numeraire.baselines.mean_variance_weights#

numeraire.baselines.mean_variance_weights(mu: NDArray[float64], cov: NDArray[float64], *, normalization: Literal['budget', 'none'] = 'budget') → NDArray[float64][source]#

Plug-in mean-variance (tangency) weights ∝ S^-1 mu, normalization made explicit.

normalization:

  • "budget" (default): divide by 1' S^-1 mu so the weights sum to one — the DeMiguel-Garlappi-Uppal convention. The divisor passes through zero for a near-cash-neutral tangency portfolio, which is why sample mean-variance weights and turnover explode.

  • "none": the raw proportional direction S^-1 mu (no budget rescaling).