Theorem 1, p. 12 — if a portfolio dominates the benchmark by τ-AASSD under some discretization scheme, it dominates it by τ-ASSD
ProvedAlmostSSDPortfolio.AASSD.theorem_1Consider the finite-scenario portfolio model of §2 under its standing assumptions: returns in with , scenario probabilities forming probability vectors. Let be a portfolio with return , let be the benchmark, and let .
If there exists a discretization scheme (grid containing every benchmark return) under which dominates by -AASSD, that is
then dominates by -ASSD:
In short, .
AASSD needs only means, variances and the LPM at finitely many grid points, so it can be imposed as a tractable constraint in portfolio optimization; this theorem guarantees that every portfolio it admits is a genuine -ASSD improvement on the benchmark.
Formalization Note. Both dominance notions are written with their ratios cross-multiplied (Definition 3 in the paper's form (5d)); see the definition file for why this is the faithful reading when a denominator vanishes. is the complement of inside .
import Mathlib import Definitions.Def_AlmostSSDPortfolio_AASSD_Model open MeasureTheory Set
namespace AlmostSSDPortfolio.AASSD
theorem theorem_1 {n T Tb : ℕ} (S : Setting n T Tb) (hS : S.Standing)
(lam : Fin n → ℝ) (hlam : lam ∈ simplex n) (τ : ℝ) (hτ : 1 < τ)
(h : ∃ D : Discretization S, D.IsAASSD lam τ) : S.IsASSD lam τ := by sorry
end AlmostSSDPortfolio.AASSD
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.