Totalis: derivative layer for prediction markets
Derivative layer for prediction markets
Status: Active
Location: San Francisco
Batch: Spring 2026
Website: https://www.totalis.trade
Overview
Totalis lets you trade correlated events together through parlays. Instead of trading isolated prediction market events one at a time, you can combine 2–5 outcomes into a single trade based on your thesis about how they move together.
People think in second and third-order effects: if Google has the best AI model, does the S&P price go up? If Trump talks about crypto, does BTC move? Totalis lets you trade the whole thesis rather than individual events.
A parlay is simple: pick 2–5 things you think will happen together, combine them into one trade, and get a larger payout if all legs hit. Instead of putting $100 into each individual market, you can put $100 into the combined thesis, making it more capital efficient when your view is really about the events happening together.
Example theses
Google wins AI → SPY rips
- Google has the top AI model - GOOGL closes above target - SPY closes above target
Geopolitics influencing asset prices
- Trump discusses "Iran" in China - Trump says "Crypto" - Oil price above target - BTC above target
How it works
Browse markets across crypto, politics, stocks, sports, weather, and more. Add legs to a parlay. The system sends it through RFQ and market makers return a quote. You accept or reject. There's no orderbook and no central "house" that controls payout structure.
Why now
Prediction markets have exploded, but they still mostly look like single-event betting. In mature financial markets, people don't just trade single assets — they trade baskets, spreads, options, hedges, and structured products. Parlays are the first obvious primitive for prediction markets.
Totalis starts with markets that resolve fast, are data-rich, recurring, and have deep liquidity, where correlations are easiest to test.
Founders
Pravesh Mansharamani
Building derivative layer for prediction markets. Previously math @ Waterloo.
Eric Liu
Building the derivatives layer for prediction markets. Previously data @ Coinbase, MLE @ Faire, math @ UWaterloo.
Get started
Try the app at app.totalis.trade
Note: For regulatory reasons, Totalis is only available to people outside the US.