Saved in:
Bibliographic Details
Main Authors: Yang, Han, Hao, Dong, Wang, Zhuohan, Shi, Qi, Li, Xingtong
Format: Preprint
Published: 2026
Subjects:
Online Access:https://arxiv.org/abs/2601.22119
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866908798669553664
author Yang, Han
Hao, Dong
Wang, Zhuohan
Shi, Qi
Li, Xingtong
author_facet Yang, Han
Hao, Dong
Wang, Zhuohan
Shi, Qi
Li, Xingtong
contents Automatically discovering formulaic alpha factors is a central problem in quantitative finance. Existing methods often ignore syntactic and semantic constraints, relying on exhaustive search over unstructured and unbounded spaces. We present AlphaCFG, a grammar-based framework for defining and discovering alpha factors that are syntactically valid, financially interpretable, and computationally efficient. AlphaCFG uses an alpha-oriented context-free grammar to define a tree-structured, size-controlled search space, and formulates alpha discovery as a tree-structured linguistic Markov decision process, which is then solved using a grammar-aware Monte Carlo Tree Search guided by syntax-sensitive value and policy networks. Experiments on Chinese and U.S. stock market datasets show that AlphaCFG outperforms state-of-the-art baselines in both search efficiency and trading profitability. Beyond trading strategies, AlphaCFG serves as a general framework for symbolic factor discovery and refinement across quantitative finance, including asset pricing and portfolio construction.
format Preprint
id arxiv_https___arxiv_org_abs_2601_22119
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Alpha Discovery via Grammar-Guided Learning and Search
Yang, Han
Hao, Dong
Wang, Zhuohan
Shi, Qi
Li, Xingtong
Computational Finance
Artificial Intelligence
Machine Learning
Automatically discovering formulaic alpha factors is a central problem in quantitative finance. Existing methods often ignore syntactic and semantic constraints, relying on exhaustive search over unstructured and unbounded spaces. We present AlphaCFG, a grammar-based framework for defining and discovering alpha factors that are syntactically valid, financially interpretable, and computationally efficient. AlphaCFG uses an alpha-oriented context-free grammar to define a tree-structured, size-controlled search space, and formulates alpha discovery as a tree-structured linguistic Markov decision process, which is then solved using a grammar-aware Monte Carlo Tree Search guided by syntax-sensitive value and policy networks. Experiments on Chinese and U.S. stock market datasets show that AlphaCFG outperforms state-of-the-art baselines in both search efficiency and trading profitability. Beyond trading strategies, AlphaCFG serves as a general framework for symbolic factor discovery and refinement across quantitative finance, including asset pricing and portfolio construction.
title Alpha Discovery via Grammar-Guided Learning and Search
topic Computational Finance
Artificial Intelligence
Machine Learning
url https://arxiv.org/abs/2601.22119