Unraveling Financial Fragility of Global Markets Using Machine Learning
This study examines systemic financial risk forecasting in global markets using machine learning-based heterogeneous panel regression and multiple risk indices.
What it examines
This paper uses advanced machine learning panel regressions to examine systemic financial risk in global markets. It analyzes how geopolitical, climate, and economic uncertainties drive market stress after COVID-19 and employs various forecasting methods to improve prediction accuracy.
What it concludes
Results show that economic and financial factors best forecast long-term systemic risk, while geopolitical risks aid short-term predictions. The study suggests updating risk frameworks and exploring emerging economies to better prepare for financial stress and avoid prolonged recessions.
Evidence objects
Research reveals systemic financial risk impacted by instability, climate hazards, and economic uncertainties, with surprising findings: short-term forecasts improve with geopolitical risk, while long-term predictions depend on economic data significantly.
key_findings bullet 1 · key_findings · validation V0
The study introduces an advanced machine learning heterogeneous panel regression capturing crosssectional dependencies and nonlinear patterns, while defining indices for climate risks and geopolitical components to describe risk spillovers effectively.
key_findings bullet 2 · key_findings · validation V0
Extensive out-of-sample rolling window analyses from 2006 to 2022 compared econometric and machine learning methods including TWFE, MG, SVR, and Panel SVR with kernels, exposing overfitting and instability during periods.
key_findings bullet 3 · key_findings · validation V0
The paper uses state-of-the-art ML heterogeneous panel regression addressing nonlinear financial stress drivers from geopolitical, climate, and economic sources. It overcomes conventional model limitations through innovative integration; its originality, novelty, and potential global risk management and policy impact render the work highly engaging and influential for Quantitative Risk Management specialists.
key_findings bullet 4 · key_findings · validation V0
Raw abstract and provenance
- … the revival of a rise of a widespread (systemic) financial risk in financial markets, in this paper we evaluate the potential sources of financial systemic risk, utilizing a state-of-…
Source row: 2102 · abstract type: snippet