Framing, Overconfidence and Regret in Italian Mortgage Banking Litigations

This paper aims at analyzing banking litigations according to a behavioral perspective. The purpose is to uncover intensity of affection for distortions or cognitive biases suffered by cus-tomers and to make an impact assessment. In fact, perfect rationality during the decision-making process might not be applicable due to several interferences. We focused on mortgage litigations, which, according to a pre-sampling procedure, resulted to be one of the most prone categories of case law to show a behavioral distortion. We exploited the dataset managed by the Italian alternative dispute resolution mechanism, i.e. the Arbitro Bancario Finanziario (hereinafter, ABF), and we focused on three well-known cognitive biases in a financial envi-ronment: narrow framing, overconfidence and regret. In our empirical analysis, firstly, we built up the sample by randomly selecting 75 decisions from the ABF archive through an extraction algorithm. This procedure guarantees a homoge-neous distribution and subsequently allowed to apply a statistical hypothesis testing of propor-tions to the sample. Secondly, we studied each litigation singly, to state its degree of distortion. Guidelines used for this goal have been stated ex ante in the form of fundamental parameters for each kind of bias, to ensure an objective and impartial statement. Thirdly, we run a statisti-cal hypothesis testing, to infer the sample to the whole population. Finally, but not marginally, we run a multivariate analysis to assess the impact of the presence of such biases. Our results indicate that if a litigation arises with the presence of a cognitive bias, the probability that the appeal is successful is low, after controlling for time and for the variability of the judge com-mittee composition.