A sophisticated quantitative finance project employing advanced econometric techniques to analyze the nonlinear relationship between geopolitical risk (GPR) indices and financial markets. This research goes beyond simple correlation to capture complex, non-linear dependencies.
This project utilizes cutting-edge econometric approaches:
- Analyzes long-range correlations and market memory effects
- Measures persistence in volatility patterns during risk events
- Identifies scaling properties in gold/SPY time series
- Quantifies information flow between GPR index and asset returns
- Captures non-linear, non-parametric causal relationships
- Reveals directional information transfer (GPR → Gold vs GPR → SPY)
- GPR Index: Academic geopolitical risk indices from existing literature
- Market Data: Gold prices and SPY ETF historical data
- Time Period: Multi-decade analysis for robustness
GPR Index → [Data Preprocessing] → [DFA Analysis] → [Transfer Entropy] → [Non-linear Modeling] → Trading Insights
- Non-stationary time series analysis
- Multi-fractal DFA for multi-scale analysis
- Conditional transfer entropy controlling for market factors
- Regime-switching models for different volatility environments
- Does GPR contain predictive information beyond what's captured by linear models?
- How does information flow differ between safe-haven (gold) and risk-on (SPY) assets?
- Are there critical thresholds where GPR impact becomes non-linear?
- How persistent are GPR shocks across different time horizons?
- Beyond Correlation: Moves to causal information transfer analysis
- Multi-scale Analysis: Different time horizons (short-term vs long-term memory)
- Regime Detection: Identifies market conditions where GPR matters most
- Risk Management Applications: Dynamic hedging based on entropy measures
- Econometric Analysis: R/Python (MFDFA, transfer entropy packages)
- Time Series Modeling: ARCH/GARCH, regime-switching models
- Statistical Testing: Bootstrap methods for entropy significance
- Data Visualization: Interactive plots of multi-scale relationships
- Dynamic Portfolio Allocation: Adjust gold/equity exposure based on GPR entropy signals
- Risk Management: Early warning signals from transfer entropy spikes
- Trading Strategies: Non-linear pattern recognition in risk-off environments
- Academic Contribution: Methodological advancement in financial econometrics
This project bridges financial economics and complex systems theory, demonstrating how physics-inspired methods (DFA, entropy) can extract actionable insights from financial markets during geopolitical turbulence.
Part of a portfolio showcasing advanced quantitative research capabilities with practical financial applications.