Financial crises: Uncovering self-organized patterns and predicting stock markets instability

Financial markets are complex systems where investors interact using competing strategies that generate behaviours in which herding and positive feedbacks may lead to endogenous instabilities. This paper develops a novel methodology to detect the emergence of such phases by quantifying the intensity of self-organizing processes arising from stock returns' co-movements and self-similarities. Our methodology identifies a group of stocks, the Leading Temporal Module, whose statistical properties reflect the transition of the market into a crisis state. We define a topological indicator of the emergence of market discontinuity based on the autocovariance of the stocks in the Leading Temporal Module and on the ratio between the stocks' correlations within this group and the correlations between these stocks and those outside the leading module. This indicator provides early-warning market signals useful for policy-makers and investors by mapping the evolution of the topological properties of the leading module in different points in time.

Automated summary

This paper introduces a novel methodology to detect endogenous instabilities in financial markets by analyzing co-movements and self-similarities of stock returns. The methodology identifies a group of stocks, the Leading Temporal Module, whose statistical properties reflect the transition of the market into a crisis state. It defines a topological indicator of market discontinuity based on autocovariance and correlation ratios within this group, providing early-warning market signals for policy-makers and investors.