# Portfolio Optimization on Classical and Quantum Computers Using PortFawn

**Authors:** Moein Owhadi-Kareshk, Pierre Boulanger  
**arXiv:** [2112.08998](https://arxiv.org/abs/2112.08998)  
**Date:** 2021-12-16  
**Summary page:** https://www.summarizepaper.com/en/paper/2112.08998v1/

## Key points

- Importance of portfolio diversification in minimizing investment risk
- Creating portfolios with high returns and low correlation
- Factors considered: historical performance, fundamentals, predictions, news sentiment
- Mean-variance (Markowitz) portfolio optimization algorithm commonly used
- Original version had simplified assumptions and constraints
- Introduction of PortFawn, an open-source Python library for mean-variance portfolios
- PortFawn supports classical computing (real-valued asset weights) and quantum annealing computing (binary asset weights)
- Customization of target portfolios based on specific investment goals
- Comprehensive overview of mean-variance portfolio optimization algorithm limitations
- Architecture and functionalities of PortFawn described in detail
- Demonstration of tool's application through a simple investment scenario
- PortFawn as a valuable resource for optimizing portfolios using classical and quantum computing approaches

## Summary

The paper titled "Portfolio Optimization on Classical and Quantum Computers Using PortFawn" discusses the importance of portfolio diversification in minimizing investment risk. Individuals and fund managers strive to create portfolios consisting of assets with high returns and low correlation. To achieve this, various factors such as historical performance, fundamentals, predictions, and news sentiment are considered. The well-known mean-variance (Markowitz) portfolio optimization algorithm is commonly used to manage and optimize portfolios. This algorithm takes expected returns and risks (volatility) as inputs and provides optimized weights for each asset in the target portfolio as output. However, the original version of this algorithm had simplified assumptions and constraints that limited its practical applicability. To address these limitations, the authors introduce PortFawn, an open-source Python library designed to create and backtest mean-variance portfolios. PortFawn offers simple-to-use APIs for creating and evaluating mean-variance optimization algorithms using classical computing (real-valued asset weights) as well as quantum annealing computing (binary asset weights). The tool allows customization of target portfolios based on specific investment goals. The paper provides a comprehensive overview of the background and limitations of the mean-variance portfolio optimization algorithm. It also describes the architecture and functionalities of PortFawn in detail. Additionally, the authors demonstrate how this tool can be applied in practice through a simple investment scenario. Overall, this paper presents PortFawn as a valuable resource for investors seeking to optimize their portfolios using both classical and quantum computing approaches. By addressing the shortcomings of traditional mean-variance optimization algorithms, PortFawn offers enhanced flexibility in tailoring portfolios to meet individual investment objectives.