Artemis: The Financialization of Uncertainty

Author: Kaviish, Artemis Analytics Data Analyst; Source: Artemis; Translated by: Shaw Golden Finance

Overview

• Prediction markets turn uncertainty into tradable assets, where prices reflect probabilities backed by capital—not just opinions.

• The industry is expanding rapidly, but current growth is still driven mainly by sports betting rather than high-information-content markets that can demonstrate its long-term value.

• Crypto rails break down geographic and settlement constraints, enabling global participation and improving signal effectiveness through scale.

• Regulatory direction will determine whether liquidity consolidates into deep global markets or fragments into inefficient regional ones.

• The gap between prediction markets’ positioning (information infrastructure) and today’s profit model is widening.

Prediction markets will reach $63.5 billion in trading volume in 2025.

This figure is more than three times the processing volume from the prior year. In only the first 86 days of 2026, the combined trading volume across Kalshi and Polymarket already totals $52.7 billion, with Kalshi at $28.3 billion and Polymarket at $24.3 billion (data from Artemis). On an annualized basis, the full-year scale will reach $223.0 billion. The industry’s last year’s scale more than doubled; this year, it is poised to triple again.

Three years ago, monthly trading volume in prediction markets was under $100 million. Today, the same scale can be achieved in just a few hours. Federal Reserve decisions, elections, escalating geopolitical conflicts, and corporate earnings results—any event can now be transformed into a priced market. That price is not a simple forecast; it is a probability backed by capital, continuously updated as new information flows in.

Pollsters gather opinions, models process historical data, and analysts publish judgments—yet they all share the same structural flaw: prediction errors come at no cost, so accuracy is not strictly required. Prediction markets remove this flaw by attaching financial consequences to forecasting: you profit if you’re right, and you lose money directly if you’re wrong.

The Federal Reserve has already confirmed that prediction markets are more accurate and more responsive than traditional methods. In a Federal Reserve research report in January 2026, the Fed found that, in forecasting overall CPI data, Kalshi’s prediction market performance was “significantly better than Bloomberg consensus expectations.” Since 2022, on the day of every Federal Open Market Committee (FOMC) meeting, the mode of the probability distribution in Kalshi markets has matched the最终 realized federal funds rate exactly (Federal Reserve FEDS working paper 2026-010).

Why now

Prediction markets have existed for decades. For example, Iowa Electronic Markets went live as early as 1988; another platform called Intrade also operated for years before shutting down in 2013. Its core idea can be traced even further back to Friedrich Hayek, who argued that markets are the most efficient mechanism for aggregating dispersed information. Although this idea has long been highly persuasive, implementation has never been satisfactory due to structural constraints: limited channel coverage, fragmented liquidity, slow settlement, and restricted scope to local markets. Today, all of these problems are improving quickly.

Native channel penetration

Prediction markets are now directly embedded into brokerages, media platforms, and API interfaces. Robinhood has added event contracts beyond stocks and options; media platforms such as CNN have started labeling probabilities alongside news headlines, turning each news cycle immediately into tradable opportunities.

In the past, trading based on macro judgments required opening accounts, funding them, and waiting for trading hours—now these friction costs have largely disappeared. Prediction markets are becoming the default entry point for people to interact with and trade uncertainty.

Globalized settlement

The second shift is even more critical. In traditional financial systems, prediction markets face layered limits from geography, regulation, and settlement speed; on crypto infrastructure, these constraints do not exist.

Users in New York, Lagos, or Jakarta can see the same market, the same price, the same opportunity—and execute near-instant settlement.

The higher the participation, the more accurate prediction markets become. Confining a market to a single country naturally weakens signal quality; global expansion increases the reliability of outputs. Crypto technology not only makes prediction markets more efficient, it also makes their price signals more real.

Regulatory chess

Regulation grants prediction markets legitimacy, but it also reveals clear economic interest conflicts. Prediction markets sit at the intersection of finance and gambling, unable to be neatly classified as either—precisely this mismatch creates the contradictions.

This conflict is most evident in sports. In the United States, each state directly benefits from sports betting; in 2025, the market size reached $165.0 billion and generated billions in tax revenue (SportsHandle data). The current system distributes licenses state by state. Operators such as DraftKings and FanDuel must pay high fees, partner with casinos, and bear high tax-rate operating burdens (Legal Sports Report data).

Although prediction markets fall into the same broad category as sports betting, they operate under a different set of rules. From the perspective of state governments, prediction markets divert business away from tax-paying sports betting platforms and bypass regulatory systems that took years to build. This is also why traditional gambling platforms strongly oppose them—not because prediction markets have defects, but because they disrupt the industry interest structure built into the existing mature regulatory framework.

Meanwhile, the U.S. Commodity Futures Trading Commission (CFTC) is moving to regulate prediction markets, while courts are defining the boundaries of its regulatory authority. This battle will determine the final landscape for prediction markets. If forced into a state-by-state patchwork model, liquidity will become more fragmented. But if allowed to operate under a unified financial framework, liquidity will consolidate, and prediction markets can grow into a global system for pricing uncertainty.

Structural risk

At present, the vast majority of revenue for prediction markets comes from sports-related areas.

In the Kalshi platform, sports contracts account for about 83% of its nominal total trading volume. Polymarket’s categories are more diversified, but even so, sports still dominate, accounting for 38.4% of its nominal total trading volume. (Data source: Artemis)

This creates a clear mismatch. Current industry growth is driven by sports markets, which leads outsiders to broadly equate it with the gambling business. Yet this is exactly the area facing the heaviest regulatory pressure. The result is that the current revenue structure is much closer to sports betting than to financial infrastructure.

This concentration of business brings timing-node risk. If sports-related regulation tightens before other categories have formed scale, the industry’s core revenue source will shrink early, before its long-term business model is fully established.

But market valuations are clearly already pricing in future potential. Kalshi’s valuation is $22.0 billion, and Polymarket’s target valuation is also in a similar range—meaning the market expects the scale to far exceed sports betting and expand into institutional-level information markets. Yet this transition has not truly occurred, creating the core contradiction: prediction markets are being valued as information infrastructure, but at this stage they are still monetized through sports products.

Sports markets drive growth, but they are not the long-term moat. Their real value lies elsewhere.

The true core strength of prediction markets

Beyond sports, prediction market products have genuine and strong value across multiple other areas.

Politics

The 2024 U.S. election is strong proof. Polymarket’s presidential election winner markets generated $3.7 billion in total trading volume. The Monday week before the election, Polymarket showed Trump at a 58% win probability and Harris at 42%, while traditional polling (CNN) had both sides essentially tied. In the end, the market prediction was accurate. A peer-reviewed study from the Universidad Carlos III Madrid’s IMDEA Networks Institute (IMDEA Networks Institute), analyzing 86 million bets, found that one full month before the outcome was determined, Polymarket’s prediction accuracy exceeded 94%. (arXiv:2603.03136)

Prediction markets do not replace polls; they incorporate them into the system. Polls become input variables, while prices aggregate all information under an incentive mechanism of “you lose money when you’re wrong” to produce the final judgment.

Economics and policy

This is the most important lane—and where the real moat lies.

Kalshi now provides intraday trading data for macroeconomic indicators, including CPI (month-over-month, year-over-year, and calendar-year), core CPI, PCE inflation, unemployment rate, nonfarm employment data, GDP growth rate, and recession probabilities.

Federal Reserve research has already validated the value of products like this. In 2025, the trading volume of economics markets surged 905%, and technology markets skyrocketed even more—up 1,637%. Although trading volume is still relatively small today, signal quality has reached an institutional level. The implied probability of rate hikes is not just a data point; it is a key signal that affects asset allocation, risk management, and strategic decisions at top financial institutions. It is precisely here that prediction markets shift from consumer products to financial infrastructure.

Why crypto infrastructure is crucial

Moving from consumer products to financial infrastructure hinges on scale, and scale depends on accessibility.

Kalshi operates only within the United States, with a market population of about 330 million. Polymarket is global; it relies on crypto infrastructure so that any user with a network and a wallet can participate. This difference fundamentally changes market size.

Users in New York, Lagos, or Jakarta can all participate in the same market and trade at the same price. No permission is required, and coverage achieves truly global reach. This is the channel advantage brought by crypto.

Crypto also changes how markets are created: it no longer depends on centralized operators; instead, independent adjudication bodies and oracle systems define and settle the outcomes—allowing market scale to far exceed the limit a single platform could support.

All kinds of trading platforms are moving toward this model: beyond stocks, options, and crypto assets, Robinhood has integrated event contracts; around the same time, Hyperliquid also launched prediction-market-style contracts via HIP-4.

Market landscape

Although there are many market participants, the market structure already shows a duopoly oligopoly from the start. Kalshi and Polymarket together control 97.5% of trading volume. The industry’s current change is not further consolidation, but stratification: around these two platforms, a clear tiered structure is taking shape.

Disruption of the super-app

Platforms such as Robinhood and Coinbase are no longer just channel distribution layers—they are moving directly into the trading platform layer.

Robinhood has validated market demand. Prediction markets have become one of its fastest-growing products: in 2025, contract trading volume exceeded 12.0 billion contracts, and users surpassed 1.0 million. Its trading volume also constitutes an overwhelming majority of Kalshi’s total trading volume, giving Robinhood a clear incentive to internalize traffic. In January 2026, Robinhood, through a joint venture with Susquehanna, acquired the MIAXdx exchange and clearing entity, which holds a license from the U.S. Commodity Futures Trading Commission (CFTC).

Coinbase is also copying the same playbook. It first integrated prediction markets by connecting Kalshi contracts, then acquired The Clearing Company to build its own infrastructure. Its strategy is straightforward: first aggregate demand, then perform vertical integration. If it succeeds, the existing market structure will be rewritten. Kalshi’s core positioning is the licensed exchange layer, but if distribution platforms also control both users and the trading venue, this advantage will be weakened. The industry moat will shift from licenses to distribution-channel capability.

Polymarket faces another constraint. In 2022, after settling with the CFTC for $1.4 million, it exited the U.S. market, and then relied on crypto infrastructure to expand globally. When it returned to the U.S. market through QCEX, it launched a completely different product: compared with its unlicensed global service, this version has more restrictions, requires going through broker channels, and enforces comprehensive identity verification (KYC). As a result, the product format that succeeded massively in international markets is not the same as the version now competing for participation in the U.S. market—while U.S.-native channels and capital rails have long been controlled by existing players.

This setup will very likely lead to a split in role specialization: distribution platforms control users and order flow, exchanges provide infrastructure and compliance services, and native crypto platforms lead global market access. Prediction markets will keep growing across these three tiers, but value capture will concentrate in the segments closest to users.

Winners and losers

The emergence of this system reshapes the winner hierarchy.

Winner: distribution platforms

Platforms such as Robinhood and Coinbase directly reach end users, controlling account opening, capital, and order flow. As they vertically integrate exchange infrastructure, they will capture both trading volume and profit margins. They are no longer just participants in prediction markets; they are becoming core entry points.

Survival but value shrinking: infrastructure providers

Kalshi has the best chance to become the regulated core pillar within the system, responsible for contract authorization, transaction clearing, and compliance infrastructure. But its role will shift from facing users directly to servicing user holders; its business model will be closer to Visa than to a retail bank.

Maintain differentiation: global-native crypto platforms

Polymarket still retains an advantage in markets outside the U.S. In those regions, permissionless access, instant settlement, and broader market coverage matter more than regulatory compliance. It will become the default trading venue for global users, high-controversy markets, and native crypto capital.

Losers: participants lacking channels and liquidity

As probabilities backed by capital replace viewpoint-based forecasts, the importance of traditional polling organizations continues to decline. Small prediction market platforms that lack liquidity and are embedded in channels will struggle to compete. Sports bookmakers face structural pressure; prediction markets overlap increasingly with their core business. Estimates suggest this has already caused about $600 million in tax revenue losses.

Final form

Prediction markets expand the pricing boundary of financial markets. Today, financial markets price assets: stocks represent ownership of a company, and bonds represent claims on future cash flows. Prediction markets, in contrast, price the outcomes of events.

Anything uncertain can become a tradable underlying: political events, economic indicators, technological progress, environmental outcomes. Ultimately, uncertainty itself is financialized.

Data clearly shows the industry’s development trajectory:

• 2024: total trading volume of about $16.0 billion—validated by elections, proving the business model’s viability

• 2025: total trading volume of about $63.5 billion—growing 4x driven by sports markets

• 2026: annualized scale exceeding $200.0 billion—regulatory clarity will become the key to the industry’s survival or collapse (Data source: Artemis)

Conclusion

Prediction markets signal a shift in how the world processes uncertainty. They will transform the otherwise intangible thing—probabilities of future events—into tradable, priceable assets that can inform action.

The core question is whether the element that makes prediction markets valuable (pricing uncertainty) aligns with the forces that drive their scaling. Right now, the answer is no. Scale is driven by sports markets, while information value gives it its real meaning. The gap between the two will determine the future of the entire industry. Once scaled, it won’t just be a new market—it will be a new model for pricing information itself.