Prediction markets for targeted earnings hedges
Jon Ma, CEO of Artemis, on building 24/7 AI agents for trading & investing
This matters because prediction markets let funds hedge one exact earnings thesis instead of buying or shorting an entire stock. A manager can trade a contract on a narrow outcome, like Tesla deliveries or a line item mentioned on a call, and isolate that view from everything else that moves the share price, like multiple compression, macro moves, or management tone. That makes these markets closer to single issue insurance than a normal equity hedge.
-
The products already exist in concrete form. Kalshi lists Tesla event markets and earnings mention markets tied to what a company says on its call, and Polymarket has run Tesla delivery markets that resolve off Tesla investor relations releases. The workflow is simple, buy yes or no on one metric, then settle when the company reports it.
-
This is useful when the thesis and the stock reaction can diverge. Tesla can beat on deliveries, but the shares can still fall if margins miss, guidance weakens, or investors were positioned for an even bigger beat. A prediction contract pays on the specific datapoint, not on the market's full interpretation of the quarter.
-
The current bottleneck is market depth, not idea generation. Earnings markets are growing on Kalshi and Polymarket, and platforms like Dome are normalizing those feeds into APIs, but institutional adoption still depends on enough daily volume for hedge funds to move meaningful size without moving the price themselves.
The next step is these contracts becoming a standard input beside consensus estimates and options data. As liquidity improves, funds will increasingly trade around FDA decisions, deliveries, guidance, and call language as separate exposures, and research tools that combine market implied odds with company fundamentals will become much more valuable.