Insights
The hidden 2.4% tax on your prediction market trades
For large prediction market orders, the best displayed price can be misleading. Using a 50,000-contract Fed decision trade, we examine how smart order routing and passive execution can reduce slippage and fees.
Read articleKalshi Paid Out the Wrong Side Before the Michigan Game Was Over
What happens when you get the outcome right, but the exchange pays the other side? Kalshi prematurely settled Western Michigan vs. Michigan, then reversed the payouts.
Read articleA Unified API for Prediction Markets
Connecting to one prediction-market venue is straightforward. Connecting to several is an infrastructure problem. River's unified API makes Kalshi, Polymarket, and Polymarket US behave like one system.
Read articleTrading the economic calendar with prediction markets
Economic releases can now be traded as discrete events. The contracts are simple. Building a repeatable strategy around them is not.
Read articleTrading earnings-call mentions using LLMs
Prediction markets already aggregate information into a probability. An LLM should not throw that signal away. Market-Conditioned Prompting combines the market's prior with textual evidence.
Read articleSemantic correlations and how to trade them
A two-stage framework uses statistics to discover candidate lead-lag trades and an LLM to test whether they make economic sense. The main benefit was smaller losses.
Read article