Prediction markets spent the week being institutionalized from both ends. Regulators tightened the supply side as Kalshi rolled out mandatory employer disclosures and an insider-trading trial moved toward setting the ecosystem's defining precedent, while fresh research legitimized the product by showing decentralized markets out-forecast the institutions they compete with. Underneath the rules fight, the liquidity told its own story: as much as $34 billion of offshore activity traces back to American traders, and 2026 World Cup contracts pulled eight-figure daily volume on Polymarket. The vertical's defining tension is now clear, because the open participation that makes these markets accurate is the same thing regulators are pushing toward gated access.
The compliance era arrives
Kalshi implemented mandatory employer disclosure for traders in markets where material non-public information is a live risk, including Federal Reserve policy and regulatory announcements, reportedly affecting about 15 percent of its active markets and marking its first major compliance upgrade since CFTC approval. A December 2026 criminal trial of a US soldier accused of trading on classified intelligence through Polymarket is set to establish three precedents worth tracking: US prosecutors claiming jurisdiction over an offshore platform's trades, blockchain transaction data serving as primary evidence, and investigators tracking positions across multiple platforms. Senator Warren's formal CFTC inquiry questioned the agency's capacity to monitor decentralized venues, with open interest approaching $500 million across major platforms. Sports protocols pre-complied, with Azuro adding identity verification above $10,000 in monthly volume and Thales partnering with Chainalysis. The cautionary analogue: after PredictIt's 2021 restrictions, spreads widened 23 percent and retail volume fell 18 percent. Compliance trades efficiency for legitimacy.
A $34B offshore market built on regulatory arbitrage
New analysis put as much as $34 billion of offshore prediction-market activity in American hands, implying US traders represent 35 to 45 percent of global volume despite domestic restrictions. The cause is a three-way jurisdictional split: the CFTC asserting federal authority, including its challenge to New Mexico over sports betting, prior SEC support for state authority, and Kalshi's own litigation over election markets. The fragmentation does measurable damage to the thing prediction markets are supposed to be good at, because identical events show meaningful price divergence between Kalshi and Polymarket when restrictions block the cross-platform arbitrage that would normally close the gap. Liquidity concentrates offshore while domestic markets stay thin under participation limits.
The accuracy case gets its evidence
Research gave the vertical its evidentiary spine. Polymarket posted a Brier score of 0.187 on political markets against 0.234 for polling averages, 99.2 percent resolution accuracy across 2,847 resolved markets, and identified winners in 89 percent of gubernatorial races against 82 percent for polling aggregators. A separate analysis found decentralized markets out-aggregate centralized ones on structure: PredictIt caps positions at $850 while Polymarket runs uncapped with single positions above $500,000, decentralized markets adjusted to new information about 14 minutes faster than Kalshi equivalents, and participation was flatter, with Kalshi concentrating 67 percent of volume among its top 100 traders against 43 percent on Polymarket. The weakness is as instructive as the strength: Fed policy markets systematically underprice changes because the user base is retail crypto traders rather than fixed-income specialists.
Where the volume actually went
While regulators argued, traders moved into the World Cup. Three single Polymarket contracts each posted eight-figure days this window: a United States outcome market at $23.61 million, a Turkey match contract at $15.21 million, and a Curacao qualifier at $8.12 million, single-market shares of roughly 6 to 15 percent of platform-wide daily volume, which is unusually concentrated for sports contracts this far from resolution. International elections added demand, with Peru's presidential market hitting $5.73 million in a day and Egypt's World Cup market reaching $5.55 million alongside a Polymarket partnership with OneFootball, Europe's largest football platform at more than 100 million monthly users. Platforms also opened 2026 midterm Balance of Power contracts early, best read as infrastructure rather than signal, since meaningful liquidity historically arrives 6 to 12 months out.
Oracles carry the weight
All of that volume ultimately depends on resolution, and the window underlined oracle design as the load-bearing layer. UMA's optimistic oracle resolved 99.3 percent of contested outcomes without requiring governance intervention, the track record that makes uncapped, high-stakes markets credible to participants who would otherwise fear arbitrary settlement, while Chainlink provides redundancy for objective data feeds at higher latency. International sporting events introduce real edge cases, where results can face protest or disciplinary review before they finalize, which keeps oracle reliability the soft underbelly of the entire vertical.
The composable read
Through Fensory's lens, the Home for Composable Finance, prediction markets are not a standalone betting category but a composable layer in the same stack as DeFi and RWA. Their liquidity is denominated and settled in stablecoins, mostly USDC, and their outcomes are resolved by DeFi-native oracle infrastructure like UMA, so a forecasting market is already a DeFi application wearing a different label. The composability runs further than settlement, toward leveraged prediction trading once markets connect to lending protocols and hedging instruments once they connect to derivatives venues, which is where RWA reconnects: a tokenized Treasury could serve as the margin or resolution collateral behind a forecasting position, turning idle on-chain yield into productive prediction-market capital. The risk travels the same rails, because an oracle failure or smart contract exploit does not just void a bet, it impairs whatever collateral and leverage were stacked behind it, and regulatory fragmentation pushes that liquidity into the least-governed venues.
Risk Considerations: Prediction-market participation carries risk of total capital loss, regulatory enforcement, resolution disputes, and manipulation in low-liquidity events, and offshore platform use by US residents may violate federal and state rules. Oracle manipulation is an attack vector absent from centralized systems, concentrated volume in long-dated or low-probability contracts can reflect speculation rather than true probability, and accuracy statistics reflect historical records that do not guarantee future calibration.
Sources
Fensory Intelligence briefs consolidated into this weekly:
- Predict Intelligence Brief: Jun 8 to Jun 10, 2026
- Predict Intelligence Brief: Jun 13 to Jun 15, 2026
- Predict Intelligence Brief: Jun 14 to Jun 16, 2026
External sources cited in the underlying reporting: Polymarket, Kalshi, UMA, The Block, CoinDesk.
Analysis window: June 9 to June 16, 2026. Research, not advice.