Using Analyst Earnings Views to Select Chinese Growth Stocks

Summary

This Chinese equity research note argues that quarterly profit growth alone can misidentify growth stocks. Small comparison-period earnings, denominator effects, and nonrecurring gains can create misleading growth rates. It proposes using analyst research titles and forecast revisions after earnings announcements as additional evidence: classify analysts’ views as positive, negative, or neutral, then combine counts of positive versus negative views with the balance of upward versus downward profit estimate revisions.

The note tests this approach among high-growth Chinese stocks and describes a periodically rebalanced, equal-weight portfolio that filters for earnings growth and analyst confirmation. It also builds four analyst-approval factors, combines them, and evaluates stock selection and industry rotation using information coefficients, grouped returns, and benchmark comparisons. Reported historical results are strong, including a claimed annualized return of 33.41% for the selected portfolio and positive factor statistics across broad-market and index universes. These are the authors’ backtest findings, not independently verified outcomes. The method depends on a proprietary analyst-report database and title-pattern classification, and the excerpt does not fully establish robustness to publication timing, transaction costs, or changing market conditions.

Key ideas

• Year-over-year earnings growth can be distorted by small prior-period earnings and nonrecurring items.
• Analyst report language and earnings forecast revisions are used to assess whether high growth is broadly supported.
• The stock-selection process combines analyst approval counts and the balance of upward versus downward revisions.
• The study reports positive historical factor and portfolio results across several Chinese equity universes.
• The results rely on historical tests and a proprietary report database, so live performance is not established.

Tags

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