PredictIt

Operated by Victoria University of Wellington (NZ) under CFTC No-Action Letter 14-130 (2014); status contested in litigation since 2022 but currently live and serving public market data.

Country: United States
Taker fee: 1000 bps
Typical gas: —

§ Notes

Undocumented public REST feed at /api/marketdata/all/ has been used by RealClearPolitics, FiveThirtyEight, and academic researchers since 2014. Returns {markets: [{contracts: [...]}]} where each contract is a binary YES/NO outcome — Storm explodes contracts into individual binary markets. Per-trader contract caps ~$850, ~5000 traders/market = thin actionability for arb but useful long-tail US politics signal. Prices are dollars 0.01-0.99 (same probability space Storm uses). PredictIt's effective trading fees are large (5% withdrawal + 10% on profits) — the 10% taker_fee_bps approximation here treats PredictIt's profit fee as a coarse hurdle for spread engine; per-contract liquidity isn't in the bulk feed and is left null. Regulatory fragility: PredictIt could be ordered offline; ingest failures should not page operator. See docs/predictit-legal/storm-compliance-notes.md.

§ Coverage

Markets ingested: 991
Active: 973
Delisted: 18
Events linked: 22
Last ingest: 15m ago

§ Top events by volume

§ Schema health

Storm runs a canary against every venue that compares live API responses against the last-known schema shape. Drift, missing fields, and unexpected null-rates raise alarms routed to the operator queue.

§ Recent markets

Who will win the next French presidential election? — Jordan Bardella predictit_contract_34691 · $0 vol · first seen 19h ago
Who will win the next French presidential election? — Édouard Philippe predictit_contract_34692 · $0 vol · first seen 19h ago
Who will win the next French presidential election? — Jean-Luc Mélenchon predictit_contract_34693 · $0 vol · first seen 19h ago
Who will win the next French presidential election? — Marine Le Pen predictit_contract_34694 · $0 vol · first seen 19h ago
Who will win the next French presidential election? — Bernard Cazeneuve predictit_contract_34695 · $0 vol · first seen 19h ago
Who will win the next French presidential election? — Gérald Darmanin predictit_contract_34696 · $0 vol · first seen 19h ago
Who will win the next French presidential election? — Sébastien Lecornu predictit_contract_34697 · $0 vol · first seen 19h ago
Who will win the next French presidential election? — Gabriel Attal predictit_contract_34698 · $0 vol · first seen 19h ago
Who will win the next French presidential election? — François Bayrou predictit_contract_34699 · $0 vol · first seen 19h ago
Will any additional country announce that it is leaving OPEC in 2026? predictit_contract_34700 · $0 vol · first seen 19h ago

§ Ingest cadence and policy

Storm ingests from this venue on the standard pull cadence (default hourly for the daemon; on-demand via storm markets pull predictit). The ingester stores the raw payload alongside the normalized fields, so schema changes can be replayed without re-fetching.

Policy note: Undocumented public REST feed at /api/marketdata/all/ has been used by RealClearPolitics, FiveThirtyEight, and academic researchers since 2014. Returns {markets: [{contracts: [...]}]} where each contract is a binary YES/NO outcome — Storm explodes contracts into individual binary markets. Per-trader contract caps ~$850, ~5000 traders/market = thin actionability for arb but useful long-tail US politics signal. Prices are dollars 0.01-0.99 (same probability space Storm uses). PredictIt's effective trading fees are large (5% withdrawal + 10% on profits) — the 10% taker_fee_bps approximation here treats PredictIt's profit fee as a coarse hurdle for spread engine; per-contract liquidity isn't in the bulk feed and is left null. Regulatory fragility: PredictIt could be ordered offline; ingest failures should not page operator. See docs/predictit-legal/storm-compliance-notes.md.