Same Question, Different Contract

1. Introduction: the question is not the contract

Event contracts are no longer a curiosity. According to the Congressional Research Service, designated contract markets (DCMs) listed an average of five new event contracts a year between 2006 and 2020. They listed about 1,600 in 2025.¹ Industry trackers put Kalshi at roughly three-quarters of tracked prediction-market volume in the third quarter of 2026. Both Kalshi and Polymarket US reported record monthly volume in October, ahead of the 3 November midterms.² Institutions now take these prices in as data, quote them in research and, increasingly, trade them.

The usual way to read these markets is by question. Will the Fed hike in December? Which party will control the Senate? Will a given leader leave office by March? Venues encourage this. Their titles are short and look alike across platforms, and aggregators line them up side by side.

But a question doesn't pay anyone. A contract does. Each venue writes its own rules for what counts as the event and which source decides it. The rules also set when the observation is taken, what happens if something unexpected occurs, and who has the final word. On an ordinary day, the rules on different venues lead to the same result. This paper is about the days when they don't, and about how to tell in advance which days those could be.

We draw on three bodies of evidence:

1. Matched cases. In 2025–2026, Kalshi, Polymarket International and, where relevant, other venues listed contracts on the same real-world questions. We read each venue's rules, contract filings and settlement records for these cases.
2. A resolution dataset. It covers 96,414 resolved markets from the 2,100 highest-volume closed Polymarket International events with end dates on or after 1 October 2025. For each market it holds the recorded resolution history: proposals, disputes and final outcome.
3. Settlement timestamps for nine FOMC rate decisions on both Kalshi and Polymarket International.

Section 2 starts with the clearest case of divergence we found. Section 3 sorts divergence into four families. Section 4 measures how often Polymarket resolutions are contested, and what that costs in time. Section 5 compares the three adjudication models. Section 6 turns the evidence into a checklist to use before trading. Our take, the methodology and the limitations follow.

2. One event, four settlements

On 28 February 2026, Ayatollah Ali Khamenei was killed in US–Israeli strikes on Tehran. Iranian state media confirmed his death either late that day or early on 1 March; accounts differ.³ Several venues had listed contracts on whether he would leave office. They did not settle the same way.

Kalshi. Kalshi listed its "Ali Khamenei out as Supreme Leader?" markets under its WLEADEROUT contract terms, which it self-certified to the CFTC in May 2025.⁴ The terms pay YES if the named leader leaves office before the stated date, with one exception. If the leader leaves "solely because they have died", the market settles at "the last traded price (prior to the death)". If that price is missing or unfair, an Outcome Review Committee decides on a fair allocation instead.

Kalshi halted trading at about 2:59 p.m. ET on 28 February and closed the markets at 10:06 p.m. ET.⁵ Its public market records show the YES side of the contract dated "before March 1, 2026" settling at $0.02. The contracts dated before 1 April, 1 July and 1 September settled at $0.29, $0.45 and $0.51.⁶ Kalshi said it reimbursed fees and net losses so that no trader ended up net negative.⁷ A proposed class action, Risch v. KalshiEX LLC, was filed on 5 March in the Central District of California. It alleges the carve-out was not adequately disclosed.⁸ On 2 March, Kalshi filed a rulebook amendment with the CFTC. It formally codifies how Kalshi settles contracts whose primary subject dies.⁹

Polymarket International. Polymarket listed "Khamenei out as Supreme Leader of Iran by February 28?". It resolved YES if Khamenei was removed from power before 11:59 p.m. ET on 28 February, with a consensus of credible reporting as the resolution source. Its text has no death carve-out, so death counts as losing the position. The dispute was about timing instead: had credible reporting reached a consensus before the deadline? The proposed YES outcome was disputed twice and went to a vote of UMA token holders. The market resolved YES at 01:05 UTC on 4 March (the evening of 3 March in New York). It had traded $131 million.¹⁰

Gemini Predictions. Gemini, a CFTC-registered venue, framed the question the other way round: would Khamenei remain Supreme Leader through a given date? Its listed "remain through" contracts started with 31 March 2026, and they resolved NO.⁹

PredictIt. PredictIt operates under a CFTC no-action letter, not as a DCM. It asked whether Khamenei would "resign from, be removed from, leave, or otherwise vacate" office by 1 May. Under that wording, death counts as vacating office.⁹

The divergence was not only about the size of each payout. Read literally, the contracts could pay both sides of a supposed hedge. Take someone who held Polymarket YES and Kalshi NO as an offsetting pair. They collected $1.00 on the first leg and $0.98 on the second. The mirror-image pair lost almost everything on both legs, before Kalshi's discretionary reimbursement. Neither venue broke its own rules. Each paid exactly what its contract said. The difference came from drafting choices made months earlier, when the contracts were listed. Kalshi says it does not list markets that settle on death, a stance in line with the long-standing US scrutiny of event contracts involving assassination and war.¹

3. Four places where same-question contracts diverge

The Khamenei case is extreme, but its cause is ordinary: contracts that read alike, written to different rules. Across the cases we reviewed, divergence came from four families of terms.

3.1 Trigger: what counts as "happened"

Every contract chooses, explicitly or not, the kind of evidence that triggers a payout. Four types recur:

• Event-based: the thing occurred by the deadline, whenever it became known.
• Observation-based: a state holds at a fixed observation time.
• Announcement-based: a named body announces it.
• Confirmation-based: named or credible sources report it as confirmed before the deadline.

These sound like fine distinctions. They aren't, as three recent cases show.

The 14 February 2026 shutdown. Funding for the Department of Homeland Security lapsed and, according to the Associated Press, a partial government shutdown began on Saturday 14 February.¹¹ Kalshi's contract asked whether the federal government was "at least partially shut down due to a lapse of appropriations at 10:00 AM ET" that day. It settled YES at 15:33 UTC on 14 February. Polymarket's "Government shutdown on Saturday?" required the Office of Personnel Management to announce a new shutdown by 11:59 p.m. ET. It excluded any continuation of the partial shutdown that began on 31 January. It resolved NO.¹² The lapse was the same one, and so was the date. Kalshi's contract was triggered by an observation; Polymarket's needed an announcement from a named source.

Strategy's bitcoin sale. Polymarket's "MicroStrategy sells any Bitcoin by ___?" contracts resolved YES if the company sold any of its bitcoin by the date in the title. On 1 June 2026, Strategy disclosed in an 8-K filing that it had sold 32 BTC between 26 and 31 May.¹³ Polymarket then posted "additional context". It said confirmation that arrived after the market's timeframe did not qualify. The May contract had traded $376 million. It was disputed twice and resolved NO. The June 30 and December 31 contracts resolved YES on 1 June without dispute.¹⁰ So one sale, made in May, counted for the June contract but not for the May one. As Galaxy Research observed, the original text read as event-based, while the clarification read as confirmation-based.¹³ Kalshi's comparable contract family is filed under a template named EVENTCONFIRMED. Under it, the outcome depends on whether named news organizations report the event as confirmed before the date. That makes it confirmation-based from the start.¹⁴

The Fed chair nomination. Both venues paid on Kevin Warsh's nomination, but each waited for different evidence. Polymarket's market defined a formal nomination as "the submission of a nomination message to the U.S. Senate". Kalshi's resolved on the first person "formally nominated by the President". DeFi Rate reported that Kalshi waited until the Senate had formally received the nomination.¹⁵ Polymarket's market was disputed twice before resolving YES at 19:25 UTC on 4 March. Kalshi settled at 14:52 UTC on 5 March.¹⁰⁶ The outcome was the same, and the payments arrived about 19 hours apart.

The trigger also sets how long a contract can stay open after the outcome is obvious. WIRED reported that Polymarket's market on Ethiopia's 2026 election resolved only after the prime minister was formally appointed and sworn in. It stayed open for 123 days after the vote.¹⁶

3.2 Source and fixing: where and when the world is measured

Predict Here's public demo records a pair of Bitcoin contracts for 14 October 2026.¹⁷

• Kalshi paid YES if the simple average of CF Benchmarks' Bitcoin Real-Time Index (BRTI) over the 60 seconds before 5:00 p.m. EDT was above $85,499.99.
• Polymarket International paid YES if the 1-minute BTC/USDT candle on Binance closing at 12:00 ET was above $86,000. Its rules note that only Binance BTC/USDT counts, not other exchanges or trading pairs.

The two contracts differ in five ways at once:

• the source: an index built from several venues against a single exchange;
• the method: a 60-second average against one candle close;
• the fixing time: five hours apart;
• the quote currency: US dollars against USDT;
• the threshold: $500 apart.

There is also a governance difference. CF Benchmarks is authorized by the UK Financial Conduct Authority as a benchmark administrator. It describes the BRTI as a registered benchmark under the UK Benchmarks Regulation.¹⁸ A one-minute candle on a single exchange has no comparable oversight.

On the day, the difference didn't matter. Kalshi's records show a BRTI fixing of $81,711.75, so both contracts settled NO.⁶¹⁰ On most days, basis like this stays latent. It turns real when it matters most. If both fixings had read between $85,500 and $86,000, the contracts would have paid opposite sides. Divergence also becomes possible whenever bitcoin crosses either threshold between noon and 5:00 p.m., or when USDT trades away from the dollar. A pair of contracts that agree on 95% of days still disagree on the other 5%, and those tend to be the volatile days.

The source itself can change. Kalshi's FOMC contract terms let the exchange designate a new source agency, and change the related specifications, if an event affects the reliability of the original source.¹⁹ Polymarket's up-or-down crypto markets settle on a Chainlink time-weighted average price. Its other markets settle through the oracle process described in Section 5.²⁰

3.3 Carve-outs and fallbacks: when a binary contract stops being binary

A binary contract pays $1 or $0 only until a carve-out applies. The ones we found most often:

• Death and incapacity.
  • Kalshi's leader-out terms settle at the last traded price before death, as above. Its 2 March 2026 rulebook amendment extends the same approach to contracts whose primary subject dies before expiration. It also lets Kalshi halt trading when it believes a death has occurred or is imminent.⁹
  • Polymarket US's rulebook (dated 30 September 2026) contains Rule 10.5. It applies when a contract's primary subject dies or is incapacitated, so the qualifying outcome can no longer occur or be objectively determined. Affected contracts then settle at the last traded price before the event. If trading was affected, the price used is the last one before the circumstances became known, based on the earliest objective evidence.²¹
  • Polymarket International's market text contains no general clause of this kind.
• Cancelled or rescheduled events.
  • Kalshi's FOMC contracts: if a scheduled meeting is cancelled and does not occur on its date, "Fed maintains rate" resolves YES.
  • Polymarket's FOMC markets: if no statement is released by the end date of the next scheduled meeting, the market resolves to "No change". A change that falls between the listed brackets is rounded up to the nearest 25 basis points.¹⁷
• Equal splits. Polymarket can resolve a market 50/50, with each share paying $0.50.²⁰ In our sample, 19,250 of 96,414 resolved markets (20.0%) carried an "equal split" fallback flag in Polymarket's API. 1,137 markets actually resolved 50/50: 1.18% of markets, but only 0.08% of volume. Almost all were game-level esports markets (a fourth game in a best-of-five series, for example) and fight markets. Markets like these resolve 50/50 when the specific game or bout does not take place.
• Committee allocations. Kalshi's Outcome Review Committee can decide a "fair allocation" when its other rules do not produce a fair settlement value.⁴

Carve-outs are not defects. Most protect market integrity or comply with regulation. But each one adds a state to the payout. A risk system that models a binary contract as having two states will mis-measure it.

3.4 Adjudication and timing: when the money arrives

The FOMC is the cleanest test of settlement speed. The outcome is unambiguous. It is published at a known time, 2:00 p.m. ET, by a source both venues name: the Federal Reserve. Neither venue had to interpret anything.

Kalshi's FOMC terms stop trading at 1:55 p.m. ET and set expiration at 2:05 p.m. ET after the decision is released. Kalshi then settles the contracts directly.¹⁹ Polymarket's markets resolve when someone proposes an outcome and a challenge window passes. Polymarket's documentation describes that window as two hours.²⁰ Since July 2026, Polymarket's Fed markets have settled about two hours after the statement. When a proposal is disputed, the clock restarts. One October 2025 market was disputed twice and settled 33.6 hours after the statement, although its outcome, no hike, was never in doubt. In April 2026, all four of Polymarket's Fed markets were disputed once and still settled within 3.4 hours.

For an institution, time-to-cash is part of the price. It decides how quickly capital can be recycled and when a rolled hedge is truly flat. It also shapes how a position is marked during the window, while the outcome is public but the cash is not.

4. How often are resolutions contested?

Polymarket International is the only major venue whose resolution history is public at the level of individual markets, because it runs on an onchain oracle. Kalshi and Polymarket US publish final outcomes and outcome-review notices but no dispute log. That transparency lets us measure what the other venues keep internal. It should not be read as evidence that disputes happen only on Polymarket.

Data. We retrieved, through Polymarket's public Gamma API, the 2,100 highest-volume closed events with an end date on or after 1 October 2025. That is the maximum the endpoint returns for this query. The events contain 96,605 markets, 96,414 of them resolved, with $37.6 billion of reported volume. For each market we counted the "disputed" entries in its recorded resolution history. Full definitions are in the methodology section.

The count is small. The money is not. About one dollar in five traded in the non-sports markets of our sample went through at least one formal dispute.

The concentration grows with the size of the market.

The pattern follows from how the process is priced. The bond to propose or dispute an outcome is fixed and small. Polymarket's API recorded a $500 bond on the largest disputed markets in our sample, and Polymarket's help documentation describes a typical counter-bond of $750.²⁰ The value riding on an interpretation grows with the market. The Strategy market's bond came to about 0.0001% of the $376 million it traded. The larger the market, the more it pays to contest a reading.

Disputed value is concentrated in a small number of markets. The ten largest account for 43.9% of disputed volume, and the 25 largest for 66.9%.

Several of these questions are the kind institutions use as macro hedges or as policy signals: ceasefires, government funding, central-bank decisions, leadership changes. Those are also the questions whose wording is hardest to make airtight.

The cost in time. We measured how long markets took to resolve after their stated deadline. We used non-sports markets with a defined end date that resolved within 120 days of it.

What a dispute is, and is not. A dispute is a formal challenge to a proposed outcome, backed by a bond. It is not proof that a market resolved wrongly. Some disputes challenge proposals made too early. Others end with the original proposal upheld. Our data records that a dispute occurred, not which outcome was proposed. What disputes do measure reliably is interpretive friction: how much value had its settlement depend on adjudication beyond the first proposal, and how much time that took.

Other researchers and reporters have looked at who adjudicates. In May 2026, the Wall Street Journal reported three findings for most disputed Polymarket markets. The ten largest wallets cast more than half of UMA votes. At least 60% of active voters could be linked to Polymarket trading accounts. And about one dispute in five involved a voter with a financial stake in the outcome.²² Bloomberg reported that nine wallets held about half of UMA voting power over three years.²³ Galaxy Research counted more than 1,150 disputed Polymarket markets in 2026 by early June, more than in all of 2025. That count covers all markets, not only the high-volume events in our sample.¹³ Polymarket's help center says it expects to roll out a new reward and oracle-resolution system later this year.²⁴

5. Three adjudication models

Everything above depends on who decides. Today's largest venues use three models.

Each model puts discretion somewhere. Regulated venues offer finality and a decision-maker who can be held to account. They also give the exchange broad latitude: it can modify specifications, substitute sources and make final determinations without an external appeal on the trade itself. The offshore model offers an open process and a public record. In exchange, its final word comes from a token-weighted vote whose concentration and conflicts have been documented. Neither model is free of discretion. They give it to different people.

The word "Polymarket" now covers two different adjudication regimes. Polymarket International resolves through UMA. Polymarket US resolves through exchange officers under a DCM rulebook. A data pipeline that treats "Polymarket" as one source is mapping two kinds of contract.

Regulation is pushing in one direction. In June 2026, the CFTC proposed amendments to its Rule 40.11, which governs event contracts that touch enumerated activities. The proposal lists "settlement integrity deficits", such as a lack of clear and objective resolution criteria, among the factors that could make a contract contrary to the public interest.¹ Regulated venues already write more explicit carve-outs and evidentiary standards: Kalshi's death-rule amendment and Polymarket US's Rule 10.5 are recent examples. Offshore, text-based contracts will keep relying on clarifications and votes. In the near term, the drafting gap between "same question" contracts on regulated and offshore venues is likely to widen, not narrow.

6. Before treating two contracts as the same exposure

The cases above reduce to eight questions. They take minutes per contract to answer, and much longer to answer after settlement.

7. Our take

What trades is the contract, not the event. Pricing the event without pricing the contract leaves out part of the risk. Prices on two venues can be compared only after their terms have been.

Most cross-venue spreads are basis, not mispricing. The Khamenei case shows the tail: two contracts on one fact paid $1.00 and $0.02. The FOMC data shows the everyday cost: the same correct answer, paid minutes apart on one venue and hours apart on the other. Neither shows up in a side-by-side table of headline prices.

Resolution risk is weighted by value and sits where institutions care most. Disputes touched 0.36% of the Polymarket markets we examined. They touched one non-sports dollar in five, and 41% of non-sports markets above $50 million. Macro, policy and geopolitical contracts are exactly where institutional demand for event hedges is growing.

Regulated and offshore venues are drifting apart on drafting. Expect regulated venues to keep codifying carve-outs and evidentiary standards in response to the CFTC's proposal and to litigation. Expect offshore venues to keep resolving ambiguity after the fact. Until there is a shared definition of the event, the burden of reconciling contracts falls on the people trading them.

What would help. Five changes would make settlement risk measurable rather than anecdotal:

• each contract's trigger type and evidentiary standard classified at listing;
• carve-outs published as structured fields, not only in prose;
• public, timestamped versions of rules and clarifications;
• settlement-time statistics, published by each venue;
• for oracle-based venues, recusal of voters who hold positions.

Until then, someone has to read the terms and keep a record of what they read and when.

Predict Here records the terms each contract actually pays on: threshold, source, fixing time, carve-outs and rule versions. It keeps the source and observation time with every figure. Our contract comparison checklist and our side-by-side reading of the midterm control contracts apply the approach in this paper.

Methodology, data and limitations

Polymarket resolution dataset.

• Retrieval. We used the public Gamma API endpoint /events with closed=true, ordered by volume (descending) and filtered to events with an end date on or after 1 October 2025. We paginated to the endpoint's limit: 2,100 events, retrieved on 31 October 2026 between about 07:56 and 08:00 UTC.
• Size. The events contain 96,605 markets, 96,414 of them resolved. Volume is the volume field Polymarket reports for each market.
• Disputes. A market is "disputed" if its umaResolutionStatuses history contains at least one disputed entry; we also count how many it contains.
• 50/50 resolutions. A market resolved 50/50 if its final outcomePrices are [0.5, 0.5].
• Categories. We assigned each event to the first tag that matched, in this order: Sports, Politics, Crypto, Economy, Finance, Geopolitics, Culture, Tech, Business, World, Elections; anything else is Other. Because events carry several tags, category boundaries are approximate. We therefore report the sports versus non-sports split as the main result.
• Delays. For Exhibit 7, delay is the market's recorded close time minus its end date. We kept non-sports markets with a set end date and a delay between −1 hour and 120 days.

Kalshi data. We used Kalshi's public Trade API v2: the /markets and /historical/markets endpoints for settlement values, results and timestamps, and /series for series metadata. We read contract terms from the PDF filings that Kalshi links from each series, all retrieved on 31 October 2026.

Matched cases. We chose cases purposively: questions listed on more than one venue, where the terms and outcomes are documented. They show the mechanisms; they are not a random sample and do not support rates of divergence. Some details come from third-party reporting, cited in the notes: DeFi Rate, Decrypt, Bloomberg Law, Galaxy Research, the Associated Press and WIRED.

Limitations.

• The Polymarket sample is restricted to high-volume events. Lower-volume markets may have different dispute rates.
• Reported volume can include wash trading or activity farming, which has been documented on Polymarket.¹⁶
• Volume on Kalshi is counted in contracts and volume on Polymarket in dollars, so we do not compare them.
• A dispute does not establish that an outcome was wrong.
• End-date semantics differ across Polymarket markets.
• Kalshi and Polymarket US do not publish dispute logs, so Section 4 cannot be repeated for them.

Reproducibility. The scripts and derived data behind every exhibit are available on request.

Notes and references

1. Congressional Research Service, "CFTC Issues Proposed Rule Regarding Prediction Markets," Legal Sidebar LSB11441 (2026). https://www.everycrsreport.com/reports/LSB11441.html ↩
2. Prediction News, "Kalshi holds 76% of tracked volume as Polymarket US grows 66%"; DeFi Rate prediction-market volume tracker; Coinpaper, "Kalshi hits $65 billion in October trading as Polymarket US breaks records" (October 2026). Venues and trackers measure volume differently; figures are indicative. ↩
3. Wikipedia, "Assassination of Ali Khamenei"; The New Arab, "What we know about the US-Israeli killing of Iran's Supreme Leader Ali Khamenei"; ABC News / Good Morning America report, 28 February 2026. ↩
4. KalshiEX LLC, CFTC Regulation 40.2(a) notification for the "Will <leader> leave office before <date>?" contract (WLEADEROUT), 15 May 2025. https://assets.kalshi.com/regulatory/product-certifications/WLEADEROUT.pdf ↩
5. DeFi Rate, "Kalshi Halts Khamenei Market, Polymarket's Contract Enters Second Dispute" (March 2026). https://defirate.com/news/kalshi-halts-khamenei-market-polymarkets-contract-enters-second-dispute/ ↩
6. Kalshi Trade API v2, /historical/markets and /markets (series KXKHAMENEIOUT, KXGOVTSHUTDOWN, KXFEDCHAIRNOM, KXFEDDECISION, KXBTCD), retrieved 31 October 2026. ↩
7. Decrypt, "Kalshi Sued Over Refusing to Pay Out Prediction Market After Iran Leader's Death" (March 2026). https://decrypt.co/360293/kalshi-sued-refusing-pay-out-prediction-market-iran-leader-death ↩
8. Bloomberg Law, "Kalshi Sued Over Death Carveout in Iran Leader Prediction Market" (March 2026); Risch v. KalshiEX LLC, No. 2:26-cv-02390 (C.D. Cal., filed 5 March 2026). ↩
9. DeFi Rate, "Kalshi Codifies Death Settlement Rule in CFTC Filing Amid Iran Market Backlash" (March 2026). https://defirate.com/news/kalshi-codifies-death-settlement-rule-in-cftc-filing-amid-backlash-over-iran-market/ ↩
10. Polymarket Gamma API, /events and /markets (market rules, umaResolutionStatuses, outcomePrices, closedTime), retrieved 31 October 2026. ↩
11. Associated Press via PBS NewsHour, "Partial government shutdown drags on amid immigration oversight battle" (15 February 2026). https://www.pbs.org/newshour/politics/partial-government-shutdown-drags-on-amid-immigration-oversight-battle ↩
12. Kalshi market KXGOVTSHUTDOWN-26FEB14 (rules and settlement record); Polymarket event "Government shutdown on Saturday?" (rules and resolution record). Both retrieved 31 October 2026. ↩
13. Galaxy Research, "Strategy Sold Bitcoin in May. Polymarket Says It Didn't." (4 June 2026). https://www.galaxy.com/insights/research/strategy-bitcoin-sale-polymarket-resolution-dispute-may-2026 ↩
14. KalshiEX LLC, CFTC Regulation 40.2(a) notification for the "Will <event claim> get confirmed?" contract (EVENTCONFIRMED). https://assets.kalshi.com/regulatory/product-certifications/EVENTCONFIRMED.pdf ↩
15. DeFi Rate, "Kevin Warsh Fed Chair Nomination Official, Settling $807M Prediction Market" (March 2026). ↩
16. WIRED, "Over $600 Million in Polymarket Election Trades Took Place After Races Ended" (October 2026). https://www.wired.com/story/over-dollar600-million-traded-on-polymarket-election-contracts-after-races-ended/ ↩
17. Predict Here public demo, recorded Bitcoin and Fed contract terms (captured October 2026). https://predicthere.com/terminal-demo ↩
18. CF Benchmarks, BRTI index page and FCA authorization notice. https://www.cfbenchmarks.com/indices/BRTI ↩
19. KalshiEX LLC, FEDDECISION contract terms. https://assets.kalshi.com/regulatory/product-certifications/FEDDECISION.pdf ↩
20. Polymarket documentation, "How Resolution Works." https://docs.polymarket.com/concepts/resolution ↩
21. QCX LLC d/b/a Polymarket US, Rulebook (version dated 30 September 2026), Chapter 10. https://polymarketexchange.com/files/legal/latest/rulebook ↩
22. Wall Street Journal investigation of UMA voting on Polymarket disputes (May 2026), as summarized by Finance Magnates and Galaxy Research. Figures are as reported. We have not reproduced the Journal's analysis. ↩
23. Bloomberg reporting on UMA voting concentration (2026), as summarized by KuCoin News. ↩
24. Polymarket Help Center, "Holding Rewards." https://docs.polymarket.com/polymarket-learn/trading/holding-rewards ↩

Disclosures

This paper is general research published for information only. It is not investment, legal or tax advice. It does not recommend any position, venue or strategy, and is not tailored to any reader's circumstances. Predict Here is operated by Odyssea Research LLC. Predict Here does not list markets, hold client funds, take the other side of trades, or execute or route orders.

The venue data cited here comes from public interfaces and documents, each identified by its source and retrieval date. Statements about venues' rules describe the documents cited, as of the dates given. Rules change: always read a venue's current terms before acting.

Persons associated with Predict Here may hold positions in event contracts, including on venues discussed in this report.

Errors will be corrected openly, and each correction will be dated.