MIF Sectoral Investment Performance Attribution Model

A comprehensive financial analytics project demonstrating portfolio performance analysis, risk-adjusted returns, and Brinson-Fachler attribution methodology for Mutapa Investment Fund's multi-sector portfolio.

📊 Project Overview

This project implements a sophisticated performance attribution system for analyzing investment returns across 8 sectors (Mining, Energy, ICT, Transport, Agriculture, Financials, Real Estate, Manufacturing) with $375.98 billion in assets under management.

Note: This is an independent project using simulated data for demonstration purposes. It serves as a tribute to Mutapa Investment Fund and showcases financial data science capabilities.

🎯 Key Features

• Brinson-Fachler Performance Attribution Model

  • Allocation Effect Analysis
  • Selection Effect Analysis
  • Interaction Effect Decomposition

• Risk-Adjusted Performance Metrics

  • Sharpe Ratios by Sector
  • Volatility Analysis
  • Risk-Return Profiles

• Time Series Analysis

  • Rolling Performance Windows
  • Stationarity Testing
  • Trend Decomposition

• Scenario Analysis

  • Commodity Price Shock Simulations
  • Market Downturn Scenarios
  • Monte Carlo Simulations

• Interactive Power BI Dashboard

  • Executive Overview
  • Performance Attribution Breakdown
  • Sector Deep Dive
  • Market Context Analysis

🛠️ Technical Stack

Programming Languages:

• Python 3.9+ (Data Processing & Analysis)
• DAX (Power BI Measures)
• SQL (Data Relationships)

Libraries & Tools:

• Data Analysis: Pandas, NumPy
• Statistical Modeling: Statsmodels, SciPy
• Visualization: Matplotlib, Seaborn, Power BI
• Testing: Pytest

Development Environment:

• Jupyter Notebook
• Power BI Desktop
• Git/GitHub

📁 Project Structure

├── data/                      # Simulated portfolio and market data
├── python_scripts/            # Core Python modules
├── notebooks/                 # Jupyter analysis notebooks
├── powerbi/                   # Power BI dashboard files
├── tests/                     # Unit tests
├── screenshots/               # Dashboard visualizations
└── requirements.txt           # Python dependencies

🚀 Getting Started

Prerequisites

Python 3.9+
Power BI Desktop
Jupyter Notebook

Installation

1. Clone the repository:

git clone https://github.com/YourUsername/mif-performance-attribution.git
cd mif-performance-attribution

2. Create virtual environment:

python -m venv venv
source venv/bin/activate  # On Windows: venv\Scripts\activate

3. Install dependencies:

pip install -r requirements.txt

4. Generate data:

python python_scripts/generate_portfolio_data.py
python python_scripts/fetch_zse_data.py
python python_scripts/fetch_commodity_prices.py

5. Run analysis:

jupyter notebook notebooks/MIF_Performance_Dashboard.ipynb

6. Open Power BI Dashboard:

• Open powerbi/MIF_Dashboard.pbix in Power BI Desktop

📈 Key Results

• Portfolio Annualized Return: 12.4%
• Portfolio Sharpe Ratio: 0.892
• Total Active Return: 2.5%
• Best Performing Sector: ICT (15% annualized)
• Largest Allocation: Mining (30.14%)

📊 Dashboard Screenshots

Executive Overview

Performance Attribution

Sector Deep Dive

Market Context

🧪 Testing

Run unit tests:

python -m pytest tests/test_attribution.py -v

📚 Methodology

Brinson-Fachler Attribution

The performance attribution follows the Brinson-Fachler methodology:

Total Active Return = Allocation Effect + Selection Effect + Interaction Effect

Where:
- Allocation Effect = (Portfolio Weight - Benchmark Weight) × Benchmark Return
- Selection Effect = Benchmark Weight × (Portfolio Return - Benchmark Return)
- Interaction Effect = (Portfolio Weight - Benchmark Weight) × (Portfolio Return - Benchmark Return)

Data Generation

This project uses simulated data with realistic characteristics:

• 5 years of monthly data (2020-2024)
• 8 sectors with distinct risk-return profiles
• 23 companies across the portfolio
• Commodity price correlations for Mining sector
• Zimbabwe Stock Exchange benchmark data

Disclaimer: All data is simulated for demonstration purposes and does not represent actual MIF holdings or performance.

⚠️ Limitations & Disclaimers

1. Simulated Data: This project uses synthetically generated data and does not reflect actual Mutapa Investment Fund performance or holdings. It is a factor to consider why Market Context page in the dashboard is incomplete

2. Educational Purpose: Created as an independent project to demonstrate financial data science skills.

3. Not Financial Advice: This analysis is for educational and portfolio demonstration purposes only.

4. Simplified Assumptions: Real-world factors such as transaction costs, taxes, and liquidity constraints are not modeled.

🎓 Skills Demonstrated

• Financial Data Analysis
• Portfolio Performance Attribution
• Risk Management & Metrics
• Time Series Analysis
• Statistical Modeling
• Data Visualization
• Business Intelligence (Power BI)
• Python Programming
• SQL Database Design
• Unit Testing & Validation

📝 Future Enhancements

• Machine learning-based return predictions
• Real-time data integration
• Additional factor models (Fama-French 5-factor)
• Web-based interactive dashboard
• Automated report generation

👤 Author

Adonis Chiruka

• LinkedIn: linkedin.com/in/adonis-chiruka-70b265323
• Email: chirukakatakudzwa61@gmail.com
• GitHub: @stilhere4huniid

🙏 Acknowledgments

• Inspired by Mutapa Investment Fund's multi-sector investment approach
• Brinson-Fachler methodology for performance attribution
• Zimbabwe financial markets context

📄 License

This project is licensed under the MIT License - see the LICENSE file for details.

---

Note: This is an independent educational project and is not affiliated with, endorsed by, or representing the Mutapa Investment Fund.