Alpha Discovery via Grammar-Guided Learning and Search

2026-01-29

AI Summary

This work proposes AlphaCFG, a novel framework for automatically discovering quantitative trading alpha factors that satisfy syntactic and semantic constraints while ensuring interpretability and computational efficiency. The approach constructs a scalable tree-structured search space grounded in an alpha-oriented context-free grammar and formulates factor discovery as a tree-structured language Markov decision process. For the first time, it introduces a grammar-guided mechanism that integrates grammar-aware Monte Carlo tree search with deep value and policy networks to guarantee the validity, financial interpretability, and efficiency of generated factors. Empirical evaluations on both Chinese and U.S. stock markets demonstrate that AlphaCFG significantly outperforms existing methods in terms of search efficiency and trading performance.

Technology Category

• Search and Optimization: Algorithm Configuration
• Machine Learning: Probabilistic Circuits and Graphical Models
• Natural Language Processing: Discourse, Pragmatics & Argument Mining

Application Category

• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• Economics, Online Markets and Human Computation: Economic ramifications for generative AI infrastructure and applications
• Search and Retrieval-Augmented AI: Agentic search

Abstract

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.

Problem

• alpha discovery
• quantitative finance
• symbolic factor
• grammar-guided search
• financial interpretability

Innovation

• grammar-guided learning
• context-free grammar
• Monte Carlo Tree Search
• symbolic factor discovery
• alpha factor