Titolo
Estimating Micromotives from Macrobehavior
Periodico
Università degli studi di Torino. Dip. Di Economia e Statistica Cognetti de Martiis. Working paper series
Anno: 2011 - Volume: 11 - Fascicolo: 11 - Pagina iniziale: 1 - Pagina finale: 26

In large economic systems the observed data are usually macro-data; an agent-based model is a set of individual decision makers that in uence the macro behavior of the system both trough their direct actions and trough the in uence they have on the other decision makers. The complexity of the model impedes an explicit analytical relation between the micro-parameters and the macro-behavior, and in turn the use of "traditional" econometric tools. This paper develops an estimation procedure that allows the estimation of the agents' micromotives in an agent-based model using the emerging behavior of the system. Starting from a simple stock market model, in which agents learn by interacting, the properties of simulation-based estimations are investigated. As a rst step in the direction of extending and adapting the econometric literature to agent-based models, the paper is focused on the consistency and on the bias of the estimates.



Testo completo: http://www.est.unito.it/do/home.pl/Download?doc=/allegati/wp2011dip/11_wp_grazzini.pdf

Esportazione dati in Refworks (solo per utenti abilitati)