Measuring Benchmark Damages in Antitrust Litigation: Extensions and Practical Implications

Abstract

Abstract This paper compares the “forecasting approach” and the “fully interacted approach” to estimation of cartel damages. We investigate the impact of relaxing the assumptions of exogeneity and stationarity, both theoretically and in Monte Carlo simulations. The results suggest that the advantages of the fully interacted approach are less clear and that the forecasting approach may be more robust to the relaxations of some of these maintained assumptions.