Optimizing ESG Portfolios

This repository contains research code and data for ESG portfolio optimization using the Black-Litterman model, MGARCH covariance estimation, and quadratic programming.

Overview

The project compares three portfolio optimization models:

1. Model 1: Black-Litterman portfolio optimization without ESG constraints.
2. Model 2: ESG-constrained portfolio optimization with long and short positions.
3. Model 3: ESG-constrained long-only portfolio optimization.

Methods

• Black-Litterman expected returns
• MGARCH covariance estimation
• Quadratic programming
• ESG deficiency constraints
• Rolling-window portfolio rebalancing

Files

• Model 1.ipynb — baseline portfolio optimization without ESG constraints.
• Model 2.ipynb — ESG-constrained portfolio optimization with long/short positions.
• Model 3.ipynb — ESG-constrained long-only portfolio optimization.
• ESG Scores.xlsx — ESG score data.
• Market Caps.xlsx — market capitalization data.
• Stock Prices.xlsx — historical stock price data.
• Raw Dataset EDITED TICKERS 2.xlsx — cleaned ticker dataset.

Research Area

Mathematical finance, ESG investing, sustainable finance, portfolio optimization, Black-Litterman modeling, MGARCH volatility modeling, quadratic programming, and responsible investing.