Portfolio Correlation Calculator V2

A Streamlit web app for analyzing portfolio performance, asset behavior, and cross-asset correlation using Yahoo Finance data.

Hosted app:

• https://stock-correlation-matrix.streamlit.app

---

Overview

This app helps you move from a simple ticker list to a portfolio-level view of:

• Risk and return metrics
• Asset-level performance
• Correlation structure
• Contribution by asset
• Downloadable reporting in Excel

It is designed for quick exploratory analysis with editable portfolio inputs and immediate visual and metric feedback.

---

Key Features

• Editable portfolio table (Tickers, Allocation Percentage)
• Date-range based analysis
• Portfolio summary metrics:
  • Expected Return (annualized)
  • Volatility / StdDev (annualized)
  • Sharpe Ratio
  • Max Drawdown
  • Cumulative Return
• Asset-level metrics:
  • Annualized Return
  • Volatility
  • Cumulative Return
  • Observations
• Asset contribution table:
  • Contribution (log)
  • Contribution Share
• Correlation matrix with styled heatmap
• Price visualization modes:
  • Raw price
  • Indexed to 100 (first valid date per asset)
• One-click Excel export with multiple analysis sheets

---

How to Use

1. Set Start Date and End Date.
2. Edit the portfolio table:
   • Add or remove tickers
   • Adjust allocation percentages
3. Click Calculate.
4. Review:
   • Portfolio Summary
   • Asset Metrics
   • Asset Contribution
   • Correlation Matrix
   • Price chart and data table (Price vs Indexed)
5. Download the full .xlsx report.

---

Screenshots

If the hosted app is sleeping, click the blue reactivation button:

Portfolio input section:

Portfolio summary and contribution views:

Price chart and correlation output examples:

Price view toggle (Price vs Indexed) and data table:

Download section:

---

Technical Overview

Entry Point

• main.py Streamlit app orchestration: UI, validation, calculations, rendering, and export.

Core Modules (app_lib/)

• stock_api.py Fetches closing price history from Yahoo Finance.
• data_transform.py Log-return calculation and price normalization (base=100).
• metrics.py Asset and portfolio metrics.
• corr_matrix.py Correlation matrix computation with minimum-data guardrails.
• heatmap.py Styled correlation table.
• line_chart.py Altair line chart for price and index views.
• xlsx_summary_report.py Multi-sheet Excel export builder.
• streamlit_helper.py Streamlit display helpers (for example, total-row highlighting).

Tests

• tests/test_stock_api.py
• tests/test_data_transform.py
• tests/test_metrics.py
• tests/test_corr_matrix.py
• tests/test_heatmap.py

---

Local Setup

Install dependencies:

pip install -r requirements.txt

Run app:

streamlit run main.py

Run tests:

pytest

---

Data and Calculation Notes

• Data source: Yahoo Finance (yfinance)
• Returns used for analytics: daily log returns
  • log_return = ln(P_t) - ln(P_(t-1))
• Log return calculation rejects non-positive prices.
• Portfolio calculations use allocation weights normalized across valid assets.
• Correlation requires sufficient overlapping valid observations.

---

Input Validation Rules

The app blocks analysis and shows errors when:

• Portfolio is empty
• Any ticker is blank
• Duplicate tickers exist
• Allocation contains negative values
• Total allocation exceeds 100%
• Date range is invalid (start >= end)
• Price data is unavailable for selected assets and date range

---

Excel Export Contents

The download includes these sheets:

• parameters
• portfolio_allocation
• portfo_summary
• asset_metric
• asset_contrib
• price_history
• price_history_indexed
• correlation_matrix

---

Project History

This project began as a CS50P final project focused on stock correlation and evolved into a modular portfolio analytics app (V2) with broader metrics, validation, and richer export and reporting.