Here is the direct answer: the gap between the bid and the ask on a prediction market is a real cost, and it works exactly like sportsbook vig. On any binary market, add the YES ask to the NO ask and you get 100 cents plus the spread, so a 3-cent spread makes a 103% book the same way a -110/-110 line makes a 104.76% book. That cost applies whether the venue's fee line reads 7% or zero, which is why "no commission" marketing deserves the same skepticism as a sportsbook calling its odds "fees included." Nobody runs a matching engine for free. The honest venues itemize the cost; the rest of it hides in the price you click.
I spent years teaching bettors to price the juice baked into -110, so consider this the same lesson pointed at the newer window. By the end you will have one five-second check that reads the real cost off any board, and you will see it turn a sleepy futures market with a penny-sized fee line into the most expensive price on my screen.
The Quick Answer
The spread is the vig: on a binary prediction market, the two-sided cost of buying at market equals 100% plus the bid-ask spread, so a 1-cent spread is a 101% book, a 3-cent spread is a 103% book, and anything past about 4.8 cents is worse than a sportsbook's standard -110/-110 line before trading fees even enter. The conversion table, three live worked examples running from a 1-cent MLB book to a 19-cent spread that dwarfs any fee, and the exact schedule language behind "fee-free" claims are below.
What The Bid-Ask Spread Is, And Who Collects It
A prediction market has no bookmaker setting one price. It has an order book: a bid, the highest price someone will pay for YES right now, and an ask, the lowest price someone will sell it for. The gap between them is the spread, and every market order you place crosses it. Buy at the 52-cent ask when the bid sits at 51 and you paid above the book's own midpoint the moment you clicked. Reading those two numbers is the same skill as reading the price ladder, and if contract pricing itself is new to you, start with what a Kalshi price actually means or the broader primer on how prediction markets work.
When we swept real user complaints about prediction-market costs, one trader's line summed up the confusion better than any definition: "7 percent PLUS 4 percent built in vig." Our fee breakdowns cover the 7% half. The "built in vig" half is this article, and the trader had the concept exactly right while getting the number wrong, because the spread is not a fixed 4%. It floats with liquidity, from 1 cent on a busy MLB moneyline to 19 cents on a board nobody trades.
Somebody pockets that gap, and it is worth knowing who. The traders resting limit orders on both sides of a market, the makers, earn the spread as compensation for waiting and for the risk of being picked off when news breaks. Exchanges know this is a paid job: Kalshi charges resting orders a reduced maker fee on certain designated markets precisely because market making there is profitable, a dynamic we walk through in maker vs. taker on Kalshi. None of that gap evaporates into the void. It flows from the impatient side of the market to the patient side, and a market order volunteers you for the impatient side.
The One-Line Math That Turns Any Spread Into Vig
The reason I say the spread is the vig, rather than merely resembling it, is one line of arithmetic. On a binary market, buying NO at the NO ask is the same trade as selling YES at the YES bid, so the NO ask always equals 100 cents minus the YES bid. Add the two asks together:
YES ask + NO ask = YES ask + (100¢ − YES bid) = 100¢ + the spread.
That sum is the market's overround, the same number you get at a sportsbook by converting both sides of a line to implied break-even percentages and adding them. A -110/-110 spread bet asks you to break even 52.38% of the time on each side, 104.76% combined, which makes the standard sportsbook main line the exact equivalent of a 4.76-cent spread. Every prediction-market book you will ever look at translates the same way:
| Bid-Ask Spread | Two-sided cost of the book | Sportsbook translation |
|---|---|---|
| 1¢ | 101% | Tighter than any book's main line, roughly -102 each side |
| 2¢ | 102% | About -104 each side |
| 3¢ | 103% | About -106 each side |
| ~4.8¢ | 104.8% | The classic -110/-110 |
| 5¢ | 105% | Slightly worse than -110/-110 |
| 10¢ | 110% | Prop-market pricing or worse |
| 19¢ | 119% | Worse than almost anything a sportsbook posts |
The row that changes behavior is the third one. A 3-cent spread reads as trivial on screen, three pennies, but it prices each side of a coin-flip market at 51.5 cents, about -106 in sportsbook language. Bettors who would balk at laying -113 on both sides of a game will cross a 6-cent spread, the identical price, without blinking, because nobody printed the juice on the screen for them. The conversion in both directions, cents to American odds and back, is covered in our de-vig guide for Kalshi traders, and the venue-level comparison lives in Kalshi vs. sportsbooks.
Worked Examples: A Tight Book, A Thin Book And A Dead One, Priced Live
Theory is cheap, so I pulled real boards. All prices below were fetched from Kalshi's public market data in August 2026 and checked against the order book at the time; they are snapshots for teaching, not live quotes.
The tight book. An MLB moneyline market, Baltimore Orioles at Minnesota Twins. Minnesota was 51 bid, 52 ask; Baltimore 48 bid, 49 ask. One-cent spreads on both sides. Buy both asks and you commit 101 cents to collect 100 at settlement: a 101% book, tighter than any sportsbook main line, before Kalshi's taker fee. Add the published fee, about 1.75 cents per contract on each side at those prices (each order's total rounds up to the next cent), and the all-in two-sided cost lands near 104.5%, a functional tie with -110/-110. On liquid sports markets the spread is nearly free and the fee does almost all the charging, which is exactly why liquidity decides which markets are worth trading.
The thin book. Same exchange, same day: the market on the next NATO Secretary-General. Kaja Kallas was 6 bid, 11 ask, with about 615 contracts ever traded. That 5-cent spread makes the two-sided book 105% before fees, already past -110/-110, on a market whose fee line is a fraction of a cent per contract. And the two-sided number understates what a thin book does to a single trade. Market-buy at 11 when the midpoint is 8.5 and you paid roughly a 29% markup over the book's own center. Change your mind a minute later and market-sell into the 6-cent bid, and the round trip costs you 5 cents of an 11-cent position: about 45% of your money, before a single fee posts. Even one of the tightest names on that board, Alexander Stubb at 24 bid, 26 ask, prices out as a 102% book.
The dead book. The widest spread I found was a long-dated market on whether the four Bay Area teams the contract names, the Warriors, 49ers, Giants and Sharks, combine for at least one championship between June 9, 2026 and July 1, 2030. It was 30 bid, 49 ask, with 19 contracts traded in the market's lifetime. That 19-cent spread is a 119% book. No trading fee on any schedule comes within a mile of it, and it pays off the promise this article opened with: the most expensive price on my screen had almost nothing to do with fees.
One habit transfers directly from sportsbook betting: never take the first number you see. Bettors pull up the live odds screen and shop the number across every major book before laying juice; the prediction-market version of the same discipline is checking the spread, and where the same event trades both places, checking the exchange price against the books' price too.
What 'Fee-Free' Actually Covers, Per The Published Schedules
None of this means fee schedules are fake, and the two big venues are more transparent than their marketing summaries suggest. You just have to notice what the zero-fee language does and does not cover.
Kalshi's own fee schedule states, in the most recent edition we could retrieve in full while preparing this page in August 2026 (the schedule text effective February 5, 2026): "Trading fees are not charged for orders placed that are not immediately matched and are instead left as resting orders on the orderbook unless they are included in our 'Maker Fees' section." It also says flatly that "There is no settlement fee." Both statements are true, and neither touches the spread: rest an order and your fee line can read zero while the book's gap sits there un-narrowed, waiting for whoever crosses it. The taker formula, "fees = round up(0.07 x C x P x (1-P))," is itemized on your receipt; the spread never is. Kalshi revises the schedule periodically, so verify the live version before you trade; the full curve is in our Kalshi fees guide.
Polymarket's global published documentation goes further: "Makers are never charged fees. Only takers pay fees," and for one whole category, "Geopolitical and world events markets are fee-free." Taker rates elsewhere on that schedule run from 0.04 on politics to 0.05 on sports and 0.07 on crypto markets. So a geopolitics market there carries a legitimate 0% fee line, and 100% of its trading cost lives in the spread. Note the platform split, though: that schedule belongs to Polymarket's global exchange. The CFTC-regulated Polymarket US App publishes its own schedule, effective July 1, 2026: a 0.06 taker formula applied exchange-wide, with a maker rebate instead of any fee-free category. For a US trader, the fee-free treatment is the part of the marketing that does not carry over, and everywhere the cost structure is fee plus spread. A fee-free market is not a free market. It is a market where the entire toll moved into the one number the marketing does not mention.
That reframing is the whole point. "Zero commission" tells you where the cost is not. The spread tells you where it went.
How To Price A Spread Before You Click
Here is the five-second check I promised, and it works on any binary board:
- Add the spread to 100. YES ask plus NO ask equals 100 plus the spread, so a 4-cent spread is a 104% book. Compare that number to 104.76%, the -110/-110 standard. Tighter is cheap, wider is expensive, and past 110% you are paying prices no sportsbook could get away with posting.
- Halve the spread for a single trade. One market order pays roughly half the spread against the midpoint. Half a cent on an Orioles moneyline is noise; 2.5 cents against an 8.5-cent midpoint is a 29% markup, and the long-shot end of the board is where spreads hurt most.
- Count the fee on top. Spread plus taker fee is the real all-in. A 101% book with 3.5 cents of combined fees costs about the same as -110; a 105% book with tiny fees costs more.
- Use a resting order when patience is cheap. A limit order at or inside the midpoint pays a reduced fee or none, per the schedules quoted above, and stands to earn the spread instead of paying it. The cost is fill risk: the market may run off without you. Execution mechanics are in how to trade sports on Kalshi.
Run those four steps and the venue's marketing stops mattering, because you priced the market yourself. If you would rather start from plays that have already been priced against the market, our free expert picks publish daily with the numbers that justify them.
FAQ
Does the spread cost me anything if I hold to settlement? Yes. The cost was charged at entry, not exit. Buying at an 11-cent ask when the midpoint is 8.5 means you needed roughly 29% more win probability than the book's own center implied, and holding to settlement does not refund the difference.
Does a fee-free market mean free trading? No. Fee-free removes the itemized charge, not the book's gap. Run the thin-book example above as a 100-contract round trip at market: $11.69 out, $5.60 back, a $6.09 loss in which $5.00 is spread and only $1.09 is fees. A fee-free schedule erases the $1.09 and leaves the $5.00 exactly where it was.
Does the spread ever work in your favor? Yes, on the resting side. A limit order that sits on the book and gets filled earns the spread the taker paid, which is the market maker's whole business and the reason exchanges bill some resting orders a separate maker fee.
The Bottom Line
The trader who complained about "4 percent built in vig" was closer to the truth than most of the marketing aimed at winning him over. Every venue that matches buyers and sellers charges for it somewhere: sportsbooks in the odds, exchanges in a fee line, order books in the spread, usually some blend. The number is knowable in every case, which is more than bettors could say about juice for most of the industry's history. A 1-cent MLB book is a better price than your sportsbook's main line. A 19-cent book on a sleepy futures market is worse than anything the book would dare print. Same exchange, same fee schedule, and the difference between those two trades was never the commission.
Price the spread, count the fee, and let "zero commission" be a fact you verify instead of a reason you click.
One note for the compliance-minded: Kalshi and Polymarket US offer CFTC-regulated event contracts, a different legal category from sports betting, generally open at 18 and up where available; platform availability varies by state and changes often, so check each platform's own eligibility rules. Spreads, fees and thin order books are real costs, and nothing here is a promise of trading profit.
