Polymarket Bot Tutorial · Chapter 16 of 32

Statistical arbitrage on Polymarket: cross-market pairs (correlated events), Polymarket-vs-Kalshi spreads, mean reversion, and how to size stat-arb positions when markets eventually resolve.

By Harley Young, lead writer at Polymarkets.co.il. Last reviewed: May 2026.

What this chapter covers

This is chapter 16 of our 32-part series on building a Polymarket trading bot. We cover the topic in depth across the sections below. Body content for each section is being written and rolled out chapter-by-chapter; FAQ answers and references are already complete and reflect production experience from running our own trader.

  • What stat-arb means in prediction markets
  • Polymarket-vs-Kalshi spread examples
  • Pairs within Polymarket (correlated events)
  • Mean reversion vs trend continuation
  • Sizing for resolving (not perpetual) markets
  • Risk: divergence past resolution
  • Code: pairs monitor and threshold-trigger

What stat-arb means in prediction markets

This section is in active development. Want to be notified when it goes live? Contact us or watch the authors page.

Polymarket-vs-Kalshi spread examples

This section is in active development. Want to be notified when it goes live? Contact us or watch the authors page.

Pairs within Polymarket (correlated events)

This section is in active development. Want to be notified when it goes live? Contact us or watch the authors page.

Mean reversion vs trend continuation

This section is in active development. Want to be notified when it goes live? Contact us or watch the authors page.

Sizing for resolving (not perpetual) markets

This section is in active development. Want to be notified when it goes live? Contact us or watch the authors page.

Risk: divergence past resolution

This section is in active development. Want to be notified when it goes live? Contact us or watch the authors page.

Code: pairs monitor and threshold-trigger

This section is in active development. Want to be notified when it goes live? Contact us or watch the authors page.

Frequently asked questions

Can I arbitrage between Polymarket and Kalshi?
Yes, when the same event trades on both. Spreads of 2-5 cents persist between the two exchanges because of regulatory differences, fee structures, and audience differences. The catch: Kalshi is US-only and KYC-gated, so you need accounts on both - check our /guide/polymarket-vs-kalshi/ for setup. The spread arbitrage is real but operationally heavy.
What is a pairs trade in Polymarket context?
Two correlated markets where the price ratio mean-reverts. Example: "Trump wins 2028" and "Republicans win 2028 House" - if Trump trade jumps 5% but the House trade has not moved, the bot buys House and sells Trump (or vice versa) on the assumption the ratio reverts. Edge is small but consistent in liquid politics markets.
How do you size a stat-arb position when both markets resolve?
Size to your largest expected divergence - if your model says the spread can drift 8 cents wider before reverting, size such that 8 cents wider does not exceed your loss limit. Treat each pair as an independent trade with a hard loss limit. Stat-arb pretending to be free money is how accounts blow up.
Is the Polymarket-Kalshi arb persistent?
Yes, but it is shrinking as more bots discover it. Through 2026 we still see 1-3 cent spreads on liquid politics markets persist for hours. Sports markets see narrower spreads. Tail markets (low-volume) see wider but uneconomical-to-trade spreads.
What if both markets diverge through resolution?
The pairs trade goes to its full PnL based on the actual outcomes - which is why pairs trading on prediction markets is fundamentally bounded (unlike perpetuals where divergence can grow forever). Net loss on a pairs trade is capped at the price difference at entry, not infinite.
What tools help find Polymarket pairs?
A correlation matrix across active markets is the starting point. Build it from 30-day price histories pulled from gamma+CLOB. Look for pairs with rho > 0.7 historically that are currently mispriced relative to their typical ratio. Update the matrix weekly - correlations decay.