---
title: Statistical Arbitrage with Deep Learning: High Returns and Outperformance
url: https://www.ml-quant.com/papers/ssrn/4553110/
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: SSRN 4553110
source_url: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4553110
featured: 2023-08-30
citations: unknown
topic: Trading, Microstructure & Execution
---


# Statistical Arbitrage with Deep Learning: High Returns and Outperformance

High Returns and Outperformance: The research introduces a new data-driven method for statistical arbitrage that uses machine learning to maximize risk-adjusted returns, outperforming all benchmark approaches in a study on daily US equities.

- Source: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4553110
- Identifier: SSRN 4553110
- Released: 2021-06-08
- First featured: Quant Letter No. 14 (2023-08-30): https://www.ml-quant.com/issues/2023-08-30/
- Citations (Semantic Scholar): not tracked
- Published in: not yet
- Topic: Trading, Microstructure & Execution

## Related

- [A Deep Learning Approach for Trading Factor Residuals](https://www.ml-quant.com/papers/arxiv/2412.11432/): A replication of the Deep Learning Statistical Arbitrage methodology on recent U.S. equity data shows strong performance, suggesting potential overfitting, specific market conditions, or insufficient accounting for transaction costs and market impact.
- [Approximately optimal trade execution strategies under fast mean-reversion](https://www.ml-quant.com/papers/arxiv/2307.07024/): A study models market quality using uncertain volatility and liquidity, studying optimal strategy approximations and providing estimation methods for the model using high-frequency data.
- [Statistical arbitrage portfolio construction based on preference relations](https://www.ml-quant.com/papers/doi/10-1016-j-eswa-2023-121906/): The article suggests a new method for portfolio construction using preference relation graphs, which can improve the performance of statistical arbitrage methods by reconciling contradictory trading signals.
- [On parametric optimal execution and machine learning surrogates](https://www.ml-quant.com/papers/arxiv/2204.08581/): A study introduces a numerical algorithm using dynamic programming and deep learning for optimal order execution, highlighting the convenience of using neural-network substitutes in stochastic control issues.
- [Trading with Concave Price Impact and Impact Decay - Theory and Evidence](https://www.ml-quant.com/papers/ssrn/4625040/): The research examines statistical arbitrage issues, taking into account the nonlinear and temporary price impact of metaorders, and shows that simple trading rules can be established even with nonparametric alpha and liquidity signals.
- [Deep attentive survival analysis in limit order books: estimating fill probabilities with convolutional-transformers](https://www.ml-quant.com/papers/arxiv/2306.05479/): A deep learning method outperforms other approaches in estimating filltimes of limit orders.