skip to content

Research • June 04, 2026 • 10 mins

Strategy Sold Bitcoin in May. Polymarket Says It Didn’t.

When prediction market resolution diverges from what actually happened, the product stops pricing events and starts pricing how the platform will read its own rules.

This alert was originally sent directly to clients of Galaxy Trading and Galaxy Asset Management on June 3, 2026. Trade or invest with Galaxy to receive the most timely research directly in your inbox.

Strategy (formerly MicroStrategy) has sold BTC for the first time since 2022. The trade was rounding-error small. The fight over how to resolve a related $301m-volume Polymarket contract is not.

A routine corporate disclosure has turned into the highest-stakes live test of the prediction market’s resolution stack since the $237m Zelenskyy-suit market last year.

f tweet1

What happened

Polymarket offers a recurring “MicroStrategy sells any bitcoin by ___?” market. This is a set of time-stamped contracts, each resolving YES if Strategy sells any BTC by 11:59 PM ET on the date in its title, and NO otherwise. The rules name “information from MSTR and onchain data” as the primary resolution source, with a consensus of credible reporting as backup.

Until Monday, the answer was consistently NO; Michael Saylor’s company bought and held. Hence, these contracts were uncontroversial.

og rules

But on June 1, Strategy filed an 8-K disclosing the sale of 32 BTC between May 26 and May 31. The sale generated ~$2.5m at an average net price of $77,135, with proceeds earmarked for preferred stock distributions. It was the company’s first disclosed disposal since December 2022, and it dented the “Saylor never sells” narrative, but the size was trivial: 0.0038% of an 843,706 BTC treasury.

8k mstr

The trade clearly closed within the window of the market ("by May 31”).

When this news was revealed to the market, YES odds on Polymarket spiked from 10% to ~80%. One market participant bought ~700,000 YES shares around 76c apiece. He viewed this as a free arbitrage opportunity, as he described in a detailed X thread. (Each share is supposed to pay out $1 in stablecoins if the prediction comes true, and zero if not.)

Then, at 1:00 PM ET, Polymarket posted “additional context.”

"No information from MSTR, on-chain data, or consensus of credible reporting confirmed that MicroStrategy sold Bitcoin within the market's timeframe,” the team said. “Confirmation achieved outside of the market's time frame does not qualify."

Within seconds, YES shares plummeted below 1c.

1w graph updated
disputes updated

Following Polymarket’s usual process, someone posted a ~$750 bond and proposed the contract’s resolution: NO. It was disputed, by someone posting the same amount.

The outcome was proposed again: NO. It was disputed a second time.

The market then went to final review, and, after 48 hours, resolved NO for the third and last time.

As of 2:30 PM ET June 3, the top 10 holders on each side held ~120 million shares.

top holders new

Top holders of the MSTR to Sell BTC by May 31 contract on Polymarket

jun3 vol

Wednesday afternoon, the May 31 market priced NO at ~99.7c despite a confirmed, on-the-record BTC sale within the time frame posted.

The rules vs. the clarification

The original resolution text is event-based and explicit: the market resolves Yes if Strategy “sells any of its Bitcoin” by the deadline. Nothing in it requires the sale to be announced by then.

This is the crux of the dispute. The YES camp argued the rules name MSTR’s own information as the primary source, and MSTR’s own filing lists the 32 BTC as sold “during [the] period May 26 to May 31.” The NO camp argued that absent public confirmation before June 1, there was nothing to resolve on by the deadline.

In other words, does the event date govern, or the confirmation date? The original rules read event-based; the subsequent bulletin reads confirmation-based. Traders treated the gap as a retroactive rule change, which, in our view as market analysts, it was. (Note that this is not a legal opinion.)

This is not a random edge case

The structural read is doing more damage than the specific outcome. Top YES holders are publicly accusing the platform of waiting for the 8-K to print so sharp money could offload YES onto news-buyers before the market resolved NO against them. Domer, one of the top prediction market traders (he was featured in a 60 Minutes episode on Polymarket) and Adam Back, a cryptographer and prominent Bitcoiner, piled on to the discourse. The framing across crypto Twitter is that you also needed to bet Polymarket wouldn’t change the rules on the win.

domer1

Post from Domer on X.

Resolution of Polymarket contracts flows through the UMA protocol’s optimistic oracle service, and repeated disputes escalate to UMA’s Data Verification Mechanism. This is a token-weighted vote that clears in a 48- to 96-hour window. Analyst Eric Conner has called it an “awful oracle system.” A Wall Street Journal investigation published in May found that in most disputed Polymarket markets, more than half of UMA votes came from the 10 largest wallets, at least 60% of active voters could be linked to live Polymarket accounts, and roughly one in five disputes had a voter with a financial stake in the contract they were ruling on. Polymarket has logged 1,150+ disputed markets in 2026 (already past its full-year 2025 total).

Notably, the June 30 and Dec. 31 contracts in the same series resolved to YES at ~99.9% without dispute. The controversy is entirely about the deadline that happened to sit on the wrong side of the filing.

Our take

At Galaxy Research, we continue to view prediction markets as powerful tools for information discovery and hedging. (As announced Tuesday, our colleagues have set up an over-the-counter prediction markets trading desk.) Polymarket in particular demonstrated the informational value of these markets in the 2024 presidential election, putting the pollsters and legacy media to shame.

But the dispute over the Strategy contract shows the critical importance of resolution integrity for these markets to serve their intended purpose.

Everyone who bought YES predicted the future correctly, and the market just told them they were wrong. That is a failure.

Strip away the oracle mechanics and one fact survives: Strategy sold Bitcoin before May 31. Everyone who bought YES predicted the future correctly, and the market just told them they were wrong. That is a failure.

A prediction market exists to price what will happen. When the resolution diverges from what actually happened, the product stops pricing events and starts pricing how the platform will read its own rules after the fact. Those are not the same game, and the second one is worthless.

MSTR’s filing confirms the sale landed in-window. The NO resolution survives only an unwritten requirement that confirmation be public by the deadline, asserted after sized money was already wagered. There is a good-faith version of that argument (a market can’t resolve on information that didn’t exist at cutoff), but it wasn’t in the original rules either, so it explains the outcome without excusing it.

Polymarket owns the clarification, though the mechanism is subtler than an outright override. By its own rules Polymarket is non-custodial and cannot reverse a resolution once UMA finalizes it (it has done so in the 2024 Barron Trump/DJT market, but only via refunds). Clarifications are officially framed as non-binding “additional context” guidance that “should be considered” by UMA voters, with the final say resting with UMA tokenholders.

In practice, that line is largely cosmetic: UMA voters treat Polymarket clarifications as binding market rules, have never overturned one, and lean on precedent plus literal market language over public sentiment. So Polymarket doesn’t need to override the oracle; for all intents and purposes it tells the oracle how to read the rules and the oracle follows. Add “precedent” that lives in community archives and Discord threads most traders have never seen, and the effective rules become whatever the clarification says. This is the standard operating model for disputed markets, not an edge case, and in multiple high-profile cases affected traders have cried foul.

For a venue that sells itself as “the most accurate thing we have as mankind right now” (in the words of founder and CEO Shayne Coplan), that is the existential problem: resolving against reality. The fixes are unprofitable to whoever benefits from the current setup. Resolution criteria should be locked at listing, with each market explicitly designated event-based or confirmation-based and a defined evidentiary standard. No clarification should be permitted to change that standard after bets are placed. High-dollar disputes should be walled off from token-weighted voting, or at minimum recuse any voter holding a position in the market they’re adjudicating.

Where an outcome is objectively verifiable (a filing, an onchain settlement, a price feed), resolution should be deterministic, not put to a vote. None of this is novel; it’s how every regulated settlement system already works. It cuts against the permissionless, token-governed ethos that made UMA cheap and fast, which is why it hasn’t happened.

Which raises the most uncomfortable question for Polymarket specifically: how would any of this survive being CFTC-regulated? Polymarket is no longer just an offshore crypto venue. Last July, it acquired the licensed exchange and clearinghouse QCX for $112m, which now operates as Polymarket US, separate from the flagship onchain global exchange. Then, in November, Polymarket said it received an Amended Order of Designation from the CFTC, allowing it to operate as an intermediated, federally regulated Designated Contract Market. This approval “positions Polymarket for U.S. return under a fully regulated exchange structure,” the company said at the time. A DCM carries settlement-integrity and conflict-of-interest obligations under the Commodity Exchange Act that a token-weighted oracle (where adjudicators can hold the very positions they vote to settle) does not come close to meeting.

You can’t run a “fully regulated exchange” on one hand and a clarifications-on-vibes oracle with the other and expect the line between them to hold. Either the product resolves through a genuinely different, compliant mechanism (see centralized archrival Kalshi), in which case Polymarket is conceding its crypto-native resolution layer isn’t fit for purpose, or the oracle model is on a collision course with the rulebook the company just signed.

A properly functioning, fairly adjudicated decentralized oracle service is a worthy long-term goal, but it is hard to argue that this is it.

The current standard is one that a venue like Kalshi clears by design: CFTC-regulated, centralized resolution that would never route a disclosure-timing dispute through a tokenholder vote.

Ambiguity in resolution criteria is a liability that compounds with volume, and a “truth-seeking” oracle steerable by the wallets with the most at stake is not truth-seeking.

Whatever precedent this debacle sets for corporate-disclosure-timing markets will outlast the 32 BTC that started it.

You are leaving Galaxy.com

You are leaving the Galaxy website and being directed to an external third-party website that we think might be of interest to you. Third-party websites are not under the control of Galaxy, and Galaxy is not responsible for the accuracy or completeness of the contents or the proper operation of any linked site. Please note the security and privacy policies on third-party websites differ from Galaxy policies, please read third-party privacy and security policies closely. If you do not wish to continue to the third-party site, click “Cancel”. The inclusion of any linked website does not imply Galaxy’s endorsement or adoption of the statements therein and is only provided for your convenience.