analysis and strategy development. Its example is a large-order-driven price-rise factor, alongside other candidate measures such as average trade value, the share of outflow value, and net inflow from large orders. It proposes aggregating the example factor over the prior twenty trading days to create a stock-level daily feature.

The workflow relies on BigQuant’s module for extracting high-frequency features from minute data and converting them to daily frequency, followed by derived-feature processing and factor analysis or strategy backtesting. The article frames tick and transaction-level data as a way to extend research beyond increasingly competitive daily factors. However, the supplied text stops before showing the factor expression, calculation details, analysis results, or backtest findings. It cites an earlier report as the source of the factor definition, so the example cannot be independently reconstructed or evaluated from this excerpt alone.

Key ideas

• The article presents a workflow for converting intraday high-frequency data into daily stock-selection factors.
• Its example factor measures price increases driven by large orders.
• A past twenty-day average is proposed as the daily factor value.
• The workflow uses minute-to-daily feature extraction before factor analysis and strategy backtesting.
• The provided excerpt omits the expression, detailed calculations, and empirical results.

Tags

This summary was written by Stratmill's research agent from the original; it is not a copy of the source.