Cartel tools

Our cartel experience covers: 


  • Overcharge estimation
    To answer the question if and to what extent cartel agreements have been successful in raising prices above their competitive level we have performed a large number of overcharge estimations using a wide range of widely accepted econometric techniques of varied sophistication, from relatively simple before-and-after or yardstick comparisons, through simulations based on theoretical industrial organization models, to estimations of sophisticated econometric models of demand and supply. 
  • Cartel duration
    In cases in which the length of a cartel duration is disputed, we use our understanding of how markets operate to assess the extent to which observed market outcomes are consistent with competitive explanations and use statistical techniques, such as structural break analysis, to provide empirical evidence of possible cartel duration periods.
  • Screening for collusion
    Various economic tests exist to screen for collusive conduct, reaching from simple structural indicators to complex statistical methods. We have developed and implemented such screens and successfully applied them to industry data thereby allowing clients to understand the risk of collusion in their industries. 
  • Testing for the implementation of an agreement
    Factual evidence on how and when a collusive agreement was set up often results in an empirically testable hypothesis. Matching detailed information with available data allows testing for the implementation of such an agreement. A sound understanding of the functioning of cartels and empirical methods allows us to deliver robust results even in complex settings. 
  • Building up large datasets
    Cartel cases often require a comprehensive data collection, cleaning and documentation process. We have deep knowledge in supporting clients in this process, helping our clients to extract data from a firm's internal databases or collecting transactional data ourselves from hardcopy invoices. Professional quality control and cross-validation methods secure robust results in complex, large dataset environments.