Why Reproduced High-Frequency Return-Variance Factors May Differ

Summary

This forum post raises a replication problem involving a high-frequency stock return-variance factor described in a Haitong research report. The author says they implemented the report's first two calculation methods in SQL, but obtained results on a very different scale from the values in a research-factor table. They also report that an SQL built-in variance function does not reproduce the table values, and ask how the stored factor may have been calculated.

The post identifies a practical issue in quantitative research: factor definitions and implementation details can materially affect computed values. However, the text does not include the formulas, SQL queries, sample outputs, data preparation rules, or a response resolving the discrepancy. It therefore frames a reproducibility question rather than providing a method or confirmed explanation. Possible causes cannot be distinguished from the information shown, so readers would need the report's precise definitions and the platform's data and calculation conventions to investigate further.

Key ideas

• The author compares two SQL implementations of return variance with values in a research-factor table.
• The implemented values reportedly differ substantially in scale from the stored factor values.
• An SQL built-in variance calculation also fails to match the table, according to the post.
• The post does not include enough formulas, data details, or answers to determine the cause.

Tags

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