[Paper Review] Hierarchical causality in financial economics

Diane Wilcox, Tim Gebbie|RePEc: Research Papers in Economics|Aug 24, 2014

Complex Systems and Time Series Analysis4 citations

TL;DR

This paper proposes a hierarchical causality model in financial economics that integrates top-down and bottom-up causation across five distinct causation classes, using a multilevel system of coupled models to better capture market complexity, emergence, and feedback mechanisms while remaining consistent with no-arbitrage pricing. The model enables more realistic, causally complete simulations of financial systems by embedding agent behavior, market structure, and macroeconomic influences in a scalable, multi-scale framework.

ABSTRACT

Hierarchical analysis is considered and a multilevel model is presented in order to explore causality, chance and complexity in financial economics. A coupled system of models is used to describe multilevel interactions, consistent with market data: the lowest level is occupied by agents generating the prices of individual traded assets; the next level entails aggregation of stocks into markets; the third level combines shared risk factors with information variables and bottom-up, agent-generated structure, consistent with conditions for no-arbitrage pricing theory; the fourth level describes market factors which originate in the greater economy and the highest levels are described by regulated market structure and the customs and ethics which define the nature of acceptable transactions. A mechanism for emergence or innovation is considered and causal sources are discussed in terms of five causation classes.

Motivation & Objective

• To address the limitations of orthodox economic models in explaining financial crises by developing a causally complete framework that accounts for multiple levels of causality.
• To integrate top-down and bottom-up causation mechanisms—especially feedback, adaptation, and emergence—into a coherent multilevel model of financial markets.
• To provide a structured, scalable modeling approach that remains consistent with no-arbitrage pricing theory while allowing for regulatory, behavioral, and systemic influences.
• To explore how hierarchical modeling can improve forecasting and policy design by capturing interdependencies across market participants, risk factors, and institutional structures.
• To demonstrate that causality in financial systems extends beyond traditional reductionist views, incorporating algorithmic, informational, and adaptive control mechanisms.

Proposed method

• Develops a five-level hierarchical model: (1) individual agent trading, (2) asset aggregation into markets, (3) integration of risk and information factors, (4) macroeconomic influences, and (5) institutional and regulatory structures.
• Applies a coarse-graining approach to represent system levels via averaged, scale-appropriate models, enabling autonomy and reduced information loss across levels.
• Incorporates five causation classes: TDC1 (algorithmic), TDC2 (non-adaptive control), TDC3 (adaptive selection), TDC4 (feedback control of goals), and TDC5 (adaptive goal selection), each instantiated with empirical or plausible market mechanisms.
• Uses coupled differential equations and multi-scale modeling to represent interactions across levels, ensuring consistency with no-arbitrage conditions at relevant scales.
• Calibrates the model using historical data to enable forecasting and stress-testing of systemic behaviors.
• Employs graph-theoretic and network analysis techniques to identify hierarchical clusters and information flow patterns in financial systems.

Experimental results

Research questions

• RQ1 How can a hierarchical causality model integrate top-down and bottom-up causation in financial markets while remaining consistent with no-arbitrage pricing theory?
• RQ2 What are the distinct classes of top-down causation in financial systems, and how can they be empirically instantiated in a multilevel modeling framework?
• RQ3 In what ways do regulatory, institutional, and behavioral actors influence market dynamics through feedback and adaptive goal-setting mechanisms?
• RQ4 How does the emergence of systemic risk or innovation arise from interactions across multiple levels of market structure and causality?
• RQ5 Can a causally complete, multi-scale model improve the predictability and robustness of financial system simulations compared to traditional models?

Key findings

• The hierarchical model successfully integrates five distinct causation classes—TDC1 through TDC5—each representing different mechanisms of top-down influence, such as algorithmic control, adaptive selection, and feedback-driven goal adjustment.
• The model maintains consistency with no-arbitrage pricing at relevant scales, ensuring theoretical grounding despite the inclusion of complex, adaptive, and emergent behaviors.
• Regulatory and institutional actors (e.g., central banks, market rules) are shown to exert significant top-down influence through both direct enforcement and indirect effects on market participant behavior and risk-taking.
• Bottom-up causality from agent-level trading and information flows can lead to emergent market phenomena such as liquidity shifts, volatility clustering, and arbitrage opportunities, even under regulated conditions.
• The model demonstrates that feedback loops and adaptive goal-setting among traders and profiteers can lead to systemic outcomes that deviate from long-term market stability, even when rules are technically followed.
• Stress-testing reveals that the model can simulate cascading failures and market instabilities by tracing causal chains across levels, particularly when regulatory or information controls break down.