The bid-ask spread is the gap between the best price to sell (bid) and the best price to buy (ask). On a dollar binary, if YES bids 54¢ and asks 56¢, the spread is two cents and the mid is 55¢. You cannot buy at the mid—you lift the ask or improve the book with a limit.
Spread is both a cost and a signal. Tight spreads on liquid politics often mean market makers agree on fair value near the mid. Wide spreads mean disagreement, inventory risk, thin depth, or fresh shock. Order-book foundations introduced the mechanics; here we treat spread as something you read before every entry.
Mid is not your price
Headlines quote mids. Traders pay asks and receive bids. If mid is 62¢ but you lift 64¢, your breakeven thinking starts at 64%, not 62. Round-trip economics include spread plus fees. The price-equals-probability chapter warned that implied odds are always under a specific contract and fill path.
Limit orders exist to capture part of the spread when urgency is low—as covered in the chapter on limit versus market orders. Market orders buy certainty at a known cost: the spread.
Reading spread as “confidence”
Here “confidence” means market-maker certainty about fair value, not your personal forecast confidence. A one- to two-cent spread on a marquee election contract suggests a deep, competitive book. Six to ten cents on a niche prop often means the mid is a rough guess. After major news, spreads commonly widen first, then tighten as makers return—edge for takers often lives after tightening, not in the first thirty seconds.
Tight spread does not mean undervalued YES. It frequently means efficient YES. Contrarian trades need a thesis beyond “feels high” when the book is one cent wide.
What widens spreads
Scheduled macro releases, scandals, rule ambiguity, platform position caps, thin pools on automated curves, and halts all widen spreads. Illiquidity persists when open interest and volume are low. A one-sided wide market can reflect skewed inventory rather than balanced disagreement.
Resolution uncertainty chapters matter: if oracle text is fuzzy, makers widen to protect themselves. That is not an invitation to trust the mid at face value.
What tightens spreads
Market makers returning, arbitrage bots working, high-volume days, competing venues, and late clarity as resolution nears (sometimes—final hours can also blow spreads on binary risk).
Spread through a CPI spike
Before a print, YES on a Fed-cut contract might show 54¢ / 56¢ (two-cent spread). The number hits; within thirty seconds the ask lifts while the bid sags—eight-cent spread, mid near 57¢ but meaningless for a market buyer. Five minutes later spread might narrow to five cents; twenty minutes later three. Buying 61¢ with a personal probability of 58% is negative expected value even if the event later resolves YES—you paid uncertainty tax.
Patient traders anchor to p′ and executable ask, not to the mid at T+30 seconds.
The tight-spread trap
YES 71¢ / 72¢ on a famous election feels precise. If your probability is 68%, edge is negative at either touch. The crowd and makers already agree; you need independent information, not discomfort with a high number.
CLOB versus pool “spread”
Central limit books show explicit ladders. Automated market maker pools encode friction in curve slope and fees; a thin pool can move like a wide spread. Hybrids may show a tight pool and a wide book after shock—check both quotes before blaming “the market.”
Polymarket-style patterns sometimes show the pool leading while the book lags after shocks; confidence readings differ by leg.
Spreads in cross-venue work
Arbitrage gross gaps must exceed both legs’ spreads plus fees. A three-cent headline gap between apps can vanish when you model executable prices at your size. Platform comparison material emphasizes normalizing costs before calling dislocations free money.
Rules for your book
Many disciplined traders cap market orders when spread exceeds a few cents on names they care about. Require edge to clear half the spread plus fees before aggression. Log spread at entry and exit in the journal—backtests that assume mid fills fantasize performance.
Making markets earns spread; taking pays it. Match your strategy role: passive liquidity provision versus urgent taking.
Time to expiry interaction
Far-dated obscure markets often stay wide. Final week can tighten on clarity or widen on binary lottery risk. Passive entries should accept spread today, not assume it will narrow and gift you edge later.
Time-decay chapter themes apply: the last forty-eight hours are when wide spreads hurt most if you must exit.
Spread versus manipulation
Wide both sides with low volume often means illiquidity, not conspiracy. Tight spread with huge two-way volume often means informed two-way flow. Wide spread with quick price revert may be overreaction on thin depth.
Key misconceptions
“Tight market means buy” reverses efficiency logic. “Wide market means free edge” ignores that mids lie. “Spread will narrow if I wait” is sometimes false in endgame.
Edge must clear the spread
If your p′ is 58% and the ask is 56¢ but spread is 4¢, your effective buy is not two cents cheap—you pay 58¢ or worse if you lift. A quick rule many traders use: require edge at least half the spread plus fees before market aggression. Improving the book with limits is how small accounts compete with urgent takers.
Core concepts to remember
Spread is cost and signal. Mid is not fill. Post-news widening is normal. Tight means efficient, not cheap. Log spread at entry and exit.
Buying NO instead of selling YES
When spread is wide, expressing a bearish view by buying NO sometimes lands a cleaner price than hitting the YES bid. The mechanics mirror order-book chapters: two sides, one event, different liquidity on each leg.
What comes next in this module
Spread is the bridge between signal and execution. Later chapters compare venues, blend π, and demand net edge after friction. If you ignore spread, every later calculation lies.
What comes next
Spread is the friction of now. Time decay is how the contract behaves as the calendar runs out.
Next: Time Decay: How Probabilities Change as Events Approach