Risk-axis identification of factorial repair effects

Determine whether the performance gains attributed to the conditional-mean, covariance, and scenario-channel repairs move portfolios away from the risk frontier rather than merely along it by recording and analyzing maximum drawdown and turnover for every factorial cell.

Background

The factorial analysis reports Sharpe differences only. Because a higher Sharpe ratio can result from accepting greater drawdown or turnover, the reported channel effects cannot establish genuine improvement in risk-adjusted decision quality without the corresponding risk-axis measurements. The paper explicitly states that the position of each factorial cell relative to the risk frontier is not identified.

References

Table~\ref{tab:factorial} reports $\Delta$Sharpe alone: the factorial lacks MaxDD and turnover per cell, so we cannot show that a channel repair moves a portfolio off the risk frontier rather than along it, the confound our own frontier test exists to expose and which we apply everywhere else.

— Verify Claims, Not Scores: Evidence-Based Verification of Modular Agents  (2610.01348 - Alzahrani, 1 Oct 2026) in Appendix, Section 11, 'Boundaries of established knowledge', paragraph 4