Related Experiment Videos
Modeling cross correlations within a many-assets market.
H E Roman1, M Albergante, M Colombo
1Dipartimento di Fisica, Università di Milano-Bicocca, Piazza della Scienza 3, 20126 Milan, Italy.
Summary
This study introduces a market simulation model with stochastic volatility, enhancing cross-asset correlations. The improved model more accurately reflects real market behavior and price variations.
Area of Science:
- Quantitative Finance
- Computational Economics
- Statistical Modeling
Background:
- Traditional one-factor models simplify asset correlations.
- Real market data often exhibits complex cross-correlations and fat tails.
Purpose of the Study:
- To develop a more realistic market simulation model.
- To capture complex cross-asset correlations and price variation distributions.
- To improve agreement with empirical market behavior.
Main Methods:
- Simulation of a many-assets market.
- Application of the one-factor model for initial correlation analysis.
- Extension of the model with autoregressive stochastic volatility.
- Comparison with empirical data from 445 Standard and Poors 500 stocks.
Main Results:
- The extended model reproduces fat tails in logarithmic price variations.
- Introduced stochastic volatility enhances cross-correlations between time series.
- The model demonstrates improved agreement with real market cross-correlations.
Conclusions:
- Stochastic volatility is crucial for realistic market simulations.
- The proposed model offers a better representation of complex market dynamics.
- This approach advances the simulation of financial markets.