opes 0.8.1

A research-focused portfolio optimization and backtesting engine.

OPES

An Open-source Portfolio Estimation System for advanced portfolio optimization and backtesting.

---

Overview

OPES is a comprehensive Python library for advanced portfolio optimization and backtesting. It is designed for quantitative finance enthusiasts who wants to explore cutting-edge methods for portfolio research or understanding. OPES aims to be research first, trying to prioritize sheer breadth over depth for each paradigm. OPES aims provide a wide range of portfolio strategies, risk measures and robust evaluation tools for users to always stay updated with current portfolio optimization techniques. OPES's inspiration grew from my own project, MOP, along with the giants of portfolio optimization: PyPortfolioOpt and Riskfolio-Lib.

Visit the documentation for detailed insights on OPES.

---

Disclaimer

The information provided by OPES is for educational, research and informational purposes only. It is not intended as financial, investment or legal advice. Users should conduct their own due diligence and consult with licensed financial professionals before making any investment decisions. OPES and its contributors are not liable for any financial losses or decisions made based on this content. Past performance is not indicative of future results.

---

Portfolio Objectives

Classification	Name
Utility Theory	Quadratic Utility
	Constant Relative Risk Aversion (CRRA)
	Constant Absolute Risk Aversion (CARA)
	Hyperbolic Absolute Risk Aversion (HARA)
	Kelly Criterion & Fractional Kelly
Markowitz Paradigm	Maximum Mean Return
	Minimum Variance
	Mean-Variance
	Maximum Sharpe Ratio
Principled Heuristics	Uniform (1/N)
	Risk Parity
	Inverse Volatility
	Softmax Mean
	Maximum Diversification
	Return Entropy Portfolio Optimization
Risk Measures	Value at Risk (VaR)
	Conditional Value at Risk (CVaR)
	Mean-CVaR
	Entropic Value at Risk (EVaR)
	Mean-EVaR
	Entropic Risk Measure
	Worst-Case Loss
Online Learning	Cover's Universal Portfolios
	Best Constant Rebalanced Portfolio (BCRP)
	Exponential Gradient
Distributionally Robust Optimization	KL-Ambiguity Robust Maximum Mean Return
	KL-Ambiguity Robust Kelly
	KL-Ambiguity Robust Fractional Kelly
	Wasserstein-Ambiguity Robust Maximum Mean Return
	Wasserstein-Ambiguity Robust Minimum Variance
	Wasserstein-Ambiguity Robust Mean-Variance

Metrics

Portfolio Metrics	Backtest Metrics
Tickers	Sharpe
Weights	Sortino
Portfolio Entropy	Volatility
Herfindahl Index	Average Return
Gini Coefficient	Total Return
Absolute Maximum Weight	Mean Drawdown
	Maximum Drawdown
	Geometric Growth Rate
	Value-at-Risk 95
	Conditional-Value-at-Risk 95
	Skew
	Kurtosis
	Omega Ratio
	Ulcer Index
	Hit Ratio

Other Features

Slippage Models	Regularization Schemes
Constant	L1 Regularization
Gamma	L2 Regularization
Lognormal	L-infinity Regularization
Inverse Gaussian	Entropy
Compound Poisson-Lognormal	Weight Variance
	Mean Pairwise Absolute Deviation
	KL-Divergence from Uniform (Experimental)
	JS-Divergence from Uniform (Experimental)

---

Installation

Using pip

If you have pip, it is very convenient to install opes.

pip install opes

From the Source

Alternatively, you are also welcome to install directly from the GitHub repository.

git clone https://github.com/opes-core/opes.git
cd opes
pip install .

You can also install in editable mode if you plan on making any changes to the source code.

# After cloning
pip install -e .

Verification

Verify your installation by using pip.

pip show opes

You can also verify by using python.

>>> import opes
>>> opes.__version__

---

Getting Started

opes is designed to be minimalistic and easy to use and learn for any user. Here is an example script which implements my favorite portfolio, the Kelly Criterion.

# I recommend you use yfinance for testing.
# However for serious research, using an external, faster API would be more fruitful.
import yfinance as yf

# Importing our Kelly class
from opes.objectives.utility_theory import Kelly

# Obtaining ticker data
# Basic yfinance stuff
TICKERS = ["AAPL", "NVDA", "PFE", "TSLA", "BRK-B", "SHV", "TLT"]
asset_data = yf.download(
    tickers=TICKERS, 
    start="2010-01-01", 
    end="2020-01-01", 
    group_by="ticker", 
    auto_adjust=True
)

# Initialize a Kelly portfolio with fractional exposure and L2 regularization
# Fractional exposure produces less risky weights and L2 regularization contributes in penalizing concentration
kelly_portfolio = Kelly(fraction=0.8, reg="l2", strength=0.001)

# Compute portfolio weights with custom weight bounds
kelly_portfolio.optimize(data, weight_bounds=(0.05, 0.8))

# Clean negligible allocations
cleaned_weights = kelly_portfolio.clean_weights(threshold=1e-6)

# Output the final portfolio weights
print(cleaned_weights)

This showcases the simplicty of the module. However there are far more diverse features you can still explore. If you're looking for more examples, preferably some of them with "context", I recommend you check out the examples page within the documentation.

---

Testing

Tests for opes are written using the pytest module. You can run the tests easily using the following command.

cd project-root    # Navigate to project root
pytest

This will run three scripts, each dedicated to testing the optimizer, regularizer and backtesting engine. Note that the tests are heavy and can take a significant amount of time since it tests each of the available objectives. We test using a year's data for the following tickers.

GOOG, AAPL, AMZN, MSFT

The price data is stored in the prices.csv file within the tests/ directory. The number of tickers are limited to 4 since there are computationally heavy portfolio objectives (like UniversalPortfolios) included which may take an eternity to test well using multiple tickers.

Also it eats up RAM like pac-man.

---

Upcoming Features (Unconfirmed)

These features are still in the works and may or may not appear in later updates:

Objective Name (Category)
Hierarchical Risk Parity (Principled Heuristics)
Online Newton Step (Online Learning)
ADA-BARRONS (Online Learning)