Prediction Markets: A Research Guide

State of prediction markets in 2026

The industry hit $63.5B in 2025 (4× growth from 2024), Kalshi leads at $178B annualized volume with a $22B valuation, and Polymarket returned to the US in December 2025 via its CFTC-licensed QCEX acquisition. Last verified: 2026-05-31.

In 2024, Polymarket crossed $3.6B in the month around the US election — Bloomberg embedded its odds, CNN debate tickers showed implied probabilities. Then Kalshi plugged sports event contracts into Robinhood and IBKR, Polymarket acquired CFTC-licensed exchange QCEX for $112M and re-entered the US in December 2025, and ICE (NYSE parent) committed $2B total to Polymarket in a data-infrastructure bet. The category is now a recognized institutional asset class.

Total 2025 prediction-market industry volume hit $63.5B, a 4× increase from 2024's roughly $15.8B, per CertiK research. Kalshi and Polymarket together accounted for 97.5% of that figure. Kalshi recorded an all-time monthly high of $14.81B notional ($5.42B taker) in April 2026; Polymarket's April notional was $9.01B ($1.99B taker). As of late May 2026, combined weekly volume is approximately $5.9B — Kalshi $4.0B (69%), Polymarket $1.8B (31%) — per DeFiRate's live tracker.

Kalshi's annualized trading volume has risen from $52B to $178B over the prior six months, with annualized revenue exceeding $1.5B. Institutional trading volume on Kalshi is up 800% over the same period. The platform closed a $1B Series F led by Coatue at a $22B valuation on May 7, 2026 — double its December 2025 valuation of $11B.

Polymarket's total YTD volume reached $33.5B as of May 19, 2026, at a $9B valuation post the latest ICE tranche, rising to a reported ~$15B target in April 2026 raise talks. ICE completed a $2B total commitment to Polymarket in March 2026 (the final $600M tranche closed March 27) — structured as an equity investment with ICE becoming the exclusive institutional distributor of Polymarket event-data signals.

Top 4 prediction markets in 2026

Kalshi ($178B annualized, USD/CFTC, sports-dominant) vs Polymarket ($33.5B YTD, USDC global + QCEX US exchange). Augur is rebooting; Manifold is play-money. Last verified: 2026-05-31.

Kalshi — deep dive

Kalshi is the dominant CFTC-regulated US prediction market, with $22B valuation and $1.5B+ annualized revenue.

Founded 2018 by Tarek Mansour and Luana Lopes Lara (both ex-MIT, ex-Citadel). Spent four years fighting the CFTC for the right to list event contracts, won the right to list election contracts in late 2024, then scaled sports markets to ~90% of its notional volume.

Best for

US persons who want regulated, USD-denominated prediction markets. Sports markets are the dominant use case — by April 2026, roughly 90% of Kalshi notional volume sits in NBA, NFL, MLB, soccer, and tennis contracts, per CNBC and Kalshi's own disclosures. The Robinhood integration (2025) brought retail flow that Polymarket's global platform cannot access.

Trade-offs

US-only registration via SSN. Smaller political and crypto markets than Polymarket. Sports contracts face live state litigation: Nevada required Kalshi to halt sports contracts following a November 2025 federal ruling that reversed an earlier preliminary injunction; cease-and-desist orders have also been issued by Maryland, Ohio, Montana, Arizona, and Illinois. The 3rd Circuit ruled April 7, 2026 that New Jersey cannot regulate Kalshi's sports markets — but district-court splits remain. If a federal circuit ultimately rules that state gaming law can preempt the CFTC, roughly 90% of Kalshi's volume is at risk by state.

Volume and track record

$14.81B notional in April 2026 alone — an all-time monthly record; $178B annualized run rate; $1.5B+ annualized revenue; $1B raised at $22B valuation (May 7, 2026, Coatue-led with Sequoia, a16z, Paradigm, Morgan Stanley, and ARK Invest participating). Institutional trading volume up 800% in six months.

Fees

Approximately 1.2% on most contracts. No ACH deposit fees; withdrawals by USD wire or ACH.

Polymarket — deep dive

Full guide: Polymarket.

Polymarket is the deepest global prediction market for politics, crypto, and culture, and is now CFTC-licensed for US users via the QCEX exchange.

Built on Polygon by Shayne Coplan in 2020. Settled with the CFTC for $1.4M in 2022 over unregistered binary options. Then: acquired CFTC-licensed QCEX for $112M (July 2025), received CFTC designation in November 2025, re-launched in the US on December 3, 2025 — hitting no. 1 on the App Store sports chart on day one with 200,000+ waitlisted users. Full public rollout is expected mid-2026.

ICE (Intercontinental Exchange, parent of NYSE) made a total $2B equity commitment — $1B in October 2025 then a final $600M in March 2026. ICE is the exclusive institutional distributor of Polymarket event-data feeds (Polymarket Signals and Sentiment, launched February 2026). Polymarket's valuation at the October 2025 ICE round was $8B; implied secondary market valuation reached $11.6B by January 2026, and as of April 2026 Polymarket was reported in talks to raise $400M at a ~$15B valuation.

Best for

Non-US users who want the broadest global prediction markets: politics, crypto, sports, science, culture. 3,000+ active markets at any given time. US users should join the QCEX waitlist — it is CFTC-regulated but invite-only through mid-2026; the global USDC wallet-based platform no longer blocks US users outright but the QCEX regulated track is the compliant route.

Trade-offs

Polymarket US (QCEX) requires full KYC: government ID, SSN, proof of residency, live selfie. The global wallet-based platform has no KYC but is not the regulated US track. Resolution disputes on the global platform go through UMA's optimistic oracle — which has had material failures: in March 2025, a single actor controlling 25% of UMA voting power manipulated a $7M Ukraine mineral-deal market, falsely resolving it YES; a separate $237M Zelenskyy-linked market faced similar governance issues. UMA upgraded to a whitelisted-proposer model (Managed Optimistic Oracle V2) in August 2025 to limit direct manipulation, but structural conflict-of-interest risks remain. USDC on Polygon adds smart-contract and bridge risk that Kalshi does not carry.

Volume and breadth

$9.01B notional in April 2026; $33.5B YTD through May 19, 2026. Polymarket recorded $10.57B in March 2026 — first time the platform crossed $10B monthly, roughly 2.5× its October 2024 election-cycle peak. Volume distribution (May 2026 partial): sports 24%, politics/government 12%, finance/Fed 1%, entertainment 2%, other 11% — more diversified than Kalshi.

Token

Polymarket confirmed a POLY token and airdrop (CMO announcement October 24, 2025). Launch expected post-US rollout in 2026; 5–10% of supply reserved for airdrop. Exact eligibility criteria not yet public.

Fees

Currently 0% trading fee on the global platform. Post-token launch, expect a 0.1–0.3% maker/taker structure.

Augur, Manifold, and the long tail

Augur is undergoing a 2026 reboot fork. Launched on Ethereum in 2018 by the Forecast Foundation, it pioneered the decentralized oracle-resolution design that everyone copied. It never solved liquidity. In April 2026, Augur triggered a fork dispute (over an Artemis II launch market) — the intentional start of a reboot process with two competing development teams. Phase 1 (April–June): REP staked on competing outcomes. Phase 2 (June–August 2026): mandatory REP migration — holders must migrate tokens 1:1 into the new version or risk losing them. REP market cap is approximately $5M; 24h trading volume is negligible. Useful as a censorship-resistant fallback, not as a trading venue.

Manifold Markets is play-money. Real users, fake currency (Mana). Forecasting accuracy is reportedly strong — Manifold's Brier scores on geopolitical questions beat several professional forecasting firms in 2023–2024 calibration studies. Good for practice; useless for hedging.

The rest: Drift BET on Solana, Limitless on Base, Azuro for sports — combined under 3% of the market. Niche bets only.

Polymarket vs Kalshi: the decision table

Use this table to pick a venue in 30 seconds based on what you're trading. Last verified: 2026-05-31.

Best prediction market by use case

Kalshi for US persons and sports. Polymarket for global users, politics, crypto, and culture. Manifold for play-money practice. Augur for max decentralisation — near-zero volume, undergoing 2026 reboot. Last verified: 2026-05-31.

• Best prediction market for US persons — Kalshi (CFTC DCM, fully open) or Polymarket US via QCEX (CFTC-licensed, waitlist).
• Best prediction market for global users — Polymarket global (USDC/Polygon; broadest coverage, 3,000+ markets).
• Best prediction market for sports betting — Kalshi (~90% sports volume = deepest sports liquidity; unavailable in Nevada).
• Best prediction market for political events — Polymarket global (largest political markets worldwide).
• Best prediction market for crypto-related events — Polymarket global (BTC price targets, ETF approvals, hack outcomes).
• Best prediction market for Fed/macro outcomes — Kalshi (USD settlement, cleaner for hedging) or Polymarket global.
• Best prediction market for science/research outcomes — Polymarket global (broadest long-tail).
• Best prediction market for forecasting practice — Manifold Markets (play-money, no risk).
• Best prediction market for max decentralisation — Augur (Ethereum-native, censorship-resistant — but in reboot; near-zero volume).
• Best prediction market to avoid — Any venue with no clear regulatory standing and no decentralised settlement; you risk both seizure and rug.

How to read prediction market prices

Share price equals implied probability: $0.65 = 65%. Buy YES at $0.65, hold to settlement; if YES resolves, you receive $1.00 (53% return). Annualise the spread to compare contracts across timeframes. Last verified: 2026-05-31.

A YES share trading at $0.65 means the market thinks the event has a 65% chance of resolving YES. Share prices range $0.00–$1.00. If YES resolves, you receive $1.00 per share (a 53% return on a $0.65 entry). If NO, you get $0.00. The midpoint of the order book — not the last-traded print — is the cleanest probability read.

Sophisticated traders look at funding-rate-style proxies: the implied annualised yield to settlement. A YES share at $0.95 with 30 days to settlement has roughly (1 − 0.95) / 0.95 × (365 / 30) = 64% annualised yield if the event is near-certain to resolve YES. That's how to compare a 2-week NFL contract against a 6-month election contract on equal footing — convert both to annualised carry, then compare against your funding cost (approximately 4.3% on USD T-bills, or USDC lending rate on Aave for crypto-native capital).

Cheap heuristic for "is this market efficient?" — if the YES/NO spread is over 4 cents on a contract with under $50k of resting depth, the market is wide. Anything inside 2 cents with $500k+ depth (Kalshi NFL moneylines, Polymarket presidential top-line) is institutionally efficient and your edge is near zero unless you have a real information advantage.

Step-by-step walkthrough: your first trade

Pick venue, fund the account, find a liquid market, read the resolution criteria, size the position, hold or exit. Treat the first few trades as tuition — don't risk more than $200 while you learn resolution mechanics.

1. Pick venue by jurisdiction. US person: open a Kalshi account at kalshi.com (SSN + bank ACH) or join the Polymarket US waitlist. Non-US: open a Polymarket global account at polymarket.com — wallet-based (MetaMask or Rabby), USDC on Polygon.
2. Fund. Kalshi: ACH $100–500 from a US bank, clears next day. Polymarket global: deposit USDC via the in-app onramp (MoonPay, approximately 3% fee) or transfer USDC from a CEX to your Polygon address.
3. Find a liquid market. Sort by 24h volume. Anything under $100k daily volume is illiquid — fills will be wide and exits expensive.
4. Read the resolution criteria carefully. Polymarket global markets name a specific resolution source. The 2024 "Venezuela presidential election" market resolved off the National Electoral Council's official certification — a lot of traders got surprised. UMA oracle manipulation (March 2025) shows that even the settlement mechanism carries risk on thinly-governed markets.
5. Size the position. Risk no more than 1% of your account on a single contract while learning. A $500 account = $5 per market. The point is to learn resolution mechanics, not to profit on contract no. 1.
6. Hold or exit. YES and NO shares are tradeable until market close. If your thesis breaks, sell early rather than holding to a $0 resolution.
7. Book taxes. Kalshi sends a 1099-MISC (gambling income); Polymarket US/QCEX expected to do similarly; Polymarket global sends nothing — use Koinly to track USDC P&L and self-report.

Risk summary

Four risks: disputed/manipulated resolutions (UMA oracle March 2025 incident), thin long-tail liquidity (under $10k depth), state-level regulatory uncertainty for Kalshi sports (Nevada, 6 cease-and-desist states), and USDC/Polygon counterparty risk on Polymarket global. Last verified: 2026-05-31.

• Settlement and oracle risk. Polymarket global uses UMA's optimistic oracle. In March 2025, a single actor controlling 25% of UMA voting power falsely settled a $7M Ukraine mineral-deal market — moving the contract from 9% to 100% and resolving YES despite no official agreement. A separate $237M Zelenskyy-linked market faced analogous governance failures. UMA upgraded to a whitelisted-proposer model (Managed Optimistic Oracle V2, August 2025) but conflict-of-interest risks remain — token holders can trade markets they later vote to resolve. Kalshi uses internal compliance review; the 2024 "border crossings" and "Ukraine NATO" markets on Polymarket global both produced multi-day disputes with trader losses. Read the resolution source before sizing any large position.
• Liquidity risk. Long-tail markets can have under $10k of resting depth. A $5k buy on a thin contract can push the midpoint 5–8 cents, and the exit is worse. Filter: if 24h volume is under $50k, treat it as illiquid.
• Regulatory risk. Kalshi is CFTC-licensed but faces state gaming litigation in at least six states. Nevada required Kalshi to halt sports contracts after a November 2025 federal ruling. The 3rd Circuit ruled April 7, 2026 in Kalshi's favour on New Jersey, but district-court splits persist. If a federal court ultimately holds that state gaming law preempts CFTC jurisdiction, the sports contracts accounting for ~90% of Kalshi's volume are at risk by state. Polymarket US (QCEX) is now CFTC-licensed but limited to a waitlist rollout; the global platform operates internationally under Polymarket's offshore structure.
• Counterparty risk (Polymarket global). USDC custody is in smart contracts on Polygon. You are stacking smart-contract risk (Polymarket's CTF Exchange contracts), Polygon-bridge risk, and USDC/Circle issuer risk. None of these are unusual for DeFi — but together they're a real failure surface that doesn't exist on Kalshi or Polymarket US.

Looking ahead

Four signals to track over the next 12 months:

• Polymarket US full rollout. The QCEX exchange is on waitlist through mid-2026. Full public access (SSN + KYC, USD settlement) is targeted for Q3–Q4 2026. When it opens fully, Polymarket competes with Kalshi directly for US sports and political flow.
• Kalshi state litigation. The Nevada ruling (November 2025) already restricts Kalshi sports there. District-court splits may force a Supreme Court or 5th/3rd Circuit resolution. If states win, Kalshi's sports book — ~90% of volume — gets carved by jurisdiction.
• Polymarket POLY token launch. Confirmed for post-US-rollout 2026. A 5–10% supply airdrop to active traders creates incentivized wash-volume risk that will distort the YTD numbers above for the period around launch. Treat volume data skeptically around airdrop windows.
• Augur 2026 reboot. The REP migration window runs June–August 2026. REP holders who do not migrate lose their tokens. If the reboot attracts developer interest, Augur could re-enter the long-tail decentralized market niche that neither Kalshi nor Polymarket will cover. Current odds: uncertain.

---

Related: Best DEXs 2026 · Best Stablecoins 2026