Programme: University of Hamburg, Graduate School, Faculty of Social and Economic Sciences, Doctoral Programme

Course: Agent‐Based Models with Heterogeneous Agents

Course No: n.a.

Credits: 2

Lecturer:
Prof. Dr. Domenico Delli Gatti
Istituto di Teoria Economia e Metodi Quantitativi
Università Cattolica del Sacro Cuore
Largo Gemelli 1, 20123 Milano, Italy
E‐Mail: Domenico.delligatti@unicatt.it
Web: http://docenti.unicatt.it/ita/domenico_delli_gatti/

Prof. Dr. Delli Gatti is Full Professor of Economics at the Catholic University of Milan in the Faculty of Economics. He received his doctorate in political economy in 1987 from the Joint Doctoral Program of the Catholic University, Bocconi University and State University. In addition to his position at the Catholic University of Milan, he has been visiting professor at Columbia University and the Santa Fe Institute and visiting scholar in various high‐ranking universities.

From 2008‐2011, he has been European coordinator of the 7th Framework Programme research project on “Monetary, Fiscal and Structural Policies with Heterogeneous Agents” (POLHIA, www.polhia.eu).

He is currently coordinator of a new 7th Framework Programme research project: “Complexity based Research Initiative Systemic InstabilitieS” (CRISIS). This project will start in November 2011 and last until 2014. His main research interests cover models of complex systems and agent‐based macroeconomic models, including models of networks in a macroeconomic context. These models are applied, inter alia, to business cycle fluctuations and firms’ size dynamics, bankruptcy avalanches, credit chains and financial fragility.

Full list of publications: http://docenti.unicatt.it/ita/domenico_delli_gatti/

Learning Objectives and Outcome:

Most models in modern macroeconomics assume a representative agent (RA), thus ignoring direct interaction among individual agents, as any interaction or coordination occurs only indirectly through prices. However, the price mechanism can only work if complete information is assumed. By contrast, imperfect information necessarily leads to a model with heterogeneous agents and incomplete markets.

As an alternative to RA models, agent‐based (AB) models with heterogeneous agents aim at deriving the structure of aggregate macro behavior from the interaction of agents at the micro level. The economy is then regarded as a complex system of many interacting agents, where aggregate regularities may emerge as the consequence of complex patterns of interacting individuals. Using methods for instance from statistical mechanics or physics, the dynamics and equilibrium of the model economy can be described.

Important subtypes of AB models are network models, which represent the structure of partnerships within the economy. Generally, AB models with heterogeneous agents can be applied to a wide variety of economic areas, such as business cycle dynamics, credit chains and bankruptcy avalanches, and models describing interlinkages in networks of banks, firms and governments.

The objectives of the course are first, to present, compare and contrast various macroeconomic AB models and second, to implement the dynamic system which emerges as the reduced form of the macroeconomic model into a software code. Finally, the models will be calibrated and validated in order to corroborate the model’s dynamics against the empirical evidence.

Content:

1. Modeling the macro economy in an agent‐based framework. The following macroeconomic ABMs will be presented: MBU (Macroeconomics from the Bottom Up), EURACE, KS (Keynes meeting Schumpeter)

2. Translating the dynamic system of the model into software code (MATLAB)

3. Calibration and validation: bringing the model to the data

4. Modelling credit networks. The following models of a credit network will be presented: Battiston et al. (2007,2010), Delli Gatti et al. (2010)

Detailed Schedule and Further Information