Main result — layer structure of optimal indemnity
In equilibrium the marginal indemnity is 1 (full insurance) on layers where the policyholder's distortion exceeds the insurer's pricing distortion, 0 (no insurance) where it is below, and arbitrary (within feasibility) where equal, so the indemnity has a layer-type structure .
Practical implication: insurance transfers concentrate on loss layers where the insured overweights tail-probabilities relative to the insurer .
Insurer’s pricing distortion and equilibrium characterization
The insurer selects a pricing distortion g* that never exceeds the policyholder’s marginal willingness to insure tails; g* is characterized via the conjugate distortion ˜g* and satisfies ˜g*(t)=˜T(t) when ˜T(t)<t, and ˜g*(t)∈[...] otherwise; the insurer’s profit at equilibrium is V^In = ∫(F_X^{-1})'(t) (t-˜T(t)) 1_{˜T(t)<t} dt .
Consequence: when the policyholder is (weakly or strongly) risk averse (T≥id or concave), full insurance and g*=T arise as equilibrium .
Welfare and Pareto efficiency
Property
Result
Pareto optimality of SE contracts
Every Stackelberg equilibrium contract is Pareto optimal and leaves the policyholder indifferent (extracts all surplus) .
Reverse mapping
Any Pareto-optimal contract that leaves the policyholder indifferent can be implemented as a Stackelberg equilibrium by choosing g = T for that contract .
Comparative statics — effect of greater risk aversion
If T1(t) ≤ T2(t) for all t (T2 more weakly risk averse), then equilibrium indemnity and insurer profit are weakly larger under T2: I_1(x) ≤ I_2(x) and V^In(I_1,g_1) ≤ V^In(I_2,g_2) .
The same monotonicity holds for stronger risk aversion via concave transformations (Corollary 3.12) .
Examples and numerical illustrations (identifies contract forms)
Full insurance: when T(t)≥t (weakly risk averse) or T concave (strong risk aversion), equilibrium gives I(x)=x and g*=T .
Coverage-limit (VaR) contracts: if T is VaR at level α, equilibrium gives full cover up to the α-quantile and cap above it; insurer profit increases with α .
Deductible contracts (ISSD): for inverse-S-shaped T, equilibrium can be a deductible that fully covers large losses beyond a threshold determined by intersection t1 where T(t1)=t1; numerical examples show insurer profit non-monotone in t1 and dependent on loss distribution .
Practical takeaways for contract design and pricing
The leader (insurer) can extract all surplus by aligning pricing distortion to the boundary where insured's marginal willingness to pay switches, yielding Pareto-optimal but consumer-surplus-extracting outcomes under monopoly .
思维导图
深度阅读
元数据与概览(引言与报告概览)
报告标题与作者:论文题为 “STACKELBERG EQUILIBRIA IN MONOPOLY INSURANCE MARKETS WITH PROBABILITY WEIGHTING”,作者为 Maria Andraos, Mario Ghossoub, Bin Liu, Benxuan Shi(University of Waterloo),发布日期为 2026-02-19,.
結果對失真函數形狀高度敏感:均衡給付結構依賴 T 與 g 的逐點比較(T vs g 或其共軛),當 T 為凹或 T ≥ identity 時會出現 full insurance;若 T 為 VaR 類則出現 coverage limit;若為 inverse-S 則出現 deductible 結構——因此政策或實務上 T 的估計誤差會直接影響設計出的均衡契約與利潤.
非唯一性與可選區段:當 T 與 g 在某些 t 相等時邊際給付存在自由度(φ_g 可任取),這會導致多重均衡或多重給付形式,需要額外均衡選擇準則或次優原則來消除不確定性(Theorem 3.6, Lemma 3.5).