Using Macro Sensitivity and Economic Expectations in Stock Selection

Summary

This report summary explores whether macroeconomic indicators can help select individual Chinese stocks, rather than serving only for asset allocation or industry rotation. It proposes estimating each stock’s sensitivity, or MacroBeta, to economic indicators and discusses preprocessing challenges such as publication delays, irregular release frequency, missing observations, and nonstationary series. Suggested treatments include differencing, using surprises, and constructing simulated asset portfolios.

The reported tests find that MacroBeta sorts can appear to have monotonic returns before controlling for investment styles, but much of that pattern weakens after style controls. The summary attributes this partly to the relationship between positive macro sensitivity, large company size, and the returns of large-cap Chinese stocks during the sample period. Its recommended framework combines estimated sensitivity with a forecast of the relevant economic indicator: rank stocks by sensitivity multiplied by the expected direction of the indicator. The approach depends on accurate macro forecasts and stable stock sensitivities over time; the summary says sensitivities tied directly to business operations may be more stable. Evidence is described only in summary form, without detailed results or test specifications.

Key ideas

• MacroBeta is intended to measure the direction and magnitude of a stock’s relationship with an economic indicator.
• Macroeconomic data may require preprocessing because releases can be delayed, irregular, incomplete, or nonstationary.
• Apparent stock-selection effects from MacroBeta can weaken substantially after controlling for common investment styles.
• A proposed macro score multiplies a stock’s sensitivity by the expected direction of the related economic indicator.
• The approach depends on sound macro forecasts and sensitivities that remain stable through time.

Tags

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