In the last decades, the mathematical modelling of systems composed of a large number of interacting agents has emerged as a promising research field. A significant part of the research activity applies methodologies from statistical mechanics to construct Boltzmann-type equations. Of particular interest are the attempts to model a market economy and the related wealth distribution. This work aims to contribute to the development of a complete and satisfactory kinetic description of a market economy. For the first time in the literature, the theoretical results obtained using the Cordier-Pareschi-Toscani model are compared with real-world data. The analysis is carried out for the Boltzmann model with linear interactions and the data are taken from the Household Finance and Consumption Survey (HFCS) held in 2012 in the Euro area. Remarkably, the results show that the kinetic approach provides a valid but still rough description of the wealth distribution dynamics. Furthermore, based on the analogy between the Italian income distribution and the stationary solutions of the Boltzmann model with linear interaction, we introduce a kinetic model for the income distribution. New non-Maxwellian kernels are proposed and studied from a numerical point of view. It is observed that different choices of the scattering kernel lead to similar stationary solutions with power-law decay of the right tail. Of particular interest are the kernels that include the exponential function with a negative exponent, which may allow for exponential convergence to the stationary solution. Finally, a new kernel is introduced based on economic considerations. The choice is motivated by the observation that kinetic models are based on instantaneous interactions between agents while, in reality, wealth is not liquid and can not always be exchanged right away, in particular when a large amount of wealth is involved in a trade.
La modellistica matematica dei sistemi composti da un gran numero di agenti interagenti negli ultimi anni si è messa in risalto come un campo di ricerca molto promettente. Una grossa parte dell'attività di ricerca applica metodi della meccanica statistica per sviluppare modelli basati sull'equazione di Boltzmann. Di particolare interesse sono i tentativi di modellare l’economia e la distribuzione della ricchezza. Lo scopo di questo lavoro è quello di contribuire allo sviluppo di una descrizione cinetica completa e soddisfacente di un'economia. Per la prima volta, si confrontano i risultati teorici ottenuti attraverso il modello Cordier-Pareschi-Toscani con dati reali. L'analisi si sviluppa sul modello di Boltzmann con interazioni lineari a partire dai dati del "Household Finance and Consumption Survey" (HFCS) tenutosi nel 2012 nei paesi dell'Unione Europea. I risultati mostrano che i modelli cinetici descrivono in modo corretto ma ancora approssimativo la dinamica della distribuzione della ricchezza. Osservando la somiglianza della distribuzione del reddito in Italia e la forma della soluzione stazionaria nel modello di Boltzmann con interazioni lineari, si introduce un modello cinetico per la distribuzione del reddito. Si introducono kernel non-Maxwelliani. Attraverso metodi numerici, si osserva che diverse scelte del kernel di interazione producono soluzioni stazionarie similari: in particolare tutte hanno la coda destra che decade in forma di potenza. Di notevole interesse sono i kernel che presentano al loro interno la funzione esponenziale con un esponente negativo: infatti, questo tipo di kernel potrebbe assicurare la convergenza esponenziale all'equilibrio. Infine, basandosi sul fatto che nei modelli cinetici le interazioni avvengono istantaneamente mentre nella realtà la ricchezza non è liquida, si introduce un nuovo kernel. Il suo scopo è quello di penalizzare le transazioni che includono una quantità di ricchezza eccessiva.
Kinetic description of a market economy
Del Mul, Edoardo Maria
2020/2021
Abstract
In the last decades, the mathematical modelling of systems composed of a large number of interacting agents has emerged as a promising research field. A significant part of the research activity applies methodologies from statistical mechanics to construct Boltzmann-type equations. Of particular interest are the attempts to model a market economy and the related wealth distribution. This work aims to contribute to the development of a complete and satisfactory kinetic description of a market economy. For the first time in the literature, the theoretical results obtained using the Cordier-Pareschi-Toscani model are compared with real-world data. The analysis is carried out for the Boltzmann model with linear interactions and the data are taken from the Household Finance and Consumption Survey (HFCS) held in 2012 in the Euro area. Remarkably, the results show that the kinetic approach provides a valid but still rough description of the wealth distribution dynamics. Furthermore, based on the analogy between the Italian income distribution and the stationary solutions of the Boltzmann model with linear interaction, we introduce a kinetic model for the income distribution. New non-Maxwellian kernels are proposed and studied from a numerical point of view. It is observed that different choices of the scattering kernel lead to similar stationary solutions with power-law decay of the right tail. Of particular interest are the kernels that include the exponential function with a negative exponent, which may allow for exponential convergence to the stationary solution. Finally, a new kernel is introduced based on economic considerations. The choice is motivated by the observation that kinetic models are based on instantaneous interactions between agents while, in reality, wealth is not liquid and can not always be exchanged right away, in particular when a large amount of wealth is involved in a trade.File | Dimensione | Formato | |
---|---|---|---|
Tesi_DelMul_Edoardo.pdf
accessibile in internet per tutti
Descrizione: Tesi
Dimensione
1.59 MB
Formato
Adobe PDF
|
1.59 MB | Adobe PDF | Visualizza/Apri |
Executive_Summary_DelMul.pdf
accessibile in internet per tutti
Descrizione: Executive Summary
Dimensione
706.66 kB
Formato
Adobe PDF
|
706.66 kB | Adobe PDF | Visualizza/Apri |
I documenti in POLITesi sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/10589/181608