Portfolio Optimization & Risk Modeling
Developed a comprehensive portfolio optimization system by implementing methodologies from academic research papers, integrating Hierarchical Risk Parity (HRP) and Hierarchical Clustering (HC) to leverage historical market structures. Incorporated investor views using the Fama-French 3-Factor Model and Black-Litterman Model, combining data-driven insights with forward-looking market expectations.
Integrated Bayesian analysis, Markov Chains, and regime-switching models to dynamically adjust portfolios under varying market conditions, including periods like the COVID-19 pandemic and recessions. Built and trained sophisticated machine learning models for stock selection, combining sentiment analysis, fundamental ratios, and technical indicators to rank and select high-performing securities.
Extended the framework to handle multi-asset portfolios, including equities, ETFs, indices, gold and bonds. The model demonstrates robust performance and consistently outperforms the NIFTY 50 and S&P 500 benchmarks across various asset combinations and time intervals.