Affine Virtual Values and the Generalized Pareto distribution
用仿射虚拟价值刻画广义帕累托分布，证明尾部指数决定最优截断和比较静态，通过单一尾部参数连接垄断、机制设计和搜索问题。
• Characterizes Generalized Pareto distributions via affine Virtual Values. • Shows that tail index fully governs optimal cutoffs and comparative statics. • Links monopoly, mechanism design and search through a single tail parameter. • Uses Extreme Value Theory to show asymptotic linearity of Virtual Values.