Prediction Markets Split Geographically and Vertically
Kurush Dubash, CEO of Dome, on unified API for prediction markets
The big strategic point is that prediction markets are starting to look less like one global winner take all exchange, and more like a stack with a few giant liquidity hubs at the center and many smaller products built around local audiences or narrow use cases. Kalshi and Polymarket already concentrate most volume, but sports, politics, crypto, and local events each reward specialized distribution, local knowledge, and tailored market menus.
-
One path to fragmentation is geographic. Dome described inbound demand from teams in Latin America, South Africa, India, and Australia that want markets for local elections, local sports, or regional communities that a U.S. focused or global platform will not list deeply enough.
-
The other path is vertical. Even among the biggest platforms, specialization is already visible. Polymarket has been strongest in politics, while Kalshi has led more in sports, showing that liquidity can still cluster by category even before smaller niche products scale.
-
That creates the opening for infrastructure like Dome. If the same real world event appears in different formats across several venues, an aggregator can map them together, show combined depth, and route orders to the best venue, similar to odds comparison and routing in sportsbooks.
Going forward, the category is likely to settle into a barbell structure. A few large exchanges will own the deepest pools of liquidity, while regional and category specific products own distribution and community. That makes aggregation, market matching, and routing more important as prediction markets spread into more countries, apps, and industry specific workflows.