HFT veteran critiques buy‑side TCA

A veteran of high-frequency trading said buy-side transaction-cost analysis still misses the market signal that matters most before a fill: real-time order toxicity. (electronictradinghub.com) Transaction-cost analysis is the scorecard trading desks use to compare their execution against benchmarks such as the mid-price, the close, or peer trades. Major vendors including Trading Technologies, Tradeweb, and Bloomberg market tools that cover pre-trade, real-time, and post-trade analysis across asset classes. (tradingtechnologies.com) (tradeweb.com) (professional.bloomberg.com) The critique is narrower than “TCA does not exist.” It says common buy-side setups still lean on post-trade reports and benchmark comparisons, while failing to surface live microstructure signals such as Volume-Synchronized Probability of Informed Trading, or VPIN, before execution quality deteriorates. (tradersmagazine.com) (electronictradinghub.com) VPIN is a way to estimate whether the flow hitting the market looks informed rather than routine. The measure, developed by David Easley, Marcos López de Prado, and Maureen O’Hara, uses equal-volume buckets instead of equal time intervals to track imbalances in buying and selling pressure. (quantresearch.org) (questdb.com) That matters because market makers widen spreads or pull back when they think they are trading against better-informed counterparties. Research papers and later technical work tied elevated VPIN to liquidity stress, though the metric’s accuracy depends heavily on parameter choices. (quantresearch.org) (osti.gov) The veteran’s second point was about determinism before the trade. In plain terms, that means an engine should react to the same market state the same way every time, so desks can block, reroute, or resize orders before slippage compounds. (electronictradinghub.com) He argued that standard reports often isolate one symptom at a time, such as reject rates or implementation shortfall, and miss the interaction between several signals. In his April 7 case study, a desk with a 50.8% reject ratio and $182,000 in measured implementation shortfall looked broken on the surface, but the analysis said the rejects were shielding the strategy from worse fills during toxic flow. (electronictradinghub.com) The backdrop is a market where buy-side firms already rely heavily on TCA. Coalition Greenwich data published in 2024 said almost 90% of buy-side equity desks conducted TCA in the prior year, and 78% ranked post-trade analysis as the most important feature. (tradersmagazine.com) (globaltrading.net) Regulation also helps explain why TCA became a reporting staple. Under Article 27 of MiFID II, firms must take “all sufficient steps” to obtain the best possible result for clients, taking account of price, costs, speed, likelihood of execution and settlement, size, and nature of the order. (esma.europa.eu) Vendors would likely answer that their platforms already offer real-time monitoring, alerts, and pre-trade models. The dispute is over whether those features amount to live, order-book-level toxicity detection and deterministic pre-trade control, or whether they remain benchmarking tools wrapped around the trade after the fact. (tradingtechnologies.com) (tradeweb.com) (professional.bloomberg.com) The thread’s core claim is that buy-side desks do not just need a better report after the close. They need a system that can see toxic flow while the order is still optional. (electronictradinghub.com)