[Paper Review] Scale-free avalanche dynamics in the stock market

M. Bartolozzi, Derek B. Leinweber+1 more|RePEc: Research Papers in Economics|Jan 22, 2006

Complex Systems and Time Series Analysis4 citations

TL;DR

This study investigates self-organized criticality (SOC) in financial markets using multi-scale wavelet filtering to isolate avalanche dynamics from stock index returns. It finds robust power-law distributions in avalanche size, duration, and laminar times—indicating scale-free behavior—suggesting such dynamics may be a stylized fact of market indices, despite temporal correlations challenging classical SOC models.

ABSTRACT

Self-organized criticality has been claimed to play an important role in many natural and social systems. In the present work we empirically investigate the relevance of this theory to stock-market dynamics. Avalanches in stock-market indices are identified using a multi-scale wavelet-filtering analysis designed to remove Gaussian noise from the index. Here new methods are developed to identify the optimal filtering parameters which maximize the noise removal. The filtered time series is reconstructed and compared with the original time series. A statistical analysis of both high-frequency Nasdaq E-mini Futures and daily Dow Jones data is performed. The results of this new analysis confirm earlier results revealing a robust power law behaviour in the probability distribution function of the sizes, duration and laminar times between avalanches. This power law behavior holds the potential to be established as a stylized fact of stock market indices in general. While the memory process, implied by the power law distribution of the laminar times, is not consistent with classical models for self-organized criticality, we note that a power-law distribution of the laminar times cannot be used to rule out self-organized critical behaviour.

Motivation & Objective

• To determine whether self-organized criticality (SOC) underlies stock market dynamics by analyzing avalanche-like fluctuations.
• To develop and optimize a multi-scale wavelet filtering method to separate Gaussian noise from coherent market avalanches.
• To test whether the statistical properties of avalanches—size, duration, and laminar times—follow power laws, a hallmark of SOC.
• To assess whether the observed power-law distribution of laminar times contradicts classical SOC models or supports a modified SOC framework.
• To establish whether scale-free avalanche dynamics could be considered a stylized fact of financial market indices.

Proposed method

• A multi-scale wavelet transform is applied to logarithmic returns of the Nasdaq E-mini Futures (NQ) and Dow Jones (DJ) indices to decompose market dynamics into scale-specific components.
• An adaptive filtering procedure is developed to optimize noise removal by minimizing residual non-Gaussian fluctuations while preserving coherent avalanche events.
• Avalanches are identified as coherent periods of high volatility in the filtered return series, defined by thresholds in the residual signal after wavelet denoising.
• Avalanche size (V) is computed as the integrated squared volatility over each coherent event; duration (D) is the time span of each event; laminar time (L) is the interval between consecutive events.
• Statistical analysis of the probability distribution functions (PDFs) of V, D, and L is performed to test for power-law scaling.
• The optimal filtering parameters are selected based on minimizing the kurtosis of the filtered return distribution, ensuring effective noise suppression while preserving non-Gaussian dynamics.

Experimental results

Research questions

• RQ1 Do stock market indices exhibit scale-free avalanche dynamics consistent with self-organized criticality?
• RQ2 Can wavelet-based filtering effectively isolate coherent market avalanches from Gaussian noise in high-frequency and daily data?
• RQ3 Do the statistical distributions of avalanche size, duration, and laminar times follow power laws across different market indices and time scales?
• RQ4 How does the power-law distribution of laminar times challenge or support classical SOC models, which predict exponential distributions?
• RQ5 Can the observed power-law behavior in laminar times be reconciled with SOC theory, especially in systems with memory or non-random drivers?

Key findings

• The filtered time series successfully removes Gaussian noise while preserving coherent avalanche dynamics, as confirmed by reduced kurtosis and improved statistical fit to power laws.
• The probability distribution functions (PDFs) of avalanche size (V) for both NQ and DJ indices follow power laws with exponents γ ≈ -2.4 and γ ≈ -1.9, respectively.
• The PDFs of avalanche duration (D) also exhibit power-law scaling, with exponents γ ≈ -4.2 (NQ) and γ ≈ -3.5 (DJ).
• The laminar times (L) between avalanches display a power-law distribution with exponents γ ≈ -2.1 (NQ) and γ ≈ -2.3 (DJ), indicating temporal correlations in the triggering process.
• The power-law distribution of laminar times contradicts classical SOC models (which predict exponential distributions), suggesting a memory process in the market driver.
• Despite this discrepancy, the persistence of power laws in size and duration supports the possibility that the stock market may be in a near-SOC state, potentially driven by chaotic or dissipative mechanisms.