Test grouped calendars

Title : Oligopoly,Complementarities, and Transformed Potentials

Abstract : We adopt a potential games approach to study multiproduct-firm pricinggames. We introduce the new concept of transformed potential and characterizethe classes of demand systems that give rise to pricing games admitting such apotential. The resulting demand systems may contain nests (of closersubstitutes) or baskets (of products that are purchased jointly), orcombinations thereof. These demand systems allow for flexible substitution patternsand can feature product complementarities arising from joint purchases andsubstitution away from the outside option. Combining the potential gamesapproach with a competition-in-utility approach, we derive powerful results onexistence of pure-strategy Nash equilibria.

Title : How to Weight in Moment Matching: An ML Approach with Applications toEarnings Dynamics

Abstract : Following the seminal paper by Altonji and Segal (1996), empiricalstudies commonly adopt equal or diagonal weighting in minimum distanceestimation to mitigate finite-sample bias arising from sampling error in theweighting matrix. We propose a new weighting scheme that combines cross-fittingwith regularized estimation of the weighting matrix, in the spirit of de-biasedmachine learning. We also propose a new formula for cross-fitted standarderrors. We show that several canonical models in the earnings dynamicsliterature satisfy exact or approximate  sparsity conditions that can beexploited by graphical lasso estimation of the weighting matrix. Within amany-moment asymptotic framework, we characterize the asymptotic distributionof the structural parameters. Extensive simulation studies demonstrate that ourapproach outperforms commonly used alternative weighting schemes. Finally, anempirical application using data from the Panel Study of Income Dynamicsillustrates the practical gains of our method.