Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

From Deep Learning to LLMs: A survey of AI in Quantitative Investment

Bokai Cao, Saizhuo Wang, Xinyi Lin, Xiaojun Wu, Haohan Zhang, Lionel M. Ni, Jian Guo

TL;DR

This survey maps the transformation of quantitative investment from human-crafted features and traditional statistics to deep learning and, more recently, large language models (LLMs). It analyzes how DL enhances every stage of the alpha pipeline—data processing, prediction, portfolio optimization, and order execution—while outlining the rise of LLMs as both predictors and autonomous agents capable of processing unstructured data and coordinating end-to-end decision workflows. The authors synthesize current methods, identify limitations (e.g., interpretability, latency, and alignment between linguistic sentiment and market impact), and propose future directions such as AutoML, knowledge-driven AI, and end-to-end modeling. By providing an integrated framework and practical insights, the paper aims to bridge research and real-world investment practice, guiding the development of AI-driven, adaptive quantitative-investment systems.

Abstract

Quantitative investment (quant) is an emerging, technology-driven approach in asset management, increasingy shaped by advancements in artificial intelligence. Recent advances in deep learning and large language models (LLMs) for quant finance have improved predictive modeling and enabled agent-based automation, suggesting a potential paradigm shift in this field. In this survey, taking alpha strategy as a representative example, we explore how AI contributes to the quantitative investment pipeline. We first examine the early stage of quant research, centered on human-crafted features and traditional statistical models with an established alpha pipeline. We then discuss the rise of deep learning, which enabled scalable modeling across the entire pipeline from data processing to order execution. Building on this, we highlight the emerging role of LLMs in extending AI beyond prediction, empowering autonomous agents to process unstructured data, generate alphas, and support self-iterative workflows.

From Deep Learning to LLMs: A survey of AI in Quantitative Investment

TL;DR

This survey maps the transformation of quantitative investment from human-crafted features and traditional statistics to deep learning and, more recently, large language models (LLMs). It analyzes how DL enhances every stage of the alpha pipeline—data processing, prediction, portfolio optimization, and order execution—while outlining the rise of LLMs as both predictors and autonomous agents capable of processing unstructured data and coordinating end-to-end decision workflows. The authors synthesize current methods, identify limitations (e.g., interpretability, latency, and alignment between linguistic sentiment and market impact), and propose future directions such as AutoML, knowledge-driven AI, and end-to-end modeling. By providing an integrated framework and practical insights, the paper aims to bridge research and real-world investment practice, guiding the development of AI-driven, adaptive quantitative-investment systems.

Abstract

Quantitative investment (quant) is an emerging, technology-driven approach in asset management, increasingy shaped by advancements in artificial intelligence. Recent advances in deep learning and large language models (LLMs) for quant finance have improved predictive modeling and enabled agent-based automation, suggesting a potential paradigm shift in this field. In this survey, taking alpha strategy as a representative example, we explore how AI contributes to the quantitative investment pipeline. We first examine the early stage of quant research, centered on human-crafted features and traditional statistical models with an established alpha pipeline. We then discuss the rise of deep learning, which enabled scalable modeling across the entire pipeline from data processing to order execution. Building on this, we highlight the emerging role of LLMs in extending AI beyond prediction, empowering autonomous agents to process unstructured data, generate alphas, and support self-iterative workflows.

Paper Structure

This paper contains 58 sections, 9 figures, 1 table.

Table of Contents

  1. Introduction
  2. Evolution of Alpha Strategy Investment
  3. Motivation and Contribution of this Survey
  4. Background
  5. Alpha Strategy
  6. Alpha Pipeline
  7. Deep Learning in Alpha pipeline
  8. Data Processing
  9. Raw Data
  10. Numerical data
  11. Relational data
  12. Alternative data
  13. Simulation data
  14. Features
  15. Model Prediction
  16. ...and 43 more sections

Figures (9)

  • Figure 1: The evolutionary process of Alpha investment across different stages.
  • Figure 2: The overall framework of this paper.
  • Figure 3: A typical pipeline of quantitative investment.
  • Figure 4: Examples of numerical data.
  • Figure 5: Relational data examples.
  • ...and 4 more figures