A Comparison of Economic Agent-Based Model Calibration Methods

Platt, Donovan

doi:10.2139/ssrn.3335397

Cited by 10 publications

(6 citation statements)

References 32 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…More elaborated and simulation-oriented, large-scale agent-based models, studying the interplay between many different and evolving speculator types, have been advanced, for instance, by Palmer et al (1994), Arthur et al (1997), LeBaron et al (1999), Chen and Yeh (2001) and Raberto et al (2001). While it is still important to better understand the forces that may create financial market havoc, current research increasingly addresses questions that revolve around input validation (Anufriev et al 2016, Fagiolo et al 2017, model estimation (Lamperti et al 2018, Platt 2020, Kukacka and Kristoufek 2020, policy applications (Stanek and Kukacka 2018, Diem et al 2020) and prediction (Demirer et al 2019, Zhang et al 2019, Westphal and Sornette 2020. See Delli Gatti et al 2018, Dieci and He (2018), Iori and Porter (2018) and Lux and Zwinkels (2018) for up-to-date surveys.…”

Section: Introductionmentioning

confidence: 99%

Heterogeneous speculators and stock market dynamics: a simple agent-based computational model

Schmitt

Schwartz

Westerhoff

2020

The European Journal of Finance

View full text Add to dashboard Cite

We propose a simple agent-based computational model in which speculators' trading behavior may cause bubbles and crashes, excess volatility, serially uncorrelated returns, fat-tailed return distributions and volatility clustering, thereby replicating five important stylized facts of stock markets. Since each speculator bets on his own (technical and fundamental) trading signals, stock prices are excessively volatile and oscillate erratically around their fundamental value. However, speculators' heterogeneity occasionally vanishes, e.g. due to panic-induced herding behavior, yielding extreme returns. Lasting regimes with high volatility originate from the fact that speculators extract stronger trading signals out of past stock price movements when stock prices fluctuate strongly. Simulations furthermore suggest that circuit breakers may be an effective tool to combat financial market turbulences.

show abstract

Section: Introductionmentioning

confidence: 99%

Heterogeneous speculators and stock market dynamics: a simple agent-based computational model

Schmitt

Schwartz

Westerhoff

2020

The European Journal of Finance

View full text Add to dashboard Cite

show abstract

“…This method sets up an objective function, measuring the distance between the simulated and observed values of a set of target variables. Then, an optimization process is run in order to find that parameter combination that minimizes the distance, so that the chosen set of (endogenous) model variables are most closely resembled by the simulations (Platt 2020). This process was started by constructing an initial position for agents: This was done by mapping those relationships that already existed in 2010 (before the survey was conducted), into the two-dimensional social space.…”

Section: Model Calibrationmentioning

confidence: 99%

Dynamics of collaboration among high-growth firms: results from an agent-based policy simulation

2022

View full text Add to dashboard Cite

This study aims to show how supporting new business formation could promote the emergence of new ties in the knowledge network. Stimulating start-ups is conducive to the innovation system and economic development, however, we show how it has additional beneficial effects by the increasing the density of the innovation network. An innovation network is a complex socioeconomic phenomenon, which emerges from the decisions of many heterogeneous agents, that justifies the use of an agent-based model (ABM). We introduce an application of an ABM that is appropriate for simulating network formation among organizations. Agents are placed in the two-dimensional abstract social space where they are moving toward each other to find cooperation partners, following the gravity principle. The attraction is determined by the mass and the distance of agents. These parameters are specified by regression analysis, the result of which shows that the geographical, social, and technological distance has a negative impact on innovation-related cooperation. For the empirical underpinning of the model, we used survey data on the Hungarian high-growth firms’ (known as gazelles) egocentric network that contains information about innovation-purpose cooperation.

show abstract

“…The advantage of the black-box function methods lies at hand: Since they do not require any knowledge about the model, the same methods can be applied to a variety of different models. A good comparison on gradient-free calibration of ABMs can be found in Platt (2020) and Carrella (2021). On the other hand, a gradient-based calibration is model-specific.…”

Section: 8mentioning

confidence: 99%

“…We also want to stress the drawback of the gradient-based calibration method: While other researchers worked on improving the calibration of ABMs on a very general level (Lamperti et al 2018;Platt 2020;Carrella 2021), our method is only applicable in the described ABM of Innovation Diffusion and can make no statement about other types of models. Additionally, to use gradients, the ABM has to be rather simple and no complex decision rules for the agents are possible.…”

Section: 4mentioning

confidence: 99%

Calibrating Agent-Based Models of Innovation Diffusion with Gradients

Kotthoff

Hamacher

2022

JASSS

View full text Add to dashboard Cite

Consumer behavior and the decision to adopt an innovation are governed by various motives, which models find difficult to represent. A promising way to introduce the required complexity into modeling approaches is to simulate all consumers individually within an agent-based model (ABM). However, ABMs are complex and introduce new challenges. Especially the calibration of empirical ABMs was identified as a key difficulty in many works. In this work, a general ABM for simulating the Diffusion of Innovations is described. The ABM is differentiable and can employ gradient-based calibration methods, enabling the simultaneous calibration of large numbers of free parameters in large-scale models. The ABM and calibration method are tested by fitting a simulation with 25 free parameters to the large data set of privately owned photovoltaic systems in Germany, where the model achieves a coefficient of determination of R 2 ≃ 0.7.

show abstract

A Comparison of Economic Agent-Based Model Calibration Methods

Cited by 10 publications

References 32 publications

Heterogeneous speculators and stock market dynamics: a simple agent-based computational model

Heterogeneous speculators and stock market dynamics: a simple agent-based computational model

Dynamics of collaboration among high-growth firms: results from an agent-based policy simulation

Calibrating Agent-Based Models of Innovation Diffusion with Gradients

Contact Info

Product

Resources

About