Econoinvest • MSc research
Monte Carlo modeling for nonlinear financial data
A Matlab extension to traditional Monte Carlo simulation using entropy and coentropy alongside linear statistics.
Research question
Traditional simulation methods can preserve familiar linear statistics while missing important nonlinear characteristics of financial time series.
Contribution
For my MSc thesis at Econoinvest, I developed a Matlab extension that incorporated entropy and coentropy measures alongside linear statistics to reproduce a broader set of observed data characteristics.
Application
The work was applicable to automated-trading backtesting and value-at-risk analysis, where the quality of simulated market behavior directly affects conclusions.