Literature by the same author
plus at Google Scholar

Bibliografische Daten exportieren
 

On Modeling Heterogeneous Agents Avoiding Stiff Equations

Title data

Baumann, Michael Heinrich ; Baumann, Michaela ; Grüne, Lars ; Herz, Bernhard:
On Modeling Heterogeneous Agents Avoiding Stiff Equations.
2019
Event: 43rd Annual Meeting of the AMASES Association for Mathematics Applied to Social and Economic Sciences , September 9-11, 2019 , Perugia, Italy.
(Conference item: Conference , Speech )

Related URLs

Abstract in another language

One method to explain several stylized facts in financial markets is to model prices via market maker models with heterogeneous agents, usually fundamentalists and chartists. A basic approach going back to Day and Huang (1990) and Huang and Day (1993) is to assume the log-prices to be linear in the demand of the agents (cf. Brock and Hommes (1997, 1998), Lux (1995)). The demand function of the fundamentalists is assumed to be linear in the deviation of the log-price from the log-fundamental and the demand function of the chartists is assumed to be linear in the trend of the log-price.

When assuming the log-fundamentals as well as the number of the traders to be constant, we get a linear recurrence relation for the log-prices, where the log-fundamental is the unique equilibrium. Such models are used to determine the instability threshold of the fundamentalists to chartists ratio. We show that this model is an explicit Euler discretization of a properly chosen differential equation.

It turns out that even when there are no chartists, the equilibrium is unstable if there are "too many" fundamentalists: The price jumps in each point in time over the fundamental and the distance are even increasing. It is imaginable that when there are too many fundamentalists, the price jumps beyond the fundamental, however, in a long run, the price should converge to the fundamental value in the absence of chartists. That means, when this market becomes unstable it is not clear if too many chartists cause the instability or if the instability is a numerical artifact. The diverging behavior caused solely by the fundamentalists cannot be observed on real markets. This leads us to the question how to model heterogenous agents without these effects, which are well known from technical problems and are called stiff equations. We use an implicit Euler discretization for the ordinary differential equation mentioned above and observe that the model is stable in absence of chartists for arbitrarily many fundamentalists.

Further data

Item Type: Conference item (Speech)
Refereed: Yes
Keywords: Heterogeneous Agents Model; Stiff Equation; Implicit Euler Scheme
Institutions of the University: Faculties > Faculty of Mathematics, Physics und Computer Science > Department of Mathematics > Chair Mathematics V (Applied Mathematics)
Faculties > Faculty of Mathematics, Physics und Computer Science > Department of Mathematics > Chair Mathematics V (Applied Mathematics) > Chair Mathematics V (Applied Mathematics) - Univ.-Prof. Dr. Lars Grüne
Faculties > Faculty of Law, Business and Economics > Department of Economics > Chair Economics I - International Economics and Finance > Chair Economics I - International Economics and Finance - Univ.-Prof. Dr. Bernhard Herz
Profile Fields > Advanced Fields > Nonlinear Dynamics
Research Institutions > Research Centres > Forschungszentrum für Modellbildung und Simulation (MODUS)
Faculties
Faculties > Faculty of Mathematics, Physics und Computer Science
Faculties > Faculty of Mathematics, Physics und Computer Science > Department of Mathematics
Faculties > Faculty of Law, Business and Economics
Faculties > Faculty of Law, Business and Economics > Department of Economics
Faculties > Faculty of Law, Business and Economics > Department of Economics > Chair Economics I - International Economics and Finance
Profile Fields
Profile Fields > Advanced Fields
Research Institutions
Research Institutions > Research Centres
Result of work at the UBT: Yes
DDC Subjects: 300 Social sciences > 330 Economics
500 Science > 510 Mathematics
Date Deposited: 21 Oct 2019 12:10
Last Modified: 21 Oct 2019 12:10
URI: https://eref.uni-bayreuth.de/id/eprint/52796