Tainá Fernandes De Brito
Doctorate – Attribute pricing using hierarchical linear models: an application to the popular motorcycles market in Brazil
Advisor: Prof. Dr. Nuno Manoel Martins Dias Fouto
Comission: Profs. Drs. Luiz Paulo Lopes Fávero, Claudio Felisoni de Angelo and André Taue Saito
Assuming that motorcycles are heterogeneous goods, which possesses multiple attributes that yield utility to the consumer (LANCASTER, 1966), this research sought to reveal the implicit prices of relevant attributes of popular motorcycles, by applying the 1st stage of the hedonic procedure proposed by Rosen (1974). Based on data from Tabela FIPE and information in the user’s manuals provided by the manufacturers, a database was created with 61 motorcycle versions (grouped into 23 motorcycle models). The estimation of linear hierarchical models with repeated measures identified individual trajectories of the retail prices for these motorcycles over 48 months (January/2015 to December/2018). The final model contains random-intercept effects and random-slope effects. It reveals the semi-elasticities of the relevant attributes: brand, category, start system and type of front brakes.
*Abstract provided by the author