AI Trading Models Falter in Complex Markets

New research reveals that large artificial intelligence models struggle to replicate human trading efficiency in complex markets. In traditional double auctions, where buyers and sellers meet to reach a fair price, these AI models often prioritize securing tiny profits over completing transactions, causing trading volumes to drop.

The study found that smaller AI models were more adaptable and effective at closing gaps between buy and sell prices than their larger counterparts. This creates a paradox for financial institutions: simply deploying the most complex or expensive AI technology does not guarantee better market performance or faster trade execution.

Regulators are already paying close attention to these risks, with the Bank of England's Financial Stability Board raising concerns about 'black box' decision-making and the potential for AI systems to trigger disorderly market corrections. The National Stock Exchange (NSE) in India is actively developing oversight mechanisms to detect risks like potential market manipulation and misinformation.