Testing Analyst Consensus Factors for Stock Selection in China
Summary
This study evaluates 19 factors derived from sell-side analyst consensus data, including forecasts of financial measures, analyst ratings, and attention. It tests the factors in several Chinese equity universes and across industries. For financial forecast factors, it compares raw values with two transformations: within-industry ranks and quarter-over-quarter changes. The reported long-term results suggest that some consensus measures and derivatives have stock-selection value, with performance varying by universe, factor, and sector.
Industry ranks generally show higher information ratios and more stable portfolio groupings than raw measures, although their average rank information coefficients are often lower. Quarter-over-quarter changes often have stronger information ratios and long-short Sharpe ratios, but also higher turnover; the reported gains frequently come from short positions. Financial factors perform best overall in the CSI 300 universe, while analyst-attention measures have limited coverage and modest results. Banking stands out in the sector tests. These are historical single-factor tests, so results may not persist if market relationships change or wider adoption erodes the effect.
Key ideas
- The study tests analyst consensus forecasts and attention measures as stock-selection factors in Chinese equities.
- Within-industry ranks often improve information ratios and stability compared with raw forecast values.
- Quarter-over-quarter changes can improve long-short metrics while increasing turnover, with results often driven by short positions.
- Factor performance varies across equity universes and industries, with stronger results reported for the CSI 300 and banking sector.
- Historical single-factor results may weaken as market conditions change or investors adopt the signals more widely.
Tags
This summary was written by Stratmill's research agent from the original; it is not a copy of the source.