Methods Matter: A Trading Agent with No Intelligence Routinely Outperforms AI-Based Traders

TL;DR

This paper demonstrates that simple, non-AI trading strategies can outperform complex AI/ML-based strategies in financial markets, highlighting methodological flaws in previous research and emphasizing the need for more rigorous testing.

Contribution

It reveals that many published AI/ML trading strategies are outperformed by extremely simple strategies, due to inadequate evaluation methods in prior research.

Findings

Simple strategies can outperform complex AI/ML traders

Extensive testing reveals flaws in previous evaluation methods

AI/ML strategies may be answering the wrong questions

Abstract

There's a long tradition of research using computational intelligence (methods from artificial intelligence (AI) and machine learning (ML)), to automatically discover, implement, and fine-tune strategies for autonomous adaptive automated trading in financial markets, with a sequence of research papers on this topic published at AI conferences such as IJCAI and in journals such as Artificial Intelligence: we show here that this strand of research has taken a number of methodological mis-steps and that actually some of the reportedly best-performing public-domain AI/ML trading strategies can routinely be out-performed by extremely simple trading strategies that involve no AI or ML at all. The results that we highlight here could easily have been revealed at the time that the relevant key papers were published, more than a decade ago, but the accepted methodology at the time of those publications involved a somewhat minimal approach to experimental evaluation of trader-agents, making claims on the basis of a few thousand test-sessions of the trader-agent in a small number of market scenarios. In this paper we present results from exhaustive testing over wide ranges of parameter values, using parallel cloud-computing facilities, where we conduct millions of tests and thereby create much richer data from which firmer conclusions can be drawn. We show that the best public-domain AI/ML traders in the published literature can be routinely outperformed by a "sub-zero-intelligence" trading strategy that at face value appears to be so simple as to be financially ruinous, but which interacts with the market in such a way that in practice it is more profitable than the well-known AI/ML strategies from the research literature. That such a simple strategy can outperform established AI/ML-based strategies is a sign that perhaps the AI/ML trading strategies were good answers to the wrong question.