Abstract
On a price-time-priority futures book a market maker earns a spread for supplying liquidity and pays adverse selection to the informed traders who pick off its quotes (Glosten-Milgrom). The empirical literature lacks a direct, large-scale, trueaggressor-signed measurement of how large that cost is relative to the captured spread. We supply one on CME depth-10 data (13 roots, 234 contract-days, 353,387 maker fills): a ground-truth, fill-by-fill maker-P&L decomposition splitting each passive fill into the half-spread captured at the touch (+0.654 bp) and the post-fill adverse drift the counterparty takes back (-0.661 bp). Adverse selection consumes essentially the whole captured half-spread, leaving net maker P&L indistinguishable from zero (-0.022 bp after the 0.015 bp fee, every root within ±0.13 bp). Because passive fills are charged no slippage and every unmodelled friction pushes net more negative, this is a one-sided bound: net ≤≈ 0 (Proposition 1). Read through a minimal competitive model the cancellation is consistent with the zero-profit condition of liquidity supply, and it carries a derived sufficient tick-size survival condition. The one-tick floor clears the fee only for the coarse-tick roots ZN and SI, beyond which the per-root adverse drift decides (the ≈ 0.74 bp frontier separating the survivors is descriptive, not a structural necessity). The sample concentrates in one calm 2023 month (93% of fills), so the cancellation is established cross-sectionally, within a single volatility regime. Whether the cancellation holds across volatility regimes is confounded by the resolution of the book reconstruction, which we resolve by measuring on a single nanosecond-resolution, true-aggressor-signed trades-and-quotes (TAQ) feed that spans the regimes we test and reproduces the depth-feed decomposition to 0.3pp on matched calm sessions. On that feed adverse selection over-consumes the captured spread in three crisis weeks (adverse-to-captured 105.1% SVB, 111.5% FOMC, 109.8% COVID, CIs excluding 100%) and net maker P&L deepens with volatility.
Keywords: market making, adverse selection, market microstructure, limit order book, high frequency trading, maker-taker economics, liquidity provision, bid-ask spread, order flow imbalance
Declaration of Interest
The author declares that he has no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
The views, analysis and figures are the author’s own and do not necessarily reflect those of algoseek. This content is for educational purposes only and is not investment advice.