DraftKings volume hit $11 billion annualised as BofA upgrades to Buy

Bank of America analyst Julie Hoover upgraded DraftKings to Buy on Wednesday, holding a $27 price target on a stock that had fallen roughly 47% in the year since the company acquired Railbird Technologies. The upgrade rested almost entirely on one number: annualised trading volume growing from $2.3 billion in April to $11 billion in July, a move that happened in three months and that Hoover read as structural rather than seasonal.

The market-making composition of that volume is where the story gets interesting. Of the $11 billion annualised figure, $7.4 billion came from market-making activity and $3.6 billion from consumer trading. That split matters because market-making revenue and consumer fee revenue have completely different margin profiles and completely different legal exposures. Bank of America is projecting $40 million in prediction market fees from DraftKings in 2027, plus a separate $200 to $400 million from market-making — a range wide enough to acknowledge that the architecture of the business is still being built.

I have seen this pattern before in DeFi liquidity infrastructure: a platform that starts capturing volume through market-making before consumer flow matures ends up building a two-sided dependency that is harder to unwind than it looks. The volume numbers are real, but the consumer side — $3.6 billion annualised against 600,000 registered users by August — implies average activity per user that will compress as the novelty premium fades. Bank of America's 2027 EBITDA estimate of $1.15 billion, up from $1.05 billion before the revision, is pricing in the market-making revenue holding. That is the assumption worth pressure-testing.

The structural problem DraftKings faces is one that the Coinbase settlement in Michigan made concrete this week. Michigan's Gaming Control Board reached agreement with Coinbase to halt sports-event contracts in the state, following earlier agreements with Robinhood and a court injunction against Kalshi. The MGCB was explicit that the agreement preserves all parties' legal rights pending a Sixth Circuit ruling. That pending ruling now sits over every volume projection in every pitch deck in this sector.

DraftKings entered prediction markets after acquiring Railbird, which means it entered after the legal perimeter was already contested. Kalshi, Robinhood, and Coinbase have all absorbed state-level enforcement in Michigan, Ohio, Connecticut, and Missouri. DraftKings is arriving with 600,000 users and a market-making operation into a legal environment where the Sixth Circuit has already allowed states to enforce gambling laws against CFTC-registered entities. The volume growth is genuine. The question Bank of America's upgrade does not fully answer is what percentage of that $11 billion annualised figure is accessible in states where enforcement has not yet arrived.

The consensus read on this upgrade is that prediction markets represent a durable expansion of DraftKings' addressable market. I think the market-making revenue line is being priced with more certainty than the legal map warrants. The consumer fee business at $40 million projected for 2027 is defensible. The $200 to $400 million market-making range is the number that moves — in either direction — depending on where the Sixth Circuit lands, and that ruling is not scheduled around anyone's earnings calendar.