theorem variance_targeting_improves {sigma_prev delta beta_n m_actual m_floor : ℝ}
(hs_prev : 0 < sigma_prev) (hd : delta ≠ 0) (hb : 0 < beta_n)
(hma : 0 < m_actual) (hmf : 0 < m_floor)
(hexcess : m_actual < m_floor) :
effectiveWelfareContribution sigma_prev delta beta_n m_floor <
effectiveWelfareContribution sigma_prev delta beta_n m_actual :=
eff_welfare_decreasing_margin hs_prev hd hb hma hmf hexcessVariance Targeting Escapes Damping Cancellation