The Price of Thought: Does Test-Time Reasoning Pay in LLM Trading?

2026-09-24

AI Summary

This study investigates whether the inference costs of large language models (LLMs) can be translated into net return improvements in quantitative trading. Treating LLM inference as an economic intervention for the first time, we systematically evaluate the post-cost economic value of varying reasoning intensities in U.S. equity markets. By controlling variables and comparing models such as DeepSeek within a framework integrating multi-source data inputs, portfolio construction, and backtesting, we assess actual trading performance. Our empirical findings reveal a non-monotonic and unstable relationship between reasoning investment and trading outcomes. Specifically, additional inferential computation does not reliably enhance net returns but instead exacerbates decision volatility. This work provides critical evidence delineating cost-benefit boundaries for AI-driven financial investment.

Abstract

While inference-time reasoning in large language models (LLMs) promises better decision making, its higher computational cost may not yield better economic outcomes. Yet reasoning controls are rarely evaluated as economic interventions, where changes in model outputs must translate into better portfolios after trading costs. We conduct a controlled study of representative LLMs from the DeepSeek, GPT, and Gemini families. We vary reasoning effort while holding information available at each formation date, prompts, output formats, and portfolio construction fixed. Our evaluation covers a full year of U.S. equities under three input conditions: numerical, identifiable news, and masked news. It includes more than 800,000 asset predictions and repeated model generations. Across all three model families, additional reasoning does not produce a reliable improvement in net portfolio returns. For DeepSeek, where we examine the full progression from no reasoning to maximum reasoning, performance is nonmonotonic. Repeated generations also produce unstable treatment effects and portfolio selections, even when overall scores remain similar. These findings show that additional reasoning can change financial decisions without reliably improving their economic value, motivating validation for each task before deployment.

Problem

• Large Language Models
• Test-Time Reasoning
• Portfolio Returns
• Stock Trading
• Economic Value

Innovation

• Test-Time Reasoning
• LLM Trading
• Portfolio Returns
• Controlled Study
• Nonmonotonic Performance