Exact signed conditional-monotonicity classification for Clayton copulas #
The strict midpoint comparison with independence excludes the opposite conditional monotonicity for each nonzero parameter branch.
theorem
ProbabilityTheory.Copula.not_isCI_clayton_negative
(θ : ℝ)
(hθ : -1 ≤ θ)
(hn : θ < 0)
:
¬(claytonNegative θ hθ hn).IsCI
Negative Clayton parameters are never conditionally increasing.
theorem
ProbabilityTheory.Copula.not_isPQD_clayton_negative
(θ : ℝ)
(hθ : -1 ≤ θ)
(hn : θ < 0)
:
¬(claytonNegative θ hθ hn).IsPQD
Negative Clayton parameters are not positive-quadrant dependent.