Skip to content
All library documents

A Reflective LLM Agent for Zero-Shot Cryptocurrency Trading

Article arXiv papers · Author: Yuan Li et al.
Read the original

Summary

The document describes CryptoTrade, a cryptocurrency trading agent built around a large language model. Its approach combines on-chain information, which the authors characterize as transparent and immutable, with off-chain signals such as news, which can provide timely context. A reflective mechanism reviews the outcomes of earlier decisions to inform the agent’s daily trading choices, allowing its process to incorporate trading feedback without relying on a conventional time-series baseline alone.

The authors report experiments across multiple cryptocurrencies and market conditions, claiming that CryptoTrade achieved higher returns than traditional strategies and time-series baselines. The supplied description does not identify the assets, evaluation period, return metric, transaction costs, risk-adjusted results, or safeguards against data leakage. The performance claim should therefore be read as a summary of the reported experiments, not as evidence that the agent will generalize to live trading or other market regimes.

Key ideas

  • CryptoTrade combines on-chain data with off-chain information such as news.
  • A reflective mechanism uses outcomes of earlier trades to guide later daily decisions.
  • The authors report experiments across several cryptocurrencies and market conditions.
  • They claim higher returns than traditional strategies and time-series baselines.
  • The provided description omits evaluation details needed to assess risk and live-market generalization.

Tags

CryptoMachine learningOn-chain dataSentiment
Full text
# A Reflective LLM-based Agent to Guide Zero-shot Cryptocurrency Trading


# A Reflective LLM-based Agent to Guide Zero-shot Cryptocurrency Trading









The utilization of Large Language Models (LLMs) in financial trading has primarily been concentrated within the stock market, aiding in economic and financial decisions. Yet, the unique opportunities presented by the cryptocurrency market, noted for its on-chain data's transparency and the critical influence of off-chain signals like news, remain largely untapped by LLMs. This work aims to bridge the gap by developing an LLM-based trading agent, CryptoTrade, which uniquely combines the analysis of on-chain and off-chain data. This approach leverages the transparency and immutability of on-chain data, as well as the timeliness and influence of off-chain signals, providing a comprehensive overview of the cryptocurrency market. CryptoTrade incorporates a reflective mechanism specifically engineered to refine its daily trading decisions by analyzing the outcomes of prior trading decisions. This research makes two significant contributions. Firstly, it broadens the applicability of LLMs to the domain of cryptocurrency trading. Secondly, it establishes a benchmark for cryptocurrency trading strategies. Through extensive experiments, CryptoTrade has demonstrated superior performance in maximizing returns compared to traditional trading strategies and time-series baselines across various cryptocurrencies and market conditions. Our code and data are available at https://anonymous.4open.science/r/CryptoTrade-Public-92FC/.

Shown in full with attribution under the source's licence. Licence: abstract CC0

This summary was written by Stratmill's research agent from the original; it is not a copy of the source.