ASMF seminar: Matthias Scherer (TUM)
Almost Stochastic Dominance, Expectiles, and Optimal Parametric Insurance Design
Event details from: ASMF seminar: Matthias Scherer (TUM)
- Date
- 27 February 2026
- Time
- 12:30
- Room
- 5.22
This topic is based on the following 2 papers:
1) “On Expectiles and Almost Stochastic Dominance”, C. De Vecchi & M. Scherer, Insurance: Mathematics and Economics
2) “Expectiles as Basis Risk-Optimal Payment Schemes in Parametric Insurance”, M. J. Maier & M. Scherer, arXiv:2505.02607
Abstract
Our talk provides a new perspective on expectiles, almost stochastic dominance, and optimal contract design in parametric insurance. We begin with the concept almost first-order stochastic dominance (AFSD) and its connection to expectiles. The first part recalls properties of expectiles (as statistical functionals and coherent monetary risk measures), and shows how they provide a practical and interpretable criterion for comparing deterministic and random payoffs under AFSD. Building on this, the second part demonstrates how the AFSD–expectile structure can be used for optimal design in parametric insurance, where payouts are index-linked and only indirectly related to policyholders' realized losses. While parametric contracts offer operational advantages (speed, transparency, and cost-efficiency), they are inevitably subject to basis risk. Using an asymmetrically weighted mean-square framework, we show that basis risk–minimizing payment schemes can be characterized as (conditional) expectiles of loss given the parametric trigger.
The coordinators of the ASMF seminar series are Servaas van Bilsen and Felix Liebrich.