Subtopic Deep Dive

← Complex Systems and Time Series Analysis

Agent-Based Models of Market Crashes
Research Guide

What is Agent-Based Models of Market Crashes?

Agent-based models of market crashes simulate interactions among heterogeneous agents exhibiting herding, leverage cycles, and contagion to replicate empirical crash dynamics in financial markets.

These models depart from equilibrium assumptions by incorporating behavioral rules and network structures. Calibration to historical data like the 1987 crash or 2008 crisis tests policy interventions. Over 300 papers explore these dynamics since foundational works by Shiller (1984, 787 citations) and Rosser (1999, 375 citations).

Curated Papers

Key Challenges

Why It Matters

Agent-based models reveal systemic risk propagation through networks, as in Markose et al. (2012) analysis of CDS market fragility (313 citations), enabling regulators to design circuit breakers. Shiller (2003, 1689 citations) shows social dynamics drive excess volatility, informing behavioral policies. Gabaix (2016, 408 citations) power laws quantify crash tail risks for better capital buffers.

Key Research Challenges

Agent Heterogeneity Calibration

Matching ABM outputs to stylized facts like fat tails requires calibrating diverse agent rules to high-frequency data. Rosser (1999) notes nonlinear interactions amplify mismatches. Shiller (1984) highlights social dynamics complicating parameter estimation.

Systemic Network Emergence

Endogenous network formation leads to 'too interconnected to fail' structures, per Markose et al. (2012). Capturing contagion across asset classes demands scalable simulations. Carvalho (2014, 374 citations) stresses micro-macro linkages in production networks.

Out-of-Sample Crash Prediction

Models overfit historical crashes but fail on unseen events due to behavioral shifts. Barberis et al. (2018, 423 citations) link extrapolation biases to bubbles. Arthur (2021, 307 citations) calls for complexity economics to address non-stationarity.

Essential Papers

From Efficient Markets Theory to Behavioral Finance

Robert J. Shiller · 2003 · The Journal of Economic Perspectives · 1.7K citations

The efficient markets theory reached the height of its dominance in academic circles around the 1970s. Faith in this theory was eroded by a succession of discoveries of anomalies, many in the 1980s...

Stock Prices and Social Dynamics

Robert J. Shiller · 1984 · RePEc: Research Papers in Economics · 787 citations

macroeconomics, stock prices, assets, social movements, investments

Using wavelets to decompose the time–frequency effects of monetary policy

Luís Aguiar‐Conraria, Nuno Azevedo, Maria Joana Soares · 2008 · Physica A Statistical Mechanics and its Applications · 553 citations

Extrapolation and bubbles

Nicholas Barberis, Robin Greenwood, Lawrence J. Jin et al. · 2018 · Journal of Financial Economics · 423 citations

Power Laws in Economics: An Introduction

Xavier Gabaix · 2016 · The Journal of Economic Perspectives · 408 citations

Many of the insights of economics seem to be qualitative, with many fewer reliable quantitative laws. However a series of power laws in economics do count as true and nontrivial quantitative laws—a...

On the Complexities of Complex Economic Dynamics

J. Barkley Rosser · 1999 · The Journal of Economic Perspectives · 375 citations

Complex economic nonlinear dynamics endogenously do not converge to a point, a limit cycle, or an explosion. Their study developed out of earlier studies of cybernetic, catastrophic, and chaotic sy...

From Micro to Macro via Production Networks

Vasco M. Carvalho · 2014 · The Journal of Economic Perspectives · 374 citations

A modern economy is an intricately linked web of specialized production units, each relying on the flow of inputs from their suppliers to produce their own output which, in turn, is routed towards ...

Reading Guide

Foundational Papers

Start with Shiller (1984, 787 citations) for social dynamics in prices, then Shiller (2003, 1689 citations) for behavioral critiques of EMH, and Rosser (1999, 375 citations) for complexity foundations—these establish micro motivations for ABM crashes.

Recent Advances

Study Barberis et al. (2018, 423 citations) on extrapolation bubbles, Gabaix (2016, 408 citations) power laws, and Arthur (2021, 307 citations) complexity economics for modern ABM advancements.

Core Methods

Core techniques: agent rules for herding/leverage, network topologies (Markose et al. 2012), wavelet time-frequency analysis (Aguiar-Conraria et al. 2008), and Monte Carlo calibration to fat-tailed data.

How PapersFlow Helps You Research Agent-Based Models of Market Crashes

Discover & Search

Research Agent uses citationGraph on Shiller (2003) to map behavioral finance clusters, then findSimilarPapers uncovers ABM extensions like Markose et al. (2012) on CDS networks. exaSearch queries 'agent-based market crash herding calibration' for 50+ targeted papers beyond OpenAlex indexes.

Analyze & Verify

Analysis Agent runs runPythonAnalysis on crash time series from Aguiar-Conraria et al. (2008) wavelet decompositions, verifying power law fits via Gabaix (2016) stats. verifyResponse (CoVe) with GRADE grading flags ABM claim contradictions; statistical tests confirm fat-tail replication.

Synthesize & Write

Synthesis Agent detects gaps in leverage cycle modeling post-Barberis et al. (2018), flagging underexplored contagion paths. Writing Agent uses latexEditText for ABM pseudocode, latexSyncCitations for Rosser (1999), and latexCompile for policy report; exportMermaid diagrams herding cascades.

Use Cases

"Replicate 2008 crash leverage cycles in ABM with real data calibration"

Research Agent → searchPapers 'ABM leverage cycles 2008' → Analysis Agent → runPythonAnalysis (pandas simulation of agent debt dynamics) → matplotlib plot of crash tails vs. empirical data.

"Write LaTeX appendix on Shiller social dynamics in crash ABMs"

Synthesis Agent → gap detection in Shiller (1984, 2003) → Writing Agent → latexEditText (insert herding equations) → latexSyncCitations → latexCompile → PDF with citation-matched bibliography.

"Find GitHub repos implementing power law crash ABMs"

Research Agent → searchPapers 'Gabaix power laws ABM code' → Code Discovery → paperExtractUrls → paperFindGithubRepo → githubRepoInspect → executable NumPy scripts for tail risk simulation.

Automated Workflows

Deep Research workflow scans 50+ papers from Shiller (2003) citationGraph, producing structured report on ABM crash mechanisms with GRADE-verified claims. DeepScan's 7-step chain analyzes Markose et al. (2012) network data via runPythonAnalysis for fragility metrics. Theorizer generates hypotheses on post-2021 contagion from Arthur (2021) complexity foundations.

Try Doxa for Agent-Based Models of Market Crashes Research

Frequently Asked Questions

What defines agent-based models of market crashes?

Simulations of interacting agents with herding and leverage rules replicate crash fat tails and volatility clustering, calibrated to data like 1987 or 2008 events.

What methods dominate these models?

NetLogo or Python-based multi-agent systems incorporate behavioral heuristics (Shiller 1984), network contagion (Markose et al. 2012), and power law calibrations (Gabaix 2016).

What are key papers?

Shiller (2003, 1689 citations) on behavioral anomalies; Rosser (1999, 375 citations) on complex dynamics; Markose et al. (2012, 313 citations) on financial networks.

What open problems remain?

Out-of-sample prediction amid regime shifts (Barberis et al. 2018); scalable calibration for real-time policy (Arthur 2021); integrating ML for agent learning (Kou et al. 2019).