Overview

Metrics for the Common × Continuous cell — one time-series factor broadcast across N assets (a market-wide signal: VIX, USD index, oil, sentiment). Aggregation is time-series first: per-asset ordinary least squares (OLS) \(\beta\) over all dates, then a cross-asset \(t\) on the mean of the per-asset betas.

Metric	Role	Page
Cross-asset \(t\) on per-asset \(\beta_i\) (BJS aggregation)	Primary	ts_beta
Spread between top- and bottom-bucket assets sorted by \(\beta_i\)	Profile	ts_quantile
Sign-asymmetric slopes (positive vs negative regimes)	Profile	ts_asymmetry

At N == 1 the cross-asset \(t\) degenerates; factrix auto-routes to a TIMESERIES single-series test (null: \(\beta = 0\), not \(\mathbb{E}[\beta] = 0\)). Same StatCode.MEAN identifier, different statistical meaning — see profile.mode for which path ran.

The Common × Sparse cell shares the time-series-first aggregation shape but with a {0, R} broadcast event dummy (R unrestricted; {0, 1} for a pure event flag is the simplest form) in place of the continuous regressor; metrics live under Individual sparse.