Lineage aware search boosts discovery of predictive market signals
Lineage-Aware Search Control for LLM-Based Evolutionary Factor Mining
Artificial IntelligenceMachine Learning
Summary
Finding formulas that predict stock market trends can be challenging because methods often get stuck exploring similar ideas repeatedly. The authors propose using the concept of 'lineage,' which tracks how these formulas evolve over time, to guide the search more effectively. Their approach, LineageEvo, helps maintain a balance between exploring new types of formulas and improving existing ones. Tested on two major stock markets, this method finds better and more diverse predictive formulas than earlier approaches.
What this means in practice
- •For quantitative trading teams: Improve automated discovery of stock market signals by guiding formula evolution using lineage-aware methods to increase prediction quality and diversity.
- •For financial software developers: Build tools that control formula search processes with lineage information to generate a wider variety of predictive investment factors.
Authors
Xintong Liu, Chenkai Zhang, Yanzhao Ma, Peng Yang
Abstract
Formulaic alpha factor mining searches for symbolic expressions that transform market data into predictive signals. Recent LLM-based evolutionary factor mining methods use large language models as semantic variation operators. Under the conventional LLM-EA paradigm, stochastic variation and fitness-based selection can leave the search weakly directed and concentrated around a narrow set of structurally similar factors. We argue that these limitations arise because current methods mainly organize search at the individual and population levels, leaving lineage-level control underused. A lineage connects successive factor variants through evolutionary descent. We propose LineageEvo, which uses lineage in two coupled forms of search control: (1) lineage-conditioned LLM variation, which distills Lineage-Scoped Search Experience into operator-specific priors and fuses them with global operator priors, providing persistent, lineage-relevant guidance for LLM generation; and (2) lineage-stratified search allocation, which balances factor quality against lineage concentration by distributing the search budget across lineages during evolutionary search. Together, these mechanisms make semantic variation more direction aware and prevent branches from dominating the evolution, thereby promoting the discovery of predictive and diverse factors. We evaluate LineageEvo on CSI 500 and S&P 500 in terms of factor predictiveness, portfolio performance, search efficiency, and lineage- and expression-level diversity. LineageEvo achieves the strongest predictive performance on both markets, while LSPS produces a more balanced lineage distribution and a structurally broader factor pool. The results support lineage as a useful intermediate unit for controlling LLM-based evolutionary factor search.