Regression & Econometric Modeling

Analysis Group has extensive experience in regression analysis, including the application of econometric modeling in both litigation and non-litigation contexts.

In litigation related to finance, for example, we have used financial asset pricing models to test for “abnormal” stock price behavior, controlling for changes in the industry and stock market. We have also estimated expected mortgage defaults based on benchmark loans, controlling for loan and borrower characteristics and macroeconomic changes. In our antitrust work, we have applied econometric analysis to evaluate the impact of alleged anticompetitive conduct and prospective mergers, and to assess other competition issues. In addition to supporting affirmative testimony, we frequently provide expert rebuttal of opposing experts’ damages models. 

Outside of litigation, we have helped clients quantify operational and financial risk by modeling complex business and market environments and have provided strategic support to clients in health care and other industries. We have also applied regression analysis to public policy questions in a variety of settings.

Our work has included:

  • Estimation of financial payments made due to alleged fraud
  • Estimation of default and prepayment models for mortgage loans underlying litigated RMBS securities -- including discrete time duration models such as those used by the credit rating agencies -- to determine the “but-for” performance of the mortgage loans had their performance been similar to that of benchmark loans, after controlling for differences in loan and borrower characteristics and changes in macroeconomic conditions
  • Assessments to isolate and measure the effects of one of many explanatory variables, such as in “event study” analyses
  • Estimation of overcharges associated with price-fixing allegations 
  • Estimation of the pass through of costs in distribution chains
  • Estimation of demand models -- including the Berry-Levinsohn-Pakes (BLP) model, which was co-created by our affiliate, James Levinsohn -- to understand consumer behavior, to determine the “but-for” prices that would have prevailed in the absence of alleged anticompetitive conduct, and, in merger analysis, to understand the extent to which products are substitutes for one another and consequently whether a merger is likely to raise prices for consumers
  • Econometric modeling of pharmaceutical sales to forecast market share and revenue
  • Epidemiologic modeling to assess the relationships between pharmaceuticals and adverse events
  • Conjoint analysis of survey data
  • Development and implementation of a comprehensive market research process involving market research suppliers, product planners, and management to model complex consumer decision making
  • Analysis of vehicle depreciation rates in auto defect class actions
  • Evaluation of claims of disparate impact in reductions-in-force and employee pay
Read More