How Polymarket became Wall Street's third opinion alongside analysts and economists

Until recently, Wall Street's research stack had two recognised opinions. The first was sell-side analyst consensus — the median price-target and earnings forecasts published by Goldman Sachs, Morgan Stanley, JPMorgan and the rest, aggregated by LSEG (formerly I/B/E/S), FactSet and Bloomberg. The second was economist consensus — the Bloomberg survey medians on non-farm payrolls, CPI, Fed funds, GDP, that traders watch in the 30 seconds before every release.

There is now a third opinion. Polymarket, the blockchain-based prediction market founded by Shayne Coplan in 2020, has crossed from internet-curio status into institutional-grade alternative data — and the company that ratified the transition is one most prediction-market sceptics did not expect: Intercontinental Exchange, the NYSE's parent and one of the world's largest financial-market data businesses. On March 27, 2026, ICE completed a $2 billion equity commitment to Polymarket that began with an October 2025 announcement at a roughly $9 billion pre-money valuation. The deal is being reported as a prediction-market story. Newsorga's read, and FinTech Weekly's read in its definitive piece on the structure, is that it is fundamentally a financial data infrastructure story.

What 'third opinion' means in practice

Polymarket does something neither sell-side research nor economist surveys can do. It produces a continuously updating, incentive-aligned, liquid probability signal on outcomes that traditional financial instruments cannot directly price. A sample of what Polymarket quotes on a typical Monday morning:

• The probability that the next CPI print exceeds consensus by more than 0.1 percentage point.
• The probability that the Federal Reserve cuts rates at its June 2026 meeting.
• The probability that the Senate Banking Committee marks up the CLARITY Act by end of May.
• The probability of Ukraine-Russia ceasefire extension beyond the Victory Day three-day pause.
• The probability that any of 20 named cabinet officials are dismissed within 30 days.
• The probability of a specific earnings beat / miss for a given S&P 500 name.

Each of those probabilities is a price. Each price is set by real money posted on chain via USDC, settled by UMA's optimistic oracle. The price moves in real time as participants — some retail, some specialised, some informed insiders — adjust their positions. That is the data product ICE is buying.

What ICE is actually doing

ICE's structure has three pieces:

1. The investment. $2 billion in equity, completed in two tranches — the original commitment in October 2025 plus the final $600 million cash tranche on March 27, 2026. Valuation ~$9 billion. ICE is now Polymarket's largest institutional shareholder.

2. The distribution rights. ICE becomes the exclusive global distributor of Polymarket's event-driven data to institutional capital-markets clients. Polymarket's consumer-facing platform remains run by Coplan's team. ICE runs the institutional pipe.

3. The Signals and Sentiment integration. In February 2026, ICE launched the Polymarket Signals and Sentiment tool — a normalised institutional feed that pipes Polymarket probability data into the ICE Consolidated Feed alongside NYSE equity prices, fundamental data, and corporate actions. The same plumbing that delivers AAPL last-trade to a Citadel terminal now delivers Polymarket probability on the next FOMC decision.

That last piece is the structural change. Polymarket data is no longer a separate API a hedge fund has to ingest through a custom integration. It is a feed item on the same pipe their existing infrastructure already consumes. Newsorga's read: this is what institutional-grade data looks like in operational terms. The friction to using it is now lower than the friction not to.

The 95/5 rule

The clearest published account of how hedge funds use this data comes from Karl Mattingly, CEO of Australian forecasting firm Dysrupt Labs, in an on-record interview with Business Insider's Bradley Saacks in January 2026.

Dysrupt pulls Polymarket and Kalshi data into proprietary algorithms its hedge-fund and family-office clients use to time macro positions. Mattingly's key finding: "95% of the time, the consensus from traditional sources like economists and consultants aligns with that of prediction markets. But the other 5% is a chance for traders to make money on breaking away from the pack."

On size: "The average drift from the consensus generates up to 12 basis points of uncorrelated gains." Twelve basis points is small in absolute terms; on a $5 billion macro book it is $6 million of uncorrelated alpha, recurring through any economic-release cycle where prediction markets diverge from Bloomberg's economist survey. Across a year of weekly CPI and NFP prints, that compounds.

Mattingly's structural framing of why it works: prediction markets are "the fastest way to model a known unknown." The release date for CPI is known. The number is not. Polymarket prices the number; it adjusts continuously as news breaks, as Fed speakers comment, as energy prices move; it produces a probability surface that survey medians cannot match because surveys are static for the 48-hour window before release.

Why hedge funds use the data without trading the market

Business Insider documented an important asymmetry: hedge funds are interested in the data but mostly not trading directly on Polymarket or Kalshi. Three reasons, all confirmed across the BI reporting and Newsorga's separate read:

1. Depth. A $10 billion macro fund needs to put $50-100 million behind a single high-conviction view. Polymarket liquidity, even on the most-traded contracts, rarely supports that size without significant price impact. The platform is deep enough for retail and proprietary-trading-firm sizes but not for mainstream hedge-fund position-taking.

2. Compliance. Most institutional fund compliance desks have not yet approved direct trading on a blockchain-based platform that was, until 2025, banned in the United States by federal regulators. Even with the post-2024 regulatory clearance under the new CFTC posture, internal compliance approvals have been slow.

3. Mandate scope. Many fund mandates explicitly prohibit event-contract trading as a category. Polymarket contracts settle on a yes/no basis, which most master fund agreements categorise as binary-option exposure, separately restricted.

The data, by contrast, is a read-only signal. Compliance gets a clean approval — the fund is not opening a position, posting collateral, or taking custody. It is consuming a price feed. That asymmetry is exactly what ICE is monetising.

Who is trading the platform itself

Some proprietary trading firms have moved earlier. Susquehanna International Group, the Bala Cynwyd, Pennsylvania-headquartered firm best known for options market-making, has posted prediction-market-trader job openings through 2025-2026. Newsorga's read on Susquehanna's logic: at their balance-sheet size and with options-market-making expertise, they can profitably arbitrage prediction-market prices against Treasury rate futures, SOFR futures, VIX options, and event-related single-name equity volatility. That kind of cross-instrument arbitrage benefits from being a market participant directly, not just a data consumer.

Citadel Securities, Jane Street (which Newsorga covered separately last week on its $10 billion Q1 2026 profit print), and Two Sigma have not publicly disclosed prediction-market trading desks, but multiple market-structure sources indicate prop-desk experimentation is under way at all three at non-trivial scale.

Where Polymarket fits in ICE's broader data architecture

ICE's data and analytics business generated $608 million in a single quarter in 2025 and represents one of the company's three core profit centres alongside exchanges and clearing. The Polymarket integration is the third pillar of what ICE internally calls Signals and Sentiment — a structured alternative-data layer the company has been building on top of its conventional market-data infrastructure since 2025.

The first two pillars:

1. Reddit signals. Aggregated retail-investor sentiment from WallStreetBets and related subreddits, normalised into ticker-level sentiment scores. Distributed since 2025 on the ICE Consolidated Feed.

2. Dow Jones signals. News-flow sentiment from Dow Jones's newswire and WSJ coverage, normalised by entity and topic. Distributed since 2025.

Polymarket adds probability-weighted forward-looking sentiment — the explicit money-weighted bet on what is going to happen — to a stack that already has price (NYSE), fundamentals (ICE Data Services), retail sentiment (Reddit), and news sentiment (Dow Jones). Each piece complements the others.

Newsorga's view on why this structure matters: ICE is one of the most disciplined data businesses in finance — twenty consecutive years of record revenues, $9.9 billion in annual revenue. It does not put $2 billion into a thing for fashion. It builds platforms that monetise data spreads. The Polymarket deal is a textbook ICE product-development call dressed up in crypto colours.

Risks and limits

Three structural risks to the "third opinion" framing that are worth tracking:

1. Manipulation risk. Prediction markets remain manipulable by a sufficiently well-capitalised participant willing to lose money on the contract to move the published probability — the "signal-as-narrative" problem. Polymarket's liquidity is real but not yet large enough to fully resist coordinated manipulation on second-tier contracts. Newsorga expects ongoing academic and regulatory scrutiny of this.

2. Selection bias. Polymarket is not a representative sample of the informed universe. Its participants skew toward crypto-native, US-political, and macro audiences. On topics where the truly informed participants are absent (e.g. detailed corporate-credit outcomes, niche emerging-market scenarios), the platform's probabilities may be noisier than the Dysrupt 95% agreement metric suggests.

3. Regulatory uncertainty. The 2024-2025 CFTC posture cleared Polymarket for US participation, but the underlying classification of event contracts under the Commodity Exchange Act is being litigated. A reversal under a future administration is not unthinkable. The $2 billion ICE investment is partly an institutional bet that the regulatory direction will not reverse.

What to watch through 2026

Newsorga is tracking four data points:

1. Polymarket institutional volume. Volumes on the institutional side of the ICE-distributed Signals and Sentiment product. ICE has not publicly disclosed take-up, but mid-2026 earnings calls will be the first place that number surfaces.

2. Hedge fund mandate revisions. Whether large multi-strategy funds (Citadel, Millennium, Point72) update mandates to allow direct Polymarket or Kalshi trading at scale.

3. Tokenization side-agreement. The ICE announcement flagged a joint tokenization initiative. The structure of any tokenised event-contract product that ICE brings to market will set the legal template for the entire institutional event-contract space.

4. The Kalshi parallel. Kalshi is the CFTC-regulated US event-contract exchange that has been growing alongside Polymarket. Kalshi's institutional adoption will tell us whether the "third opinion" framing belongs uniquely to Polymarket or extends across the whole prediction-market category.

Bottom line. The phrase "third opinion" is not marketing. It is what serious hedge-fund research operations now treat Polymarket data as — a continuously updating, money-weighted forecast layer that sits structurally alongside sell-side analyst consensus and economist survey consensus in their decision stacks. ICE's $2 billion commitment did not create that role; it ratified it. The institutional plumbing is now in place. Whether Polymarket can sustain the role over the next decade depends on liquidity, regulatory stability, and whether the 12-basis-point alpha that the 5% divergence cases generate compresses as more sophisticated capital crowds in. Newsorga will keep tracking it.