Monday, May 4, 2026

Kalshi just raised $2B at an $11B valuation. The CFTC reopened the public comment window on event contracts the same week. That is not a coincidence.

We ran into the shape of this problem at Venmo in 2014. Peer-to-peer payments looked like a feature you bolt onto a bank app. Banks treated it as a feature. We treated it as a category. The category won.

Prediction markets right now look like a feature you bolt onto a brokerage. Some shops will treat them that way. The ones who do will spend the next eighteen months explaining to their boards why a $30B category formed without them.

If your derivatives book does not have a position on event contracts by Q3, you do not have a derivatives strategy. You have a backlog.

Watched my 19-year-old place a $5 bet on the Fed's next move yesterday. He used Kalshi. He has never opened a brokerage account.

Kalshi closed $2B at $11B last week. The CFTC reopened comments on event contracts the same week. Those two things together tell you what is actually happening: a generation is forming an opinion-expression habit in a venue that did not exist three years ago, and the regulator is now writing rules around behavior that already happened.

The question for anyone running a retail brokerage is not whether prediction markets are a real product. They are. The question is whether your eighteen-year-old's first dollar of risk goes through your stack or someone else's.

We got that question wrong at one company in my career. We do not get to get it wrong twice.

The hot take this morning is that Kalshi's $11B valuation is a bubble.

The cold take is that it is underpriced.

Here is what the bubble argument misses. Prediction markets have an accidental moat. Every contract written on Kalshi or Polymarket is a CFTC-supervised position. That regulatory clarity took five years of litigation to manufacture. Any incumbent who wants to enter has to either acquire one of the two existing venues or spend three years and forty million dollars rebuilding the same regulatory posture.

Incumbents will not spend three years on a category they think is a bubble. By the time they decide it is not, the price to enter is not $11B. It is a multiple of that.

We have seen this exact playbook before. Square at $3B in 2014 looked expensive. Stripe at $9B in 2018 looked expensive. The pattern is not 'category is overhyped'. The pattern is 'incumbents underprice regulatory infrastructure'.