Executive Summary
Brokerages, gaming companies, fintechs, media businesses and financial institutions increasingly face the same question: how should we enter prediction markets? The answer depends less on enthusiasm for the category than on the level of regulatory burden, distribution control, economics and execution capability the company wants to own.
Build
Building a regulated US venue offers maximum control over product, pricing and economics, but a DCM is not simply a technology platform with a licence attached. CFTC designation requires demonstrating compliance with 23 core principles covering areas including surveillance, market integrity, financial resources and systems. In March, the CFTC said it was already reviewing several pending DCM applications from companies interested in prediction markets and had received additional inquiries.
Buy or Vertically Integrate
Acquiring regulated infrastructure can shorten the path, but it introduces capital and integration requirements. Robinhood illustrates the logic: after relying on third-party exchanges, it formed a joint venture with Susquehanna that acquired MIAXdx and subsequently launched Rothera, a CFTC-licensed exchange and clearinghouse. Robinhood began routing selected event contracts to Rothera in June 2026.
Distribute
A company can instead own the customer relationship while sourcing markets from other venues. Interactive Brokers now aggregates Kalshi, CME and ForecastEx inside a single interface. Cantor's institutional model goes further in another direction, providing clients access to Kalshi block trading without needing to operate the underlying exchange.
Partner
Partnership can combine complementary assets: one company may bring distribution, another regulated infrastructure, another market-making expertise and another data or sports relationships. The disadvantage is equally clear—economics, customer ownership and strategic control must be shared.
The right model therefore depends on six questions: Who owns distribution? Who owns the regulatory burden? Who supplies liquidity? How differentiated is the product? How quickly must it launch? And which part of the economics is strategically important to retain?
PSG View
For many entrants, becoming an exchange may be the wrong starting objective. As the prediction-market stack becomes more modular, companies can participate through brokerage, distribution, liquidity, technology, data or partnerships. The strategic question is not simply whether to enter prediction markets, but which layer of the value chain is worth owning.
More exchange competition should create more partnership possibilities. The companies best positioned may be those that already control scarce assets—large customer distribution, institutional relationships, proprietary data, liquidity capability or regulatory infrastructure—and use prediction markets as an extension of those advantages rather than building a standalone product from scratch.
For informational purposes only. Nothing published by Prediction Strategy Group constitutes investment, trading or legal advice.