DraftKings geographic arbitrage faces cannibalization test

The map tells the story more cleanly than any investor call. DraftKings sells sports event contracts in the eighteen states where it has no online sportsbook. In the twenty-seven states where it does operate a sportsbook, it sells nothing. That is not a product decision. It is a legal architecture, built to survive two simultaneous conversations: one with the NFL about advertising boundaries, one with investors about whether prediction markets eat into sportsbook handle.

The NFL bans prediction market advertising across its games and stadiums. DraftKings can tell the league, truthfully, that it never runs prediction market ads where it runs sportsbook ads — because the two products never share a state. It can tell investors the products barely cannibalize each other — because the data, cleanly separated by geography, will show exactly that. Both statements are accurate. Neither requires the company to answer the underlying question, which is what would happen to sportsbook handle in New Jersey or New York if prediction market contracts were available there too.

I have seen this structure before. A firm builds a compliance wall between two product lines, uses the wall as the evidence, and the evidence is real in every way except the one that matters. The cannibalization question does not disappear because the experiment was never run. It transfers to the regulator who eventually asks why the wall exists.

The enforceability problem here is not state gambling law, where Kalshi and the courts have been fighting for two years. DraftKings is a licensed sportsbook operator. Its prediction market products sit on a CFTC-designated contract market, same as Kalshi's. The geographic firewall does not come from a federal mandate. It appears to be self-imposed, designed to manage two sets of relationships simultaneously.

That design works until one of those relationships changes. If the NFL's advertising ban softens — and the league's own media deals create pressure to eventually price event contracts rather than cede the market to competitors — the geographic logic fractures. DraftKings would face a choice between expanding prediction market access into sportsbook states, which tests the cannibalization claim against real data, or holding the firewall, which hands market share in those states to Kalshi, Robinhood, and whoever follows.

The more durable legal risk is the CFTC's definition of what a sports event contract actually is. The Ninth Circuit has now said, in the context of the tribal injunction, that it will hear federalism arguments about who governs these products. DraftKings' current structure assumes the federal shield holds everywhere. If a state successfully argues that a licensed sportsbook operator selling event contracts in-state is conducting gambling, not futures trading, the firewall's geography becomes a liability rather than protection — it demonstrates the company understood the state-by-state legal exposure and managed around it.