The last trade: When reality becomes the underlying asset

There is a question that financial regulators, technology enthusiasts, and prediction market advocates have carefully avoided asking.

Not whether prediction markets are accurate. The evidence increasingly suggests they are — often more accurate than polls, expert panels, and conventional forecasting models. Not whether they should be regulated. That debate is already underway across multiple jurisdictions. Not even whether insider trading is occurring on them. The evidence from the Iran war suggests it already has.

The question being avoided is simpler and more fundamental than any of these: should anyone be permitted to place a financial bet on catastrophic events that affect the survival, security, or wellbeing of millions of people?

That question is not rhetorical. It is the precise logical endpoint of the category of instrument that prediction markets, enabled by cryptocurrency infrastructure, have now become. The evidence from the Iran war shows it is no longer hypothetical. And the world will answer it — probably in the aftermath of something that cannot be ignored, and probably with considerable anger at the fact that it was allowed to become a question at all.

Just because the technology makes something possible does not mean it should exist. That principle is older than financial markets. Prediction markets on consequential events are about to learn it.

Part I: What We Already Know

Before the systemic argument, the facts — separated cleanly from inference.

In May 2026, CBS News and 60 Minutes published findings from data analytics firm Bubblemaps documenting nine linked anonymous accounts that accumulated more than $2.4 million betting on US military actions against Iran. These accounts placed winning bets on the specific dates of first strikes, the removal of Iran's supreme leader, and the announcement of a ceasefire. Across more than 80 bets, they maintained a 98% win rate — with many bets placed when the odds of winning were low. Bubblemaps co-founder Nicolas Vaiman was direct: "Luck alone cannot explain those numbers."

Separately, the Anti-Corruption Data Collective analysed long-shot wagers on military markets and found indications of what they called "systemic insider trading" in military-event markets. Bettors on unlikely outcomes were winning more than they lost at a rate that statistical analysis could not attribute to skill or chance.

Then there is the case of US Army Master Sergeant Gannon Ken Van Dyke, indicted on federal charges alleging he used classified information about a special operations mission to capture former Venezuelan President Nicolás Maduro to place bets totalling roughly $34,000, netting more than $400,000. He allegedly placed several bets the day before the raid, withdrew his profits immediately, and attempted to delete his betting account. Van Dyke has pleaded not guilty.

On March 23, more than $800 million was staked on dropping oil prices in a fifteen-minute window before President Trump posted that the US and Iran had engaged in productive ceasefire negotiations. The trade, if timed on non-public information, potentially netted tens of millions of dollars. Federal investigators are examining those trades. No charges have been filed.

And a journalist covering the conflict received threats after his accurate reporting voided a market prediction. Traders who lost money sent messages including: "You're going to make us lose $900,000. And we'll invest even more than that to finish you." The messages included details about his family members and their movements.

The conclusion that flows from these facts is modest, precise, and essentially incontestable: prediction markets tied to military and geopolitical events have already attracted participants seeking to profit from privileged information about those events — and in at least one documented case, a participant with operational access to classified military planning allegedly used that access for financial gain.

That is what we know. Everything that follows builds on that foundation.

Part II: Why Prediction Markets Are Structurally Different

To understand why this matters beyond conventional insider trading, consider what prediction markets actually reward — and how that differs from every prior financial instrument.

In traditional financial markets, a corrupt trader who knows a merger is coming profits from that knowledge. The underlying event exists independently. The trader benefits from information. They cannot cause the merger by placing the bet.

Prediction markets progressively collapse this separation. They reward:

Better information — the trader who knows more accurately what will happen wins. This is the epistemic function prediction markets are celebrated for.

Earlier information — the trader who knows sooner wins more. Still recognisable as information asymmetry, similar in kind to conventional insider trading.

Influence over information — the trader who can shape what others know, by suppressing a story, leaking selectively, or pressuring a journalist, can move markets in their favour. The journalist threats are direct evidence of attempts to exert such pressure. When a reporter loses control of his family's safety because accurate journalism intersected with market positions, the market has ceased being an observational tool. It is exerting financial pressure on the information environment it is supposed to reflect.

Influence over events — the trader who can affect the outcome directly holds the most powerful edge of all. At this point, the financial instrument and the real-world event are no longer separate. They are the same thing viewed from two angles — and one of them pays out anonymously.

Each step moves the market closer to the event itself. The first two describe a useful, occasionally abused, forecasting tool. The third describes a threat to journalism and public information. The fourth describes something with no precedent in financial history: an instrument that creates financial incentives for individuals to influence catastrophic real-world outcomes.

Part III: The Rational Trader Assumption — And Why It Fails

Most defences of prediction markets rest on an implicit assumption: participants are rational actors seeking to maximise financial returns.

Under this view, a trader who attempts to influence an event risks attracting scrutiny, triggering investigations, voiding contracts, or losing platform access. The financial incentive therefore encourages prediction rather than intervention. Enforcement deters manipulation.

That assumption is not unreasonable. It is incomplete.

Financial markets have never operated exclusively with rational participants. History is filled with traders who doubled down on losing positions, concealed losses, committed fraud to avoid insolvency, or took reckless risks in pursuit of recovery. Human beings do not always behave as economists expect.

A participant facing bankruptcy, criminal liability, or personal ruin may not carefully weigh whether a contract could later be cancelled. Someone who believes they have nothing left to lose may view intervention as a final opportunity. The threat of a voided payout is unlikely to deter someone already operating from desperation.

Nor is financial gain always the primary motivation. An ideologically motivated actor — a political extremist, a terrorist sympathiser, a disgruntled operative — may already desire a particular outcome for reasons entirely unrelated to profit. In such cases, a prediction market position does not create the intent. It funds it. The market becomes an additional reward attached to an outcome they already wish to see occur.

The risk compounds once participants have accumulated substantial positions. A trader who has committed significant capital to a market may begin to view intervention not as manipulation but as protection of an investment. Once enough money has been committed, some individuals stop asking whether they should influence an outcome and begin asking whether they can afford not to.

The deeper concern is therefore not that every trader will attempt to influence the events they are betting on. Most will not. Many could not even if they wished to.

The concern is that prediction markets create an additional financial incentive for those who can. And that same incentive structure will only become more concerning as technologies capable of causing outsized harm become increasingly accessible to individuals acting alone.

When the event occurs, the question of whether the contract is eventually honoured becomes largely irrelevant. The war has already happened. The attack has already occurred. No disciplinary proceeding, contract cancellation, or account suspension can reverse the underlying event. The market may punish misconduct after the fact. It cannot guarantee that the incentive did not contribute to the act itself.

Part IV: Coordination Without Communication

Prediction markets introduce a genuinely novel risk that existing legal frameworks have no vocabulary for.

Traditional criminal coordination — conspiracy, solicitation, incitement — requires communication between parties. Agreement. Planning. A shared understanding of purpose. These requirements exist precisely because they create the visibility that makes interdiction and prosecution possible.

Prediction markets can communicate incentives through prices alone.

When a market prices the probability of a political leader's death within a given month, a large position creates a public financial signal visible to every participant on a permissionless platform globally. It communicates, without any direct message to anyone: this outcome has value; here is how much value; anyone capable of influencing it stands to benefit if it occurs.

No meeting took place. No conspiracy was committed in any jurisdiction. No communication occurred between the person holding the position and any person capable of acting on it. The market mechanism performs the coordination that previously required human interaction — at global scale, instantly, anonymously, and automatically.

This is not a claim that prediction markets function as organised instruments of violence. It is a more precise and more difficult claim: prediction markets can create public financial incentives for outcomes without requiring direct coordination between parties who benefit and parties capable of influencing those outcomes.

That structural property is genuinely new. It has no analogue in the history of financial instruments. And it cannot be addressed by insider trading rules, AI surveillance, or blockchain forensics — because no insider trading, in the conventional sense, need occur.

Part V: The Deeper Challenge Is Not Market Integrity — It Is Reality Integrity

This is the insight that most commentary on prediction markets has missed, and it is the one that matters most.

Every debate about prediction market reform has focused on market integrity — whether prices are accurate, whether participants are honest, whether surveillance is adequate, whether enforcement deters misconduct. These are real and legitimate concerns. They are also the wrong level of analysis.

The deeper challenge is reality integrity.

Prediction markets work best when the events being traded are genuinely independent of the act of trading itself. A market on an election outcome functions well when no participant can influence the election by placing a bet. A market on a sports result functions well when no bettor can affect the match.

As prediction markets expand into military operations, political violence, ceasefire negotiations, and catastrophic events, that independence becomes impossible to maintain. The events being traded are precisely the events most susceptible to human influence. They involve decision-makers — generals, officials, intelligence analysts, negotiators — whose choices are not fixed by natural law but made under pressure, under uncertainty, and under conditions that a financial incentive structure can reach.

When the underlying asset is reality itself — specifically, the most consequential and fragile parts of reality — the question of whether the market is functioning with integrity becomes secondary to whether the reality being traded on is still functioning with integrity.

Traditional financial markets trade claims on things that humans produce. Prediction markets on geopolitical events trade claims on things that humans decide. That distinction is not technical. It is civilisational.

The journalist threatened over a war report was not experiencing a market integrity failure. He was experiencing a reality integrity failure — the moment when a financial instrument reached into the physical world and created a monetary incentive for a true account of events not to exist.

This is what prediction markets on consequential events actually are, stripped of the forecasting language and the epistemic theory: instruments that create financial interests in the shape of reality, held by anonymous global participants, with no accountability to the people whose reality is being traded on.

Part VI: What Cannot Be Marketised

Every society draws lines around what can be turned into a market.

Human organs cannot be bought and sold in most jurisdictions — not because the market could not price them, but because allowing them to be priced changes behaviour in ways societies have decided are unacceptable.

Judicial verdicts cannot be purchased. Military commissions cannot be auctioned. Public offices are constrained against direct sale. These prohibitions exist not because markets are incapable of functioning in these domains, but because the act of marketising them corrupts the thing being sold.

The question prediction markets now raise is whether wars, assassinations, and catastrophic events belong in the same category.

The argument for including them is not complicated. Once a financial instrument attaches a monetary value to an outcome, every actor with influence over that outcome becomes a potential target of financial incentives tied to that outcome — whether or not they chose to be, whether or not they are even aware of it.

A general whose operational decisions are being priced on an anonymous global market can be approached, pressured, or simply tempted by positions he did not create and cannot see. A negotiator whose ceasefire terms are being traded has counterparties she has never met and cannot identify. An intelligence official whose assessments move markets has a personal financial ecosystem built on the classified information he handles.

These actors did not consent to having their professional judgements financialised. They cannot opt out. The market priced them without asking.

This is the commodification argument that the prediction market debate has consistently avoided confronting directly. It is not about whether prices are accurate. It is about whether the act of pricing changes the thing being priced — and whether some things are too important to risk that change.

Derivatives before 2008 were celebrated as sophisticated instruments for distributing risk. What they actually distributed, as it turned out, was the incentive to take risks that others would bear the consequences of. The incentive structure was the problem — not the complexity of the instrument, not the failure of any individual actor, but the systematic misalignment between who held the financial interest and who lived with the outcome.

Prediction markets on consequential events create the same misalignment at a more fundamental level. The people holding the financial interests are anonymous global participants. The people living with the outcomes are everyone else.

Part VII: The Industry's Real Vulnerability

This is not fundamentally a story about Polymarket. Polymarket is simply the largest and most visible participant in a rapidly expanding sector. Regulated exchanges, decentralised protocols, and new entrants are all competing for market share in prediction markets. The issue is not whether one company succeeds or fails. The issue is what happens when an entire industry grows around the financialisation of consequential events.

The regulatory response has already begun. As of mid-2026, Polymarket has been blocked or restricted in at least 34 countries. Portugal acted after $120 million was bet on its presidential election with a suspicious late surge in activity. Ukraine restricted the platform specifically over military outcome markets. Singapore's Gambling Regulatory Authority moved in early 2025. Argentina's courts ruled it an unlicensed gambling service in March 2026. Spain ordered ISP blocks in May 2026. India classified prediction markets as prohibited online money gaming under legislation that came into force the same month. The bans are not coordinated. They are independent jurisdictions arriving at the same conclusion through their own legal and regulatory frameworks.

But here is what the ban data actually tells us — and what the industry has so far avoided confronting. These are first-wave bans. Every one of them is grounded in existing gambling regulation, financial licensing frameworks, or unlicensed operator provisions. Portugal did not ban Polymarket because it feared someone would manipulate a presidential election for financial gain. It banned it because the gambling licensing framework said it should. Ukraine did not act on the basis of a sophisticated analysis of market reflexivity or coordination-without-communication risk. It acted because prediction markets on military outcomes fell outside any recognised regulatory category.

First-wave bans can be resolved. Licences can be obtained. Compliance frameworks can be built. Polymarket's CFTC approval in the United States is exactly this dynamic playing out — the platform successfully navigating the first wave by obtaining the regulatory credential that legitimises its operation.

The second wave has not arrived yet. It will not be triggered by licensing gaps. It will be triggered by an incident — and it will be responded to not with regulatory adjustment but with prohibition. The second wave bans will come from national security ministries, not gambling regulators. They will be drafted in the language of public order and state security, not consumer protection. And they will not be resolved by obtaining a licence from the Commodity Futures Trading Commission.

The industry's current strategy — engage regulators, build compliance infrastructure, obtain licences, point to CFTC approval as proof of legitimacy — successfully addresses the first wave and does nothing whatsoever about the second. They are solving the wrong problem. And they may be solving it well enough to survive long enough to encounter the problem they cannot solve.

The prediction market industry has focused its integrity efforts on the question that lawyers and regulators are most comfortable asking: is insider trading occurring, and how should it be prosecuted?

This is the wrong question — or rather, it is a much smaller question than the one that will ultimately determine the industry's fate.

History suggests that financial innovations are not judged solely on their benefits. They are judged by their most visible failures.

Derivatives expanded rapidly until 2008 transformed them from sophisticated instruments into symbols of systemic risk. High-frequency trading promised more efficient markets and delivered them — alongside flash crashes and structural advantages that regulators are still struggling to address. Online gambling was celebrated for consumer freedom until its social costs became politically impossible to ignore.

Prediction markets may face the same dynamic — compressed and accelerated, because the potential failures are not financial collapses but real-world catastrophes.

The industry's greatest vulnerability is not that markets can be manipulated. Every financial market faces manipulation risks. It is that a sufficiently shocking real-world incident could transform a niche regulatory concern into a political crisis — and that once that happens, governments may not distinguish between responsible platforms, regulated exchanges, and decentralised protocols. They may simply decide that some categories of events should never have become tradable assets.

The triggering event need not prove that a prediction market caused harm. It merely needs to create a public perception that the market was financially connected to the harm. An intelligence official caught profiting from classified information. A government insider trading on a decision they helped shape. A political assassination market attracting attention after an attack. Evidence that participants attempted to influence journalists, officials, or information flows to protect their positions.

Whether the market actually caused the event may become politically irrelevant. The question asked by the public and by legislators will be far simpler: why was this tradeable in the first place?

Most of the mechanisms prediction market platforms have built are retrospective. They identify misconduct after it occurs. They freeze accounts after profits have been made. They investigate after information has leaked. They prosecute after the event has already happened.

For ordinary financial misconduct, this may be sufficient. For markets tied to war, political violence, or other catastrophic outcomes, the calculus is entirely different. The consequences of those events cannot be unwound by freezing an account, voiding a contract, or securing a conviction years later.

Prediction markets may require millions of successful trades to build legitimacy. They may only require one catastrophic incident to lose it — permanently, globally, and irreversibly.

The asymmetry is stark. Prediction markets may require years of responsible operation and millions of accurate forecasts to build the institutional credibility that makes them useful. They may require only one sufficiently shocking incident to lose it. A market linked to an assassination attempt, a military leak, a terrorist attack, or another catastrophic event would not be judged solely on legal liability or causation. It would be judged politically. And political systems rarely wait for perfect evidence before deciding that a category of activity has become intolerable.

Part VIII: Just Because We Can

There is a pattern in the history of technology that repeats with depressing regularity. A capability emerges. It is celebrated for what it makes possible. The ethical questions are deferred — there will be time for those later, once the technology has proven itself. And then something happens that makes those deferred questions suddenly, urgently, unavoidably necessary.

Cryptocurrency added a specific dimension to this pattern. Its foundational philosophy — permissionless by default, censorship-resistant by design, governed by code rather than institutions — was genuinely idealistic in origin. The vision was financial infrastructure that could not be captured by states, intermediaries, or gatekeepers.

What this philosophy produced, alongside genuine innovation, was also infrastructure with no mechanism for asking whether a given application should exist. Permissionless means no permission required — including no permission to decide whether anonymous global participants should be financially incentivised around the outcomes of wars, assassinations, and catastrophic events.

The technology did not ask whether it should be used this way. It made it possible. And once something is possible and profitable, markets find it.

Every civilisation draws lines around what can be commodified. The question is not whether markets can price a thing — they can price almost anything. The question is whether pricing it changes what it is, and whether that change is acceptable.

Prediction markets on consequential events price human decisions about war and peace, life and death, security and catastrophe. The act of pricing those decisions creates financial interests in their outcomes — interests held anonymously, globally, beyond the reach of any single jurisdiction, by participants who bear none of the consequences of the outcomes they are trading on.

This is not a question about Polymarket specifically. Polymarket is merely the largest current expression of a category of instrument. Were it to disappear tomorrow, the category would remain, the infrastructure would remain, and the incentive structures would remain. The question is about what the category does to reality — and whether reality, specifically the most fragile and consequential parts of it, should be an asset class at all.

Having established what prediction markets on consequential events actually do — how they reward influence over information and events, how they coordinate incentives without requiring communication, how they reach into the decisions of generals and negotiators and journalists who never consented to be priced — we can now ask the question directly: should anyone be permitted to place a financial bet on whether part of humanity will still exist next week? On whether a war starts or a ceasefire holds? On whether a leader lives or dies? On whether a society's critical infrastructure survives the month?

Modern societies have repeatedly concluded that some categories of human activity should not be transformed into markets, even when markets are technically capable of pricing them — not because the prices would be inaccurate, but because creating financial interests in certain outcomes changes the relationship between the holders of those interests and the events themselves in ways that cannot be fully governed, monitored, or reversed.

Prediction markets on consequential events are not forecasting tools that happen to carry financial incentives. They are financial instruments that happen to be priced on human survival. That distinction matters.

Nothing catastrophic has happened yet. But that is not evidence that the incentive structures are safe. It is evidence that the price is not right yet.

Every month of growth, every new geopolitical market, every additional billion dollars in open positions raises the liquidity available to whoever eventually decides the payout justifies the act. The population of people with operational influence over consequential events is not shrinking.

Human desperation, ideology, and recklessness are not variables that compliance teams can reduce. The industry's success is, in a precise mechanical sense, increasing the severity of the incident it is building toward. The longer the threshold takes to cross, the larger the markets will be when it is crossed, and the more catastrophic the event that clears it.

The world will eventually decide it matters enough to act on. The only real question is whether that decision arrives before or after the incident that makes it unavoidable — and whether, when it arrives, the lesson will have cost more than it needed to.