Quantitative#070

Adverse-Selection-Adjusted Spread Sizing

Each market's spread is set from how much it actually costs you to get adversely selected, measured as the average move against you right after a fill, plus the profit margin you want. So toxic, fast-moving markets get wider spreads and quiet ones get tighter spreads automatically. The edge is correctly pricing the risk you take on every quote across very different Polymarket markets, rather than using one fixed spread everywhere.

What you need to run it

  • Historical fill markout dataset per market
  • Adverse-selection cost estimator (Glosten-Milgrom-style)
  • Dynamic per-market spread parameterization
  • Backtest harness on CLOB tape

Where this applies

Markets on Polymarket where adverse-selection-adjusted spread sizing is the natural play:

  • Will Bitcoin exceed $140k by Dec 31, 2026?
  • Will Vladimir Putin remain Russian president through Dec 31, 2026?
  • Will the 2026 Time Person of the Year be a sitting politician?

Capabilities this demands

Model / quantData ingestionCustom code / APIRisk management

At a glance

CategoryQuantitative
Requirements4
CapabilitiesModel / quant, Data ingestion, Custom code / API, Risk management
VenuePolymarket (CLOB, Polygon)

Build it

Related quantitative strategies

This is documentation, not advice. Poly Research & Robotics publishes how these strategies work because the method should be checkable — not as a recommendation to trade them. See the full strategy database (147 strategies) or the data resources directory.
Join Discord