Quick answer
NBBO, the National Best Bid and Offer, is the consolidated best price across every US stock exchange, and regulation forces brokers to fill you at it or better. Prediction markets have no NBBO. Kalshi, Polymarket, and Predict.fun each run an isolated order book, the same event trades at different prices on each, and nothing obliges anyone to route your order to the best one. The consolidated view has to be built privately, by aggregating the books.
Buy a share of Apple through any US broker and a piece of regulation stands behind the fill: the trade must execute at the best displayed price in the country, or better, no matter which of the sixteen exchanges holds it. That consolidated best price is the NBBO, and it is so foundational to equities that most traders never think about it.
Buy YES on a Fed decision and no such thing exists. The price you get is the price on the one book you happen to be looking at. This article covers what the NBBO actually is, why prediction markets do not have one, what the gap costs you per trade, and how a synthetic version gets built.
What is the NBBO in equities?
The National Best Bid and Offer is the highest bid and lowest offer for a security across all US exchanges, consolidated into one quote. It exists because regulation built it: SEC Regulation NMS, adopted in 2005, requires exchanges to feed their quotes into consolidated data streams and requires brokers to execute customer orders at the NBBO or better. The market is fragmented across many venues, but the price is unified. Fragmented liquidity, one national quote, and a legal duty to honor it.
The plumbing matters as much as the rule. Securities information processors consolidate every exchange's top of book into a single feed, so best execution is checkable: there is one number your fill can be compared against. That number is what makes retail equity trading honest by default.
Why is there no NBBO for prediction markets?
Because the three ingredients are missing. There is no shared regulatory framework: Kalshi is a CFTC-designated contract market, Polymarket settles on the Polygon blockchain, Predict.fun runs on Solana, and no rule connects their quotes. There is no consolidated feed: each venue publishes its own order book through its own API, in its own format, and nothing merges them. And there is no routing obligation: no venue, and no interface built on one venue, is required to tell you a better price exists somewhere else.
Each market is a single order book per event on its own exchange. The result is exactly what equities looked like before consolidation: the same instrument trading at genuinely different prices in different rooms, with the spread between rooms invisible unless you stand in all of them at once.
The fragmentation is measurable. The same event routinely trades whole points apart across venues, a 54 cent bid on one book against a 58 cent offer on another, and because a contract price is an implied probability, that is the two exchanges disagreeing about the probability of the same world. Our live market data shows the venue split every day: nine figures of daily volume, divided across books that never see each other.
What does the missing NBBO cost a trader?
The spread between venues, on every trade where the venues disagree. A trader filling at 58 cents on one book while a 55 cent offer rests on another paid 3 points for looking at one screen. In equities that fill would be illegal. In prediction markets it is Tuesday, and it compounds: across a season of trades, single-venue execution quietly taxes every position the size of the cross-venue disagreement.
The flip side is the opportunity. Persistent cross-venue mispricing is the raw material of prediction market arbitrage, and it persists precisely because there is no NBBO forcing convergence. The traders who can see both books collect the spread. The traders who cannot, pay it.
The regulation will not save you
Equities got their NBBO because Congress and the SEC spent decades building a national market system. Prediction markets span a CFTC exchange, two blockchains, and jurisdictions that do not share a regulator. A mandated consolidated quote is nowhere on any rulemaking agenda. If a best-price view exists for these markets, someone has to build it privately.
How do you build a synthetic NBBO for prediction markets?
Aggregate the books. Pull every venue's order book over its API, match the same real-world event across venues, and compute the global best bid and best ask for each outcome in real time. Sub-second, because quotes decay fast around news. Matching is the hard part: the venues list the same election or the same game under different titles, rules, and settlement terms, and a wrong match produces a fake arbitrage.
This is precisely what Kairos builds. The terminal aggregates Kalshi, Polymarket, and Predict.fun into one book and displays the global best bid and best ask for every cross-listed event, then routes execution to the venue with the better price, which is the half the display alone cannot do. A consolidated quote you cannot act on is trivia. The prediction markets trading terminal treats the synthetic NBBO the way a broker treats the real one: as the price your fill has to beat. Teams that want the same consolidated feed programmatically pull it from the prediction market API.
Will prediction markets ever get a real NBBO?
The economics say the private version arrives first, and it already has. As institutional capital moves in, infrastructure firms are building fund-grade execution across venues, and cross-venue terminals do the same for individual traders. When enough flow routes through consolidated views, venue prices converge because the arbitrage gets collected instantly, which is how equity spreads tightened long before regulators finished writing rules.
Until then, the market has two classes of participant: traders who see one book, and traders who see all of them. The NBBO of prediction markets exists today. It is just not national. It is on the terminal. See you in the order books.