Portfawn: Optimizing Financial Portfolios with Classical and Quantum Techniques

Portfawn is an advanced library for constructing and optimizing financial portfolios using state-of-the-art classical and quantum optimization methods. Designed for flexibility, it allows you to fine-tune portfolio construction with customizable objectives, adaptable risk models, and cutting-edge optimization techniques—all through a single, streamlined package.

Why Choose Portfawn?

• Optimize Your Way: Achieve your portfolio goals with customizable optimization objectives:
  • Classical Objectives: Minimum variance, maximum Sharpe ratio, target returns, and more.
  • Quantum Objectives: Innovative quantum optimization for complex problems.
• Flexibility: Configure every aspect—risk models, constraints, backend systems, and more.
• Single-Command Installation: Install the package with tailored dependencies for quantum or classical setups using optional tags.
• Seamless Backtesting: Validate your strategies with automated workflows.
• User-Friendly API: Designed to simplify complex financial engineering tasks.

Key Features

Advanced Portfolio Optimization

1. Classical Optimization Objectives:

   • Minimum Variance Portfolio (MVP): Minimize portfolio risk.
   • Maximum Return Portfolio (MRP): Maximize portfolio return.
   • Maximum Sharpe Ratio Portfolio (MSRP): Maximize returns per unit risk.
   • Customizable Parameters: Define constraints, target returns, risk-free rates, and optimization precision.

2. Quantum Optimization Objectives:

   • Binary Mean-Optimal Portfolio (BMOP): Solve portfolio problems using quantum annealing.
   • Backend flexibility:
     • Simulated annealing via neal.
     • Quantum processing units (QPUs) via D-Wave.

Unified Interface

Portfawn provides a unified API for both classical and quantum optimization, making it easy to switch between methods with minimal code changes.

Installation

You can install Portfawn and its dependencies in a single command:

pip install portfawn

Getting Started

This example demonstrates how to use Portfawn to optimize portfolios with classical and quantum objectives, run backtests, and visualize results.

1. Import Libraries

import dafin
import matplotlib.pyplot as plt

from portfawn import (
    BackTest,
    EquallyWeightedPortfolio,
    MeanVariancePortfolio,
    RandomPortfolio,
    OptimizationModel,
)

2. Define Assets and Load Data

# Define assets and date range
asset_list = ["SPY", "GLD", "BND"]
date_start = "2010-01-01"
date_end = "2022-12-30"

# Load returns data
returns_data = dafin.ReturnsData(asset_list).get_returns(date_start, date_end)

3. Create and Configure Portfolios

portfolio_list = [
    RandomPortfolio(),
    EquallyWeightedPortfolio(),
    MeanVariancePortfolio(name="Minimum Variance Portfolio", optimization_model=OptimizationModel(objective="MVP")),
    MeanVariancePortfolio(name="Maximum Return Portfolio", optimization_model=OptimizationModel(objective="MRP")),
    MeanVariancePortfolio(name="Maximum Sharpe Ratio Portfolio", optimization_model=OptimizationModel(objective="MSRP")),
    MeanVariancePortfolio(name="Binary Multi-Objective Portfolio", optimization_model=OptimizationModel(objective="BMOP")),
]

4. Run Backtesting

backtest = BackTest(
    portfolio_list=portfolio_list,
    asset_list=asset_list,
    date_start=date_start,
    date_end=date_end,
    fitting_days=22,
    evaluation_days=5,
    n_jobs=2,
)
backtest.run()

5. Visualize and Save Results

fig, ax = backtest.plot_cum_returns()
plt.savefig("plot_cum_returns.png")

fig, ax = backtest.plot_dist_returns()
plt.savefig("plot_dist_returns.png")

fig, ax = backtest.plot_corr()
plt.savefig("plot_corr.png")

fig, ax = backtest.plot_cov()
plt.savefig("plot_cov.png")

Demo

Check out the demos in demo/.

License

This project is licensed under the MIT License.

Support

Need help or have feedback? Get in touch:

• Email: mkareshk@outlook.com
• Issues: GitHub Issues