Shaun Kelley noticed something unusual in Kalshi's Sunday numbers. For a platform processing nearly a billion dollars in NFL trading volume across the opening week of the season, the Sunday uplift was smaller than the college football Saturday that preceded it — by only $34 million. In a market where NFL Sundays are supposed to dwarf everything else, that is a number worth sitting with.
Kelley, the Bank of America analyst tracking week-one prediction market flows, has a theory: Kalshi may be over-indexed on college sports. That would explain why the platform's market share fell from 86% to 82% in a single week even as its raw NFL volume approached a billion dollars and the overall market hit $15.9 billion, up 16% from the prior week. Volume records do not immunize you from share loss. They just make the share loss easier to miss.
What Kelley's data suggests is that the more interesting event of week one was not Kalshi's volume — it was Robinhood's. Just before the season opened, Robinhood took a minority stake in Crypto.com and began routing NFL contracts through OG.com, Crypto.com's regulated platform. The result: Crypto.com saw a 40% weekly volume increase. That alone does not explain Kalshi's Sunday anomaly, but it fits. When a major retail brokerage with tens of millions of accounts starts directing NFL flow somewhere new, the effect does not announce itself through a press release. It shows up in a Sunday number that is $34 million lighter than it should be.
This is the part of prediction market structure that volume headlines consistently obscure. Kalshi still has a nine-fold lead over Polymarket in raw trading. Polymarket still led the category in app downloads — 762,000 in a single week, a figure its team released in a statement. But app downloads and trading volume are measuring different things. Polymarket is acquiring users. Kalshi is processing capital. The Robinhood deal is doing something else entirely: it is inserting a distribution layer between the contract and the retail customer, which is a different kind of competitive threat than either metric captures.
I have spent enough time watching infrastructure races in financial markets to know that distribution usually wins. Not the best product — the one in the most hands at the moment of decision. The question for Kalshi is whether OG.com, with Robinhood's routing behind it, represents genuine competition or a short-term spike that normalizes in week three when the novelty fades. The Sunday data suggests it was not nothing.
The CLARITY Act's failure in the Senate this week removes one of the variables that would have clarified the whole competitive picture. Without federal preemption settled, every volume record that Kalshi, Polymarket, and now Crypto.com post is being set inside a legal structure that the Ninth Circuit and several state attorneys general are actively contesting. The records are real. The legal ground beneath them is not.
Distribution usually wins in financial market infrastructure races—not the best product, but the one in the most hands at the moment of decision. When Robinhood routes customer orders through Crypto.com's OG.com platform, it inserts itself as a competitive threat that raw trading volume and app downloads alone do not capture. The question for Kalshi is whether this represents genuine structural competition or a short-term spike that normalizes as novelty fades.
The CLARITY Act's failure this week removed federal preemption as a settled variable in prediction markets. Every volume record posted by Kalshi, Polymarket, and Crypto.com is now being set inside a legal structure that the Ninth Circuit and several state attorneys general are actively contesting, meaning trading volume records remain real but their legal foundation remains uncertain and subject to ongoing litigation.