A $2 billion "game of probabilities": Is the prediction market approaching its "singularity" moment?

Prediction markets are evolving from a marginalized "crypto toy" to being regarded as a serious financial instrument.

Written by: Bitget Wallet Research

From casual discussions like "Will Zelensky wear a suit?" to global focal points such as the US presidential election and the Nobel Prize, prediction markets have always managed to periodically "catch fire." However, since Q3 2025, a real storm seems to be brewing:

• In early September, industry giant Polymarket received regulatory approval from the US CFTC, returning to the US market after three years;
• In early October, ICE, the parent company of the New York Stock Exchange, planned to invest up to 2 billion USD in Polymarket;
• In mid-October, the weekly trading volume of prediction markets hit a historic high of 2 billion USD.

The simultaneous arrival of capital surges, regulatory green lights, and market frenzy has been accompanied by rumors of a Polymarket token launch—where does this wave of enthusiasm come from? Is it just another short-lived hype, or is it a "value singularity" for a brand-new financial track? Bitget Wallet Research will take you through a deep dive into the underlying logic and core value of prediction markets in this article, and provide a preliminary assessment of the core dilemmas and development directions they face.

I. From "Dispersed Knowledge" to "Dual Oligopoly": The Evolution of Prediction Markets

Prediction markets are not a creation of the crypto world; their theoretical foundation can even be traced back to 1945. Economist Friedrich Hayek, in his classic discourse, proposed that fragmented, localized "dispersed knowledge" can be effectively aggregated by the market through price mechanisms. This idea is considered the theoretical cornerstone of prediction markets.

In 1988, the University of Iowa launched the first academic prediction platform—the Iowa Electronic Markets (IEM), which allowed users to trade futures contracts on real-world events (such as presidential elections). Over the following decades, numerous studies have consistently confirmed that a well-designed prediction market often outperforms traditional opinion polls in accuracy.

However, with the advent of blockchain technology, this niche tool found new ground for large-scale implementation. The transparency, decentralization, and global accessibility of blockchain provided prediction markets with an almost perfect infrastructure: automated settlement via smart contracts breaks down traditional financial barriers, allowing anyone worldwide to participate, thereby greatly expanding the breadth and depth of "information aggregation." Prediction markets have gradually evolved from a niche gambling tool into a powerful on-chain financial sector, becoming deeply intertwined with the "crypto market."

Source: Dune

Data from the Dune platform intuitively confirms this trend. On-chain data shows that the current crypto prediction market has formed a highly monopolized "dual oligopoly" structure: giants Polymarket and Kalshi together account for over 95% of the market share. Stimulated by both capital and regulatory tailwinds, this sector is being activated as a whole. In mid-October, the weekly trading volume of prediction markets surpassed 2 billion USD, exceeding the previous historical peak before the 2024 US election. In this round of explosive growth, Polymarket, with its key regulatory breakthrough and potential token expectations, has temporarily gained an edge in its fierce competition with Kalshi, further consolidating its leading position.

II. "Event Derivatives": Beyond Gambling—Why Is Wall Street Betting?

To understand why ICE is investing heavily in Polymarket, one must strip away the "gambling" facade of prediction markets and see their core as a "financial instrument." The essence of prediction markets is an alternative trading contract, classified as an "event derivative."

This is different from the "price derivatives" we are familiar with, such as futures and options. The latter's underlying asset is the future price of an asset (like crude oil or stocks), while the former's underlying asset is the outcome of a specific "event" (such as an election or climate event). Therefore, the contract price does not represent asset value, but rather the collective consensus of the market on the "probability of the event occurring."

With the empowerment of Web3, this difference is further amplified. Traditional derivatives rely on complex mathematical models like Black-Scholes for pricing and are settled through brokers and centralized exchanges; on-chain prediction markets, however, are executed automatically via smart contracts and settled by oracles, with their pricing (such as AMM algorithms) and liquidity pools fully transparent on-chain. This greatly lowers the entry barrier but also introduces new risks (such as oracle manipulation and contract vulnerabilities), which stand in stark contrast to the counterparty and leverage risks of traditional finance.

Comparison Table: Prediction Markets vs. Traditional Financial Derivatives

This unique mechanism is precisely what attracts mainstream financial institutions. It offers three core values that traditional markets cannot match, which is the key reason why giants like ICE are truly betting on it:

First, it is an advanced "information aggregator" that reshapes the landscape of information equality. In an era of AI-generated content, fake news, and information silos, "truth" has become expensive and hard to discern. Prediction markets provide a radical solution: truth is not defined by authorities or the media, but is "bid for" by a decentralized, economically incentivized market. It responds to the growing distrust (especially among the younger generation) of traditional information sources, offering a more honest alternative information source where "money votes." More importantly, this mechanism goes beyond traditional "information aggregation" by enabling real-time pricing of "truth," forming a highly valuable "real-time sentiment indicator," and ultimately achieving information equality across all dimensions.

Second, it turns "information asymmetry" itself into an asset, opening up a brand-new investment track. In traditional finance, investment targets are "property certificates" such as stocks and bonds. Prediction markets, however, create a new tradable asset—"event contracts." This essentially allows investors to directly convert their beliefs or "information advantages" about the future into tradable financial instruments. For professional information analysts, quantitative funds, and even AI models, this is an unprecedented profit dimension. They no longer need to express their views indirectly through complex secondary market operations (such as going long/short on related company stocks), but can directly "invest" in the event itself. The huge trading potential of this new asset class is the core interest point attracting exchange operators like ICE.

Finally, it creates a risk management market where "everything can be hedged," greatly expanding the boundaries of finance. Traditional financial instruments struggle to hedge the uncertainty of "events" themselves. For example, how can a shipping company hedge the geopolitical risk of "whether a canal will be closed"? How can a farmer hedge the climate risk of "whether rainfall in the next 90 days will be less than X millimeters"? Prediction markets provide a perfect solution. They allow participants in the real economy to convert abstract "event risks" into standardized, tradable contracts for precise risk hedging. This is equivalent to opening up a brand-new "insurance" market for the real economy, providing a new entry point for finance to empower the real economy, with potential far beyond imagination.

III. Hidden Concerns Amidst Prosperity: The Three Core Dilemmas Facing Prediction Markets

Despite a clear value proposition, prediction markets still face three interlinked real-world challenges on their path from "niche" to "mainstream," which together form the industry's ceiling.

The first dilemma: The contradiction between "truth" and "arbiter," i.e., the oracle problem. Prediction markets are "outcome-based trading," but who announces the "outcome"? A decentralized on-chain contract paradoxically relies on a centralized "arbiter"—the oracle. If the event itself is ambiguously defined (such as the definition of "wearing a suit"), or if the oracle is manipulated or makes a mistake, the entire market's foundation of trust can collapse instantly.

The second dilemma: The contradiction between "breadth" and "depth," i.e., the liquidity drought of the long tail. The current boom is highly concentrated on headline events like the "US presidential election." However, the true value of prediction markets lies in serving those vertical, niche "long-tail markets" (such as the aforementioned agricultural and shipping risks). These markets naturally lack attention, resulting in extremely poor liquidity, making prices easily manipulated and thus losing their practical function of information aggregation and risk hedging.

The third dilemma: The contradiction between "market makers" and "informed traders," i.e., the "adverse selection" problem of AMMs. In traditional DeFi, AMM market makers (LPs) bet on market volatility to earn trading fees. But in prediction markets, LPs are directly betting against "informed traders." Imagine a market on "whether a new drug will be approved," where LPs are betting against scientists with insider information—this is a sure-loss "adverse selection." Therefore, in the long run, automated market makers are unlikely to survive in such markets, and platforms must rely on expensive manual market makers to operate, greatly limiting their scalability.

Looking to the future, the breakthrough points for the prediction market industry will inevitably revolve around the above three dilemmas: more decentralized and manipulation-resistant oracle solutions (such as multi-party verification and AI-assisted review) are the cornerstone of trust; guiding liquidity into long-tail markets through incentive mechanisms and better algorithms (such as dynamic AMMs) is key to realizing real-world value; and more sophisticated market maker models (such as dynamic fees and asymmetric information insurance pools) are the engines for scaling up.

IV. Conclusion: From "Probability Game" to "Financial Infrastructure"

The CFTC's green light and ICE's entry are clear signals: prediction markets are evolving from a marginalized "crypto toy" to being regarded as a serious financial instrument. With "aggregating truth" as their core value and "event derivatives" as their financial core, they provide a new dimension of risk management for modern finance. Admittedly, the road from "probability game" to "financial infrastructure" is far from smooth. As mentioned above, the oracle dilemma regarding "arbiters," the liquidity challenges of long-tail markets, and the "adverse selection" faced by market makers are all real challenges the industry must soberly confront after the frenzy.

Nevertheless, a new era integrating information, finance, and technology has already begun. When top-tier traditional capital starts to bet heavily on this track, what it leverages will be far more than 2 billion USD in weekly trading volume. This may be a true "singularity" moment—it heralds the mainstream financial system's acceptance of a brand-new asset class (the pricing power of "beliefs" and "the future").