AI adoption tops 10 percent at proprietary trading firms
A discussion paper by researchers from the ECB, AFM, Oxford, and HEC Paris examines how artificial intelligence transforms securities markets. The study finds that AI hiring exceeds 10 percent at proprietary trading firms while warning of market concentration and financial stability risks.
Proprietary traders lead hiring push
Artificial intelligence adoption in the financial sector rose to 3.5 percent of job ads in 2024, reaching 7.5 percent at multi-manager hedge funds and over 10 percent at proprietary trading firms (PTFs).
Analyzing European market data, authors Álvaro Cartea, Jean-Edouard Colliard, Thierry Foucault, Peter Hoffmann, Rob Graumans, and Jean-David Sigaux document that algorithmic trading accounts for 70 percent of Dutch blue-chip equity turnover.
Furthermore, higher market volatility concentrates trading: a doubling of stock-level volatility increases PTF market concentration by 4 percent.
Meanwhile, traditional brokers have reduced continuous limit order book trading from 65 percent to under 45 percent.
Distortions beneath the data boom
High fixed costs in compute and data infrastructure create significant economies of scale, threatening to concentrate market power among elite trading desks.
The authors caution that AI deployment often targets short-term foreknowledge rather than fundamental discovery, which can reduce price informativeness.
Additionally, autonomous reinforcement learning algorithms can inadvertently coordinate on collusive pricing structures or manipulate order books, complicating supervisory oversight.
Smarter models, shallower markets
The paper provides a rigorous and sobering taxonomy of how artificial intelligence alters market microstructures.
By prioritizing fast rent extraction over fundamental discovery, AI risks eroding true price efficiency.
Regulators must overhaul algorithmic oversight before autonomous models entrench hidden systemic risks.
Source: Artificial intelligence and financial markets
IN: