Multi-leg strategies

Why options strategies translate to Polymarket

Options traders rarely take naked positions when they can structure a spread instead. The same instinct applies to Polymarket: a vertical caps your downside without giving up most of the upside, a pair trade hedges a directional thesis against macro noise, and box-spread arbitrage on multi-outcome markets is one of the few risk-free trades on the exchange. The Workstation's multi-leg ticket primitive lets you build any of these as one atomic unit instead of clicking together legs by hand.

Each leg is a normal CLOB order. The ticket records all legs together and submits them sequentially through the same OrderExecutor used by single-leg tickets, with rollback semantics: if any leg fails to fill, the filled legs are reversed via opposite-side market orders so you don't end up with broken structures.

Vertical spreads

Buy YES at one price and sell YES at a higher price on the same market. Your downside is capped at the net debit; your upside is capped at the difference between the two strikes minus the debit. The Polymarket version: instead of strikes, you're positioning between two prices on the same probability axis.

long YES @ 0.30, short YES @ 0.55
  max loss = (0.30 − 0.0) − net premium = ~0.30 − 0.20 = 0.10
  max gain = (0.55 − 0.30) − net premium = 0.25 − 0.20 = 0.05

Calendar spreads

Same outcome, different resolution dates. On Polymarket this maps to two markets that share an underlying question but settle at different times — for example, the daily and weekly versions of the same crypto-price question. You're positioning the term structure of implied probability, the same trade an options trader makes when they sell front-month and buy back-month vol.

Pair trades

Long one market, short a correlated one. Hedges out the macro and isolates the relative bet. Useful when you have a strong view that A will outperform B without a strong view on the absolute level. The Workstation's correlation overlay surfaces candidates from your existing positions: "you're long Trump-2024 YES; here are three negatively-correlated YES legs that hedge the systemic risk."

Box-spread arbitrage on multi-outcome markets

On multi-outcome Polymarket markets — "Who wins NYC mayor", "Who tops the World Cup group" — the YES prices of every outcome should sum to roughly $1.00 minus fees. When they don't, the divergence is a defined-profit arb that resolves to par regardless of who wins. The arbitrage scanner runs continuously across every multi-outcome market, ranking opportunities by edge after fees and one-click sending the legs to a multi-leg ticket.

Most traders never look at this because the math is fiddly. The Workstation surfaces it as a screener row — pick the row, check the spread, click submit.

Atomic submission and rollback

A multi-leg ticket isn't useful if you only get half the legs filled. The ticket persists in the order_tickets table with one trades.ticket_id back-reference per leg, then submits sequentially through the TicketExecutor worker. On per-leg failure the worker submits opposite-side market orders to reverse already-filled legs and marks the ticket rolled_back; if a rollback leg also fails, the ticket lands in failed status with notes for support reconciliation. Up to 8 legs per ticket.

Hedge assistant

Pick an existing position from your book and the assistant returns three negatively-correlated legs ranked by hedge ratio. Useful when an event is approaching and you want to lock in some of the move without closing the whole position. Built on the same correlation matrix the pair-trade builder uses.