Title: The use of prior information in very robust regression for the monitoring of EU sensitive products
Authors: Aldo Corbellini - Faculty of Economics - University of Parma (Italy) [presenting]
Andrea Cerasa - European Commission - Joint Research Centre (Italy)
Francesca Torti - European Commission - Joint Research Centre (Italy)
Abstract: It is well known in the literature that overlooking outliers might have a severe impact on the analysis, leading to biased and inconsistent estimates and misleading inference. We develop a method of detecting the patterns of outliers that indicate systematic deviations in pricing. Since the data only become available year by year, we develop a combination of very robust regression and the use of `cleaned' prior information from earlier years, which leads to sharp indication of market price anomalies. As a method of very robust regression, we use the Forward Search. A form of empirical Bayesian analysis is extended to incorporate different amounts of prior information about the parameters of the linear regression model and the error variance. As an example we use yearly imports and exports of goods traded by the European Union. We provide a solution to the resulting big data problem, which requires analysis with the minimum of human intervention.