---
title: Risk Budget Portfolios With Convex Non-negative Matrix Factorization
url: https://www.ml-quant.com/papers/arxiv/2204.02757/
site: ML-Quant (https://www.ml-quant.com)
updated: 2026-09-26
license: Summaries CC BY 4.0; links go to the original sources
index: https://www.ml-quant.com/llms.txt
identifier: arXiv:2204.02757
source_url: https://arxiv.org/abs/2204.02757
featured: 2023-06-14
citations: 1
topic: Portfolio & Allocation
---


# Risk Budget Portfolios With Convex Non-negative Matrix Factorization

A risk factor budgeting portfolio allocation method using NMF outperforms classical methods for diversification in cryptocurrency and traditional asset portfolios.

- Source: https://arxiv.org/abs/2204.02757
- Identifier: arXiv:2204.02757
- Released: 2022-04-06
- First featured: Quant Letter No. 4 (2023-06-14): https://www.ml-quant.com/issues/2023-06-14/
- Citations (Semantic Scholar): 1
- Published in: not yet
- Topic: Portfolio & Allocation

## Related

- [Portfolio Vulnerability to Systemic Risk: Vine Copula and APARCH-DCC Approach](https://www.ml-quant.com/papers/repec/spr-fininn-v-10-y-2024-i-1-d-10-1186-s40854-023-00559-2/): Vine Copula and APARCH-DCC Approach: The study assesses the sensitivity and robustness of the Conditional Value-at-Risk (CoVaR) systemic risk measure, finding that CoVaR estimates vary with portfolio strategy and are especially high for cryptocurrency portfolios.
- [Market Efficiency and Capital Allocation](https://www.ml-quant.com/papers/ssrn/4853306/): The study reveals a positive and significant impact of market efficiency on capital allocation efficiency in decentralized crypto asset exchanges.
- [Optimizing Portfolio with Two-Sided Transactions and Lending: A Reinforcement Learning Framework](https://www.ml-quant.com/papers/arxiv/2408.05382/): A new portfolio management model based on Reinforcement Learning has been created for high-risk environments, combining a new environmental formulation with a Profit and Loss-based reward function, proving effective in managing risk and maintaining profitability in volatile markets like cryptocurrency.
- [Risk Budget Portfolios With Convex Non-negative Matrix Factorization](https://www.ml-quant.com/papers/ssrn/4474100/): Proposing a portfolio allocation method based on risk factor budgeting using convex Nonnegative Matrix Factorization.
- [Combining Reinforcement Learning and Barrier Functions for Adaptive Risk Management in Portfolio Optimization](https://www.ml-quant.com/papers/arxiv/2306.07013/): Reinforcement learning and barrier functions used in portfolio management framework
- [Maximally Machine-Learnable Portfolios](https://www.ml-quant.com/papers/arxiv/2306.05568/): MACE algorithm optimizes portfolio weights for predictability and profitability in stock returns.