Skip to content
All library documents

Factor Profiles and CSI 1000 Index-Enhancement Tracking

Article BigQuant

Summary

This weekly competition report profiles the ten highest-ranked ModelScore factors and tracks a CSI 1000 index-enhancement portfolio. It summarizes average style exposures from daily multivariate regressions and reports average within-industry Rank ICs, the cross-sectional Spearman relationship between factor values and next-period returns. Factor identities and construction methods are withheld, so the profiles describe observed exposures without revealing how the signals were built.

For the portfolio, the top 100 factors are combined using regression importance, and constituent weights are tilted toward higher scores relative to benchmark weights. The report gives cumulative and annualized excess return, information ratio, tracking error, excess drawdown, daily win rate, and the change in IR after a new factor round. It also says the displayed curves omit absolute portfolio and benchmark values to conceal the backtest period. These are reported backtest statistics, not evidence of live performance; the stated tracking-error constraint and the reported annualized tracking error also differ, and the document does not explain that discrepancy.

Key ideas

  • The report aggregates style exposures and industry-level Rank ICs for the ten highest-ranked factors.
  • Factor rankings are anonymized, and factor construction and ownership are not disclosed.
  • The enhancement signal combines the top 100 factors using regression importance.
  • Portfolio weights are tilted toward higher-scoring CSI 1000 constituents relative to benchmark weights.
  • The report provides several relative-performance statistics but withholds the backtest period and absolute net values.
  • The listed tracking-error constraint and reported annualized tracking error are inconsistent, with no explanation given.

Tags

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