Factor leadership

Which factor explains the cross-section?

ProductionVerdict · WARNRecall · R²=0.77

Context · Factor decomposition · anchor metric: 7-factor attribution

What it measures

Regresses the cross-section of returns onto a 7-factor model (value, momentum, size, quality, low-vol, …) and reports which factor explains the most dispersion this quarter, plus the model's overall fit.

r_i = \alpha + \sum_k \beta_{ik} \cdot f_k + \varepsilon_i \quad\text{\# } k = 7 \text{ style factors}

R^2 = \text{explained} \,/\, \text{total dispersion}

\text{top\_factor} = \operatorname*{argmax}_k |\text{contribution}_k|

A reading of "momentum, R² 0.62" says momentum was the dominant axis of return dispersion this quarter and the factor model captured 62% of it. High R² = a factor-driven tape; low R² = idiosyncratic.

Limitations

WARN (R²=0.77 EFA OOS) — fit varies by universe and window.
Per-factor breakdown is one number — the engine emits top-factor + R², not a full factor table to the panel.

In the product

The side split-bar in the Returns chapter ("Which factor explains the cross-section?"); CONTEXT factor-decomposition voice.

Grading & stability

—

Gate rate
loading…

—

Stability / churn
not emitted

—

Mean confidence
loading…

fetching /api/panel/about/two_sigma/W/all …

Graded on the Tier-1 regime-test harness — curated macro events × 18-yr backfill × recall@K. Verdict WARN (R² 0.77 EFA OOS).

Charts it reads

Fama-French factors

fama_french—

Sector performance

sectors—

Factor leadership · The Lens methodology← all engines