🧠 Black–Litterman Portfolio Optimizer + Monte Carlo Simulation

This project implements a complete end-to-end portfolio optimization workflow using the Black–Litterman model combined with Monte Carlo simulation to visualize and analyze before/after results.

The system:

1. Reads your portfolio (stocks.xlsx)
2. Generates an optimized allocation using the Black–Litterman model
3. Updates your Excel sheet with whole-share targets
4. Runs a Monte Carlo simulation to compare risk & return before and after optimization
5. Visualizes the two simulations side by side

Part 2 of Project:

1. Pulls the top fortune 500 companies through a webscrapper

2. Uses the one hour time frame for the past six months for the prices

3. Calculates the entropy with the Shannon Entropy Model

4. Saves the calculated stocks entropy to a sql database

   To-Do's

   Running a diagnositc on a given stock Determine whether the stock is moving due to a momentum(trend) or mean-reversion Create a strategy and backtest it

---

💼 Portfolio Composition (Example)

Below is an example set of tickers and their role in a diversified portfolio:

Ticker	Company / Sector	Typical Role in Portfolio
AAPL	Apple Inc.	Large-cap tech growth, strong innovation pipeline
AMZN	Amazon.com Inc.	E-commerce & cloud (AWS) diversification
JPM	JPMorgan Chase & Co.	Financial sector exposure (banking stability & dividends)
KO	Coca-Cola Co.	Consumer defensive stock for stability & yield
META	Meta Platforms Inc.	Communication services & AI expansion
MSFT	Microsoft Corp.	Cloud, AI, and productivity software leader
PLTR	Palantir Technologies	High-growth data analytics & AI sector exposure
SPY	S&P 500 ETF	Broad U.S. market exposure and diversification anchor
T	AT&T Inc.	Telecommunications income & dividend stability
UPS	United Parcel Service	Logistics & transport sector diversification
XOM	Exxon Mobil Corp.	Energy & commodities exposure for inflation hedge

This balanced selection provides exposure across tech, finance, consumer, industrials, energy, and telecom, which helps control overall volatility.

---

⚙️ Required Packages

numpy
pandas
matplotlib
yfinance
openpyxl
PyPortfolioOpt

🧩 How It Works

1. Black–Litterman Optimization (optimization.py) Uses PyPortfolioOpt to compute a market equilibrium prior. Instead of handcrafted subjective absolute views, this repo derives absolute views programmatically from 1-year percentage returns scaled by volatility. This yields data-driven, normalized views: View_i = (1-year return_i) / volatility_i Views are clipped (e.g., between −0.3 and 0.3) for stability.

2. Portfolio Update (updatePortfolio.py) Converts optimized weights into whole-share allocations using current prices from Yahoo Finance. Uses a largest-remainder style approach to spend leftover cash fairly and keep total allocated value ≈ original portfolio value. Writes a new sheet BL_Allocation with exactly: Tickers | Shares

3. Monte Carlo Simulation (monteCarlo.py) Simulates thousands of correlated return paths using the covariance structure from historical returns. Uses actual share counts (from the Excel sheet) to compute starting market-value weights and the initial portfolio value. Returns expected final return and risk (std) and the simulated paths for visualization.

4. Visualization (main.py) Orchestrates the workflow and shows a single figure with two side-by-side plots: Monte Carlo paths before and after optimization (normalized to start = 1).

Real Output

python3 main.py
=== Running Black–Litterman Optimization ===
[*********************100%***********************]  11 of 11 completed
[*********************100%***********************]  1 of 1 completed
/Users/kece/Library/Python/3.9/lib/python/site-packages/pypfopt/efficient_frontier/efficient_frontier.py:259: UserWarning: max_sharpe transforms the optimization problem so additional objectives may not work as expected.
  warnings.warn(

Optimized Weights:
 KO      0.1501
T       0.1467
JPM     0.1217
MSFT    0.1187
SPY     0.1175
AAPL    0.1129
XOM     0.0894
PLTR    0.0873
META    0.0556
AMZN    0.0000
UPS     0.0000
dtype: float64
[*********************100%***********************]  11 of 11 completed
Targeted value: $12,913.80 | Allocated: $12,897.50 | Cash left: $16.30
✅ Wrote 'BL_Allocation' with [Tickers, Shares] to 'stocks.xlsx'.

=== Monte Carlo Simulation (Before) ===
[*********************100%***********************]  1 of 1 completed
[*********************100%***********************]  1 of 1 completed
[*********************100%***********************]  1 of 1 completed
[*********************100%***********************]  1 of 1 completed
[*********************100%***********************]  1 of 1 completed
[*********************100%***********************]  1 of 1 completed
[*********************100%***********************]  1 of 1 completed
[*********************100%***********************]  1 of 1 completed
[*********************100%***********************]  1 of 1 completed
[*********************100%***********************]  1 of 1 completed
[*********************100%***********************]  1 of 1 completed
Return: 20.93%, Risk: 31.24%

=== Monte Carlo Simulation (After) ===
[*********************100%***********************]  1 of 1 completed
[*********************100%***********************]  1 of 1 completed
[*********************100%***********************]  1 of 1 completed
[*********************100%***********************]  1 of 1 completed
[*********************100%***********************]  1 of 1 completed
[*********************100%***********************]  1 of 1 completed
[*********************100%***********************]  1 of 1 completed
[*********************100%***********************]  1 of 1 completed
[*********************100%***********************]  1 of 1 completed
[*********************100%***********************]  1 of 1 completed
[*********************100%***********************]  1 of 1 completed
Return: 37.31%, Risk: 27.05%

=== Comparison ===
Return Improvement: 16.37%
Risk Change: -4.19%

Monte Carlo Simulation Results:

🤝 Acknowledgements

Repository developed by Keller Bice, with structural and algorithmic refinements provided by ChatGPT (OpenAI). Contributions from ChatGPT include: Refactoring your procedural script into a cleaner module/class layout Implementing a stable whole-share allocation (largest-remainder) to use budget efficiently Automating absolute-view creation using 1-year returns scaled by volatility (reduces subjectivity) Integrating Monte Carlo so actual share counts drive the simulations Creating the side-by-side visualization for before/after comparison