This article is reprinted with authorization from Automatic Insight, author: Rhythm Editorial Department, copyright belongs to the original author.
On October 7, 2025, the Intercontinental Exchange (ICE) announced a $2 billion investment in Polymarket.
This financial giant, which owns the New York Stock Exchange, was founded in 1792 and has witnessed nearly all significant moments in American financial history. Meanwhile, Polymarket, a company established just five years ago, has seen its valuation soar to $9 billion.
The valuation logic behind this deal has left many people puzzled.
Polymarket has only 300,000 monthly active users, with an average valuation of $30,000 per user. In comparison, sports betting giant DraftKings has 10 million users and a market capitalization of $16 billion, with each user valued at only $1,600, about one-twentieth of Polymarket's valuation.
Even more perplexing is Polymarket's business model itself.
When you open the website, you will see a variety of "bets": Will Trump run for president again in 2028? How much will the Federal Reserve lower interest rates at its next meeting? Will a certain tech company's quarterly earnings exceed expectations?
People place bets on these questions, making money if they win and losing money if they don't. This seems no different from online gambling sites.
But why would the most conservative financial institutions on Wall Street invest $2 billion in a platform that looks like a gambling site?
The answer lies in a concept that has existed for nearly 40 years but has only truly exploded today, known as prediction markets.
It concerns how information is produced, priced, and used, who has the authority to define truth, and how we can make more accurate decisions and obtain more transparent truths in an era filled with uncertainty.
This is a story that began in 1988.
In the 1988 U.S. presidential election, Bush faced Dukakis.
That year, several professors at the University of Iowa conducted an experiment. They invited 192 faculty and students, each to wager no more than $500 on who would ultimately be elected between Bush and Dukakis.
The rules of the experiment were simple. If you thought Bush would win, you bet on Bush. As more people participated, the amount required for a single bet would increase; conversely, when someone withdrew, the amount would decrease accordingly. Those who bet early could exit at the market price after it rose, earning a profit. The maximum amount for a single bet was $100, meaning that when the amount required to bet on Bush stabilized at $65, the market's collective judgment was that Bush had a 65% chance of winning.
This experiment was named the "Iowa Political Stock Market," later renamed the "Iowa Digital Prediction Market." It was the world's first prediction market using real money.
The results shocked everyone. The collective judgment of these 192 ordinary people was more accurate than all the polls at the time, not only correctly predicting the ultimate winner but also closely matching the actual vote percentages in each state.
Interestingly, this accuracy was not coincidental; it was repeatedly validated in subsequent elections.
In 1992, Clinton vs. George H.W. Bush; in 1996, Clinton vs. Dole; in 2000, George W. Bush vs. Gore, the Iowa Digital Prediction Market consistently outperformed professional polling agencies.
Why could 192 ordinary people be more accurate than experts?
Traditional polls ask, "Who would you vote for?" But when people answer this question, they might lie due to social pressure, change their minds at the last minute, or not vote at all. Polls capture what people think at that moment, and thoughts can be fleeting.
Prediction markets ask a different question: How much would you be willing to pay for this judgment? When real money is on the line, people think more seriously and express themselves more honestly. You might verbally support a candidate, but if you have to put money on the line, you will reassess your judgment.
More importantly, the market automatically aggregates everyone's information. You might know the situation in your city, I might understand the thoughts of young people, and he might have insights into a particular industry. When we all vote with money, the information scattered across countless individual minds is aggregated into a single number. This number is not the judgment of a single expert but the weighted average of all participants' information.
This is what economists refer to as "collective intelligence." When a large number of independent individuals make judgments based on their information, the average result is often more accurate than that of any single expert.
However, this experiment soon encountered a critical compliance issue: legally, does this count as gambling or illegal securities trading?
U.S. law has strict regulations on both gambling and financial transactions. The professors at the University of Iowa were therefore particularly cautious, keeping the experiment's scale very small: each person could only invest up to $500, limited to academic research, and only open to political election markets. They hoped to avoid the legal gray area through such self-restraint.
In 1992, they finally received a "no-action letter" from the Commodity Futures Trading Commission (CFTC). The meaning of this letter was that the regulatory agency would not prosecute them, but this did not equate to formal approval of their activities.
It was precisely because of this gray identity between legality and illegality that prediction markets remained trapped in the ivory tower for the next several years.
The Iowa Digital Prediction Market continued to operate on a small scale, with only a few thousand participants in each election and total trading volumes of only a few hundred thousand dollars. It proved the feasibility and accuracy of prediction markets but also exposed its dilemma: how to survive in the cracks of the law.
In 1999, two New York futures traders, Ron Bernstein and Sean McNamara, saw the commercial potential of the Iowa experiment and founded Intrade, a prediction market open to the world. Intrade did not limit the amount of bets, topics, or nationalities and was headquartered in Ireland to try to circumvent U.S. regulations. From 1999 to 2012, Intrade grew from a niche website to the world's largest prediction market, with over 80,000 members from 162 countries.
The 2012 U.S. presidential election was Intrade's most glorious moment.
That year, every journalist, analyst, and investor following the election would keep an eye on the real-time price of "Obama elected" on Intrade. This number was more intuitive than any poll and reflected changes in public sentiment more quickly than any expert prediction.
On October 3, the first debate between Obama and Romney took place. Obama performed poorly, being on the defensive throughout. Within hours after the debate, the price of "Obama elected" on Intrade plummeted from $78 to $65. This drop far exceeded any changes in polls and sparked widespread discussion, with some believing that the market reaction on Intrade was overblown, while others thought polling agencies were too slow to respond.
However, in the end, Intrade correctly predicted the election results in all states except Florida and Virginia.
Since then, The New York Times, The Wall Street Journal, and The Economist began citing Intrade's data, viewing it as a more reliable indicator than polls. In October and November of that year, Intrade's monthly page views exceeded 50 million, receiving hundreds of media mentions. People began to believe that the market might understand politics better than experts.
But behind the glory, Intrade faced three fatal problems.
The first problem was regulation.
As early as 2005, Intrade applied to the CFTC to establish a regulated exchange in the U.S., but the CFTC's stance was ambiguous, neither clearly approving nor prohibiting it. Thus, Intrade operated in the gray area for over a decade.
After the 2012 election, the CFTC finally sued Intrade, accusing it of illegally offering commodity futures contracts, including gold and oil, to U.S. users, which violated U.S. law. Intrade had no choice but to close all U.S. user accounts a month later and stop new user registrations, losing its largest market overnight.
The second problem was liquidity.
Intrade's trading volume was actually not large. Two weeks before the 2012 election, someone used only $17,800 to push Romney's winning probability from 40% to 49%. The Washington Post at that time questioned whether a market price so easily manipulated had any reference significance.
Intrade responded that trades were spread across 40 accounts, with no single individual purchasing more than 15% of the trading volume, so this did not constitute manipulation, but was simply due to "the market order book being too thin."
However, this did not dispel people's doubts. Strangely, Intrade consistently showed significant price discrepancies with other betting sites, with Obama's winning probability on Intrade only between 60% and 70%, while betting companies' odds had already exceeded 80%.
In normal financial markets, such discrepancies would be quickly arbitraged away. Theoretically, if Obama's winning probability on Intrade was only 65%, while other betting sites implied an 85% probability, traders could place opposite bets on the two platforms, buying "Obama wins" on Intrade and betting "Obama loses" on the betting site. With proper calculations, theoretically, they could lock in profits regardless of the outcome.
But this price discrepancy persisted on Intrade for months. This was not because no one noticed, but because the market's liquidity was too poor, lacking sufficient depth to correct prices. Such markets are often the easiest to manipulate.
The third problem was financial.
On March 10, 2013, Intrade announced on its official website that it was "suspending all trading and closing accounts due to financial irregularities." The announcement did not specify details, only citing Irish law and stating that it was forced to take action.
The outside world generally believed that Intrade fell into financial trouble after losing the U.S. market. Trading volume plummeted from one million trades in 2012 to just 50,000 the following year, with revenue nearly dried up. The company had planned to relaunch as "Intrade 2.0," but ultimately failed to do so. In August 2014, Intrade announced it would permanently close and refund all user funds.
A once-glorious prediction market empire thus collapsed.
The collapse of Intrade exposed three fatal weaknesses of centralized prediction markets: regulation can shut you down overnight, insufficient liquidity makes you vulnerable to manipulation, and excessive reliance on a single market can leave you very fragile.
But Intrade also proved one thing: there is a huge market demand for prediction markets. At its peak, thousands of people used it daily, and the media cited its data every day. The problem lies not in prediction markets themselves, but in how to address these three fatal weaknesses.
Polymarket has built its entire platform on the blockchain. Unlike traditional companies, such as Intrade, whose trading records exist on servers in Ireland and can be shut down by government order, the blockchain's ledger is distributed across thousands of computers worldwide, with no single entity able to take it offline. All transaction records are public and transparent, allowing anyone to view and verify them, but they cannot be tampered with.
This architecture brings three key advantages, which correspond precisely to the three fatal weaknesses exposed by Intrade in its time.
The first advantage is decentralization.
Polymarket does not have a centralized server; all transactions are automatically executed through smart contracts on the blockchain. Even if the company itself were shut down, the platform would still operate. The contract code is open and transparent, and anyone can invoke it without needing Polymarket's permission.
However, "decentralization" was soon tested in reality. In January 2022, the CFTC accused Polymarket of providing commodity futures trading to Americans without registration and imposed a $1.4 million fine. Subsequently, the platform began to block U.S. users through technologies such as IP recognition.
But unlike Intrade, decentralization prevented Polymarket from being completely destroyed. It continued to operate globally, with trading volumes increasing rather than decreasing.
In July 2025, Polymarket acquired QCEX, an exchange licensed by the CFTC, for $112 million; two months later, it was allowed to return to the U.S., ending three years of exile.
Polymarket's second advantage is liquidity.
Intrade relied on buyers and sellers to match themselves; once the market cooled, the impact of a single transaction on price fluctuations would be magnified, which is why it only took $17,800 to push Romney's winning probability up by 9 percentage points back then.
Polymarket solved this problem with technology. It introduced an automated market-making mechanism that allows the system to provide liquidity at all times. No matter when you buy or sell, transactions can be executed immediately, even in niche markets.
This significantly increased market depth and made manipulation more difficult. To shake up a popular market on Polymarket, the required funds could be ten or even a hundred times that of Intrade. During the 2024 U.S. presidential election, Polymarket's daily trading volume once exceeded $100 million, making it nearly impossible for a single whale to manipulate prices.
The third advantage is transparency.
Intrade's finances were always a black box. Until just before the platform collapsed, users had no idea what problems the company was facing or whether their funds were safe.
In contrast, on Polymarket, all funds are on-chain, and anyone can view the platform's fund size, open contracts, trading volume, and liquidity in real-time. This level of transparency virtually eliminates the possibility of financial misconduct. If the platform misappropriated user funds, the on-chain data would immediately expose it, leaving no room for evasion.
In the 2024 U.S. presidential election, Polymarket experienced its moment of glory.
That year, Trump faced Harris. Throughout the election season, Polymarket's data was widely cited by global media. The New York Times, The Wall Street Journal, The Financial Times, and Bloomberg reported almost daily on the latest probability predictions on Polymarket. In key battleground states like Pennsylvania and Arizona, Polymarket also nearly perfectly predicted the final narrow margins.
On the evening of November 5, election day, while traditional media cautiously reported on a "tight race," Trump's winning probability on Polymarket had already soared above 90%.
A few hours later, Trump announced his victory.
No traditional polling agency could match this level of accuracy and real-time responsiveness.
That year, Polymarket's monthly active users reached 300,000, monthly trading volume hit $1.3 billion, and the on-chain locked value reached $170 million. More importantly, Polymarket transformed from a niche product in the cryptocurrency space into a focal point of mainstream media attention.
However, Polymarket also faced new challenges.
In September 2025, a competitor named Kalshi suddenly emerged. Kalshi took a completely different path: it did not use blockchain but instead followed a fully compliant route.
Starting in 2018, it underwent a six-year CFTC approval process, ultimately becoming the first prediction market exchange to receive formal approval from the CFTC, paving the way for compliant prediction markets in the U.S.
Kalshi primarily targeted the sports betting market. With frequent sporting events and quick trading turnover, it quickly accumulated a large user base and trading volume. By December 2024, Polymarket still held a 95% market share, nearly monopolizing the entire prediction market industry. But by September 2025, Kalshi's market share had soared to 65%, surpassing Polymarket.
Kalshi's rise made Polymarket realize that technological advantages alone were not enough; regulatory recognition and support from mainstream capital were equally important.
This is the context for ICE's investment.
As the parent company of the New York Stock Exchange, ICE represents the most conservative faction on Wall Street. Yet, this very institution chose to invest in a cryptocurrency platform that appears almost like gambling to outsiders.
Because ICE saw not gambling from the beginning, but the value of information itself.
In the age of AI and big data, information itself is the most valuable asset. Traditional sources of information, such as polls, expert predictions, and think tank reports, are "top-down," produced by a few experts, with ordinary people only passively receiving them.
In contrast, prediction markets provide "bottom-up" information. Thousands of people vote with real money, and the market price reflects their collective judgment. This judgment is often more accurate than any expert's because it aggregates everyone's information, experience, and intuition.
What ICE aims to do is package this "collective intelligence" into financial products to sell to institutional investors.
Hedge funds need to know whether the Federal Reserve will lower interest rates next, whether Trump's tariff policy will pass, or whether a certain geopolitical conflict will escalate. Multinational corporations need to know whether a coup will occur in a certain country or whether a specific policy will be implemented.
These questions cannot be answered by traditional financial instruments. Stock prices reflect company value, bond prices reflect credit risk, and futures prices reflect supply and demand for commodities. But no financial instrument can directly tell you "the probability of a certain event occurring."
Prediction markets fill this gap. They provide not a "yes or no," but a precise probability. This probability is updated in real-time, changing with new information, faster and more accurately than any expert report.
According to the agreement between ICE and Polymarket, ICE will become the global distributor of Polymarket data, providing institutional clients with real-time market sentiment indicators, political risk pricing, and economic event probability forecasts. This data can be integrated into investment decision models, helping fund managers, risk managers, and corporate strategy departments make better judgments.
Another consideration for ICE's investment is the openness of the data.
Although Kalshi has a larger market share, it is a centralized, regulated platform. Its data is not fully public and must be accessed through APIs, which typically require payment and authorization. Additionally, it is constrained by U.S. regulations, and certain sensitive markets may not be opened. It is a closed ecosystem.
In contrast, Polymarket is decentralized. All data is on-chain, and anyone can access and use it for free without needing Polymarket's permission. It can open any market without being restricted by a single country's regulations. Compared to Kalshi, it is a more open ecosystem where people worldwide can place bets, and anyone can develop new applications based on its data.
For ICE, Polymarket is not just a trading platform but also a data infrastructure. Just like early investors in Google were not focused on the search engine itself but on the data and network effects behind it.
The $9 billion valuation reflects this logic.
By conventional valuation logic, Polymarket does indeed seem overly expensive. But what ICE is betting on is not its current business scale but its strategic position in the future "financialization of information" era.
Polymarket is no longer just a trading platform; it is a news media outlet, a data source, and a cultural phenomenon. Its valuation reflects more of its cultural relevance and informational influence.
Those 300,000 trading users are just a small fraction of the millions consuming Polymarket's content. Every day, countless people open Polymarket to check the latest odds on events without actually trading. Now, financial institutions like Goldman Sachs have also begun to cite Polymarket's data in their research reports, no longer limited to macro events like Federal Reserve policies or government shutdowns.
This is akin to early Twitter or Facebook. Their valuations were not based on their advertising revenues at the time but on their potential to become "social infrastructure." Polymarket's valuation is based on its potential to become "information infrastructure."
ICE and Polymarket also plan to collaborate on "future tokenization" projects, exploring pathways for prediction markets to integrate into the mainstream financial system. This means that prediction market contracts are expected to be tokenized and freely traded in secondary markets; their data may also be packaged into new financial derivatives; and in the longer term, there may be possibilities for traditional financial assets to interconnect with prediction markets.
The details of these plans have not yet been announced, but the direction is clear: ICE aims to push prediction markets from niche experiments into the mainstream financial system.
The concept of prediction markets was proven feasible as early as 1988, but the real explosion awaited 2025.
Technology has finally caught up with the concept.
Early blockchain technology was slow and expensive, with a single transaction taking minutes to confirm and fees reaching tens of dollars. Ordinary users could hardly use it. At that time, some attempted to build prediction markets on-chain, but all failed due to poor user experience.
With technological advancements, transaction confirmations now take only seconds, and costs have dropped to just a few cents, making the operational experience nearly indistinguishable from ordinary websites. These improvements have made prediction markets truly usable, with faster transactions and lower costs, allowing users to hardly feel the presence of blockchain.
More importantly, there has been a shift in regulatory attitudes.
Since 2012, U.S. regulatory agencies have consistently taken a repressive stance toward prediction markets. The reason is simple: prediction markets look like gambling, and gambling is illegal in most states.
In 2022, Kalshi attempted to launch a prediction contract on "Who Will Control Congress," but it was rejected by the CFTC on the grounds of being "similar to gambling." Faced with regulatory suppression, Kalshi did not back down but chose to sue the regulators. This marked the first time in the history of prediction markets that a company directly challenged a regulatory agency.
Since then, the tide began to turn. In 2023, regulators started discussing the social value of prediction markets, and some scholars and institutions called for a reassessment of their legality.
In 2024, in the case of Kalshi v. CFTC, a federal judge ruled in favor of Kalshi, stating that prediction markets provide an information aggregation service, which is fundamentally different from gambling. This ruling opened new space for the entire industry.
In 2025, the CFTC approved Kalshi for full operation and ended its investigation into Polymarket. As a result, Polymarket was also allowed to return to the U.S. market.
The shift in regulatory attitude is the result of multiple factors.
One reason is that regulatory agencies gradually recognized the informational value of prediction markets. In the 2024 election, the predictions from Polymarket and Kalshi were more accurate than traditional polls, leading regulators to realize that prediction markets are not merely gambling; they indeed produce valuable information.
Another reason is international competition. If the U.S. excessively restricts prediction markets, the industry is bound to flow elsewhere. Countries like Singapore, the UK, and Switzerland are actively attracting fintech companies, and the U.S. is reluctant to easily lose its dominance in this emerging field.
ICE's investment further confirms this change in attitude. When the most conservative forces on Wall Street began to embrace prediction markets, it became difficult for regulators to continue viewing them as "illegal gambling." ICE's endorsement brought new legitimacy and mainstream recognition to prediction markets.
From 192 individuals at the University of Iowa in 1988 to 300,000 users on Polymarket in 2025, prediction markets have come a long way over 37 years.
These 37 years have witnessed a transfer of informational power.
The problem in modern society has never been a lack of information, but rather an excess of it, making it difficult to discern truth from falsehood.
Traditional polls frequently miss the mark. In the 2016 U.S. election, nearly all polls predicted Hillary would win, yet Trump was elected. In 2020, polls predicted a decisive victory for Biden, but the election was very close. Expert predictions are similarly flawed; in areas like the pandemic, economy, and geopolitics, the final outcomes often contradict their judgments.
Their predictions are widely reported by the media, adopted by decision-makers, and believed by the public. Yet they are often wrong, and when they are, they face no consequences.
Prediction markets offer a new answer: you don't have to believe any individual or institution because you can trust the market.
Here, anyone can participate in predictions, but you must put your money where your mouth is. If you're right, you make money; if you're wrong, you lose money. This mechanism transforms "prediction" from a power into a responsibility.
You don't need to be an economist to judge whether the Federal Reserve will lower interest rates; you don't need to be a political scientist to assess whether Trump will win. Everyone's judgment contributes to market prices, aggregating into a form of collective wisdom. This wisdom is often more accurate than that of any single expert.
Those with insider information and judgment advantages will place large bets on the outcomes they know in advance, which in turn enhances the accuracy of market predictions.
An economist may be well-versed in macro theory but may not understand the unemployment situation in a small town; a political scientist may be familiar with electoral systems but may not know what young people are thinking. Prediction markets aggregate these dispersed knowledge and experiences, forming a more comprehensive judgment.
This mechanism allows prediction markets to become a self-evolving system. It does not rely on centralized authorities to determine who is right or wrong; the market itself rewards correct judgments and punishes incorrect ones.
For ordinary people, prediction markets also mean a more transparent world.
You can see in real-time "the market's perceived probability of an event occurring," rather than passively accepting the opinions of experts. You can observe how this probability changes with new information, thereby understanding how the market digests information. This data provides new references for personal decision-making, whether in investment, entrepreneurship, or everyday life choices.
Prediction markets also imply fairer game rules.
Here, your judgment carries the same weight as that of Wall Street analysts. As long as you are willing to stake real money, your opinion can influence market prices. This is in stark contrast to traditional information production models, where only a few voices are heard, and only a few judgments can influence public perception. In prediction markets, everyone can have a voice, and everyone can express their judgment with their funds.
ICE's $2 billion investment marks the transition of prediction markets from "marginal innovation" to "mainstream infrastructure," from "gray areas" to "legitimate industries," and from "toys of the cryptocurrency circle" to "tools of Wall Street."
But more importantly, it symbolizes the arrival of a new era. In this era, information is no longer monopolized by a few experts but produced by the market; truth is no longer defined by authority but determined by collective wisdom.
This era has only just begun.
Related: Bloomberg: Polymarket's Nobel Peace Prize Bets Under Scrutiny
Original: “A $2 Billion Bet on the Next Financial Boom”
免责声明：本文章仅代表作者个人观点，不代表本平台的立场和观点。本文章仅供信息分享，不构成对任何人的任何投资建议。用户与作者之间的任何争议，与本平台无关。如网页中刊载的文章或图片涉及侵权，请提供相关的权利证明和身份证明发送邮件到support@aicoin.com，本平台相关工作人员将会进行核查。
