A Chinese Stock Factor Analysis Workflow with Group Returns and IC

Summary

This tutorial presents a workflow for evaluating a factor on Chinese equities, using the lower Bollinger Band as an example. It introduces factor analysis as a way to summarize variables and derive composite scores, then describes building daily factor data, joining trading amount and security metadata, and filtering out special-treatment stocks, recent listings, non-main-board listings, and inactive securities. Optional preprocessing clips extreme values, standardizes factors, and neutralizes industry and market-cap effects.

The analysis framework assigns stocks to factor-ranked groups, calculates forward returns, compares performance with a benchmark, and reports group returns alongside information coefficient (IC), cumulative IC, and information ratio (IR). It explains normal and rank IC, noting that rank correlation is often preferred when normality is doubtful, and lists other performance measures such as volatility, Sharpe ratio, drawdown, and hit rate. The material is an implementation tutorial, not evidence that the example factor predicts returns: it supplies no interpreted numerical results, and its conclusions depend on data handling, portfolio construction, and the forward-return assumptions used.

Key ideas

• The example constructs a daily lower Bollinger Band factor for Chinese stocks and joins it with trading and security data.
• Filtering and optional clipping, standardization, and industry and size neutralization shape the factor universe and values.
• Stocks are ranked into groups so their forward returns can be compared with each other and a benchmark.
• Rank IC uses rank correlation and is presented as a practical alternative when factor data may not be normally distributed.
• The tutorial defines multiple return and risk measures but does not report evidence that its example factor is effective.

Tags

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