Testing the Waters of Prediction Market ETFs: A Step Toward Mainstream or Playing with Fire?

Original | Odaily (@OdailyChina)

Author | DingDang (@XiaMiPP)

Recently, ETF issuers Bitwise Asset Management and GraniteShares have filed applications with the U.S. Securities and Exchange Commission (SEC) for prediction market ETFs. Among them, Bitwise submitted six products under the "PredictionShares" brand, and GraniteShares followed closely with a structurally similar proposal. Earlier, on February 13th, Roundhill Investments also filed a similar application.

The core of these ETFs is tracking the outcomes of U.S. political elections. They attempt to package the "probability of outcomes" of U.S. political elections into a financial product that can be directly traded within traditional securities accounts. Specifically, the underlying focus is on the 2028 presidential election (Democratic or Republican victory) and the control of the Senate and House of Representatives in the 2026 midterm elections.

In other words, investors in the future may no longer need to go to crypto-native platforms like Polymarket or register with CFTC-regulated Kalshi. They could simply open a Robinhood or Fidelity account and bet on "who will win the White House" as easily as buying a stock.

Screenshot from @jason_chen998

What does this leap forward signify?

Why Are Prediction Markets Always "One Step Ahead"?

The "forward-looking" nature of prediction markets regarding political events is hardly new.

Prediction markets are essentially groups of people using real money to express judgments. Participants express their confidence in an event's occurrence by buying and selling "Yes/No" contracts. The prices of these contracts fluctuate between $0 and $1, representing the market's consensus on probability. For example, if you believe a candidate has a 70% chance of winning, you might buy a "Yes" contract for $0.70. If the event occurs, the contract's value rises to $1; otherwise, it becomes worthless.

This is a form of capital-weighted collective judgment. Unlike mere verbal expression, participants must bear the profit and loss consequences of their judgments, which was vividly demonstrated during the 2024 U.S. election. Trading volumes on Polymarket and Kalshi surged rapidly, with political contracts becoming the absolute mainstay. Before Election Day, the cumulative trading volume on Polymarket for the single market "2024 Presidential Election Winner" was approximately $3.7 billion. Kalshi, a later entrant, won a key lawsuit against the CFTC in September 2024, allowing it to legally offer election-related contracts. By November, its monthly trading volume reached $127 million, with about 89% coming from political and election markets.

More noteworthy is the signal conveyed by the data itself. Weeks before the 2024 election, the probability of a Trump victory on Polymarket remained stable above 60%, while mainstream polls at the time showed a tight race, with Harris even slightly ahead. The result? The prediction market seemed to have "read" the electoral situation in advance.

This doesn't mean prediction markets are "magically accurate," but across multiple election cycles, they have indeed demonstrated strong information aggregation capabilities. Research suggests that with sufficient liquidity and broad participation, the statistical performance of prediction markets often surpasses that of traditional poll samples. The veteran platform PredictIt has also been regarded as an effective information aggregator on multiple occasions. In contrast, traditional polls are susceptible to factors like sample bias and expression bias.

The root of the difference lies in the incentive mechanism: polls express attitudes, while prediction markets bear the outcome. The former has no cost, the latter has clear profit and loss. This structural difference determines the different ways information is processed.

Although prediction markets cooled down after the election—Polymarket's daily trading volume plummeted by about 84% after the results were announced—the number of prediction market projects grew rapidly in 2025. As of now in 2026, according to data from predictionindex.xyz, there are as many as 137 prediction market projects. The leading player, Polymarket, has a total trading volume exceeding $50 billion, with a monthly trading volume of $8 billion.

From a fringe experiment to a mainstream sector, prediction markets are now a different beast. Now, imagine if participation could be made easy through ETFs—this collective intelligence could more broadly influence public perception of political events.

How ETFs Package Prediction Markets

So, how do these ETFs bring the mechanics of prediction markets to Wall Street?

What these issuers are essentially doing is translating the contract prices of prediction markets into a product structure understandable by the securities market. Dressed in the guise of an ETF, it allows you to buy and sell through a regular brokerage account, yet you're still betting on the outcome of a political event.

Taking Bitwise's six proposed ETFs as an example, four directly target the 2028 presidential election (Democrat/Republican winner), and the remaining two correspond to control of the Senate and House in the 2026 midterms. The structures of GraniteShares and Roundhill are largely similar. In simple terms, these ETFs directly map the price performance of those binary event contracts on platforms like Kalshi or Polymarket into tradable ETF shares.

Mechanically, the share price of these ETFs will fluctuate between $0 and $1, reflecting the market's real-time consensus on event probability. At least 80% of the fund's assets will be invested in derivatives linked to these political events, such as contracts obtained from CFTC-approved exchanges like Kalshi, or through synthetic swaps to replicate performance. The buying process is the same as buying a stock: through brokerage accounts like Robinhood or Fidelity, with an expected expense ratio between 0.5% and 1%, and likely trading on NYSE Arca.

Upon settlement, if the event occurs (e.g., a Democrat wins the presidency), the corresponding "Yes" ETF's value approaches $1; otherwise, it approaches $0. Bitwise's plan is for the fund to liquidate and terminate shortly after the event outcome is determined, distributing the remaining assets pro-rata to holders. Some products from GraniteShares and Roundhill are more "flexible," potentially allowing a "roll" into the next election cycle.

Compared to the Bitcoin ETFs we are familiar with, there is a clear distinction. Bitcoin ETFs, like BlackRock's IBIT, track the price of Bitcoin, with unlimited upside or downside potential, making them suitable as part of an asset allocation. Prediction market ETFs, however, lean more towards binary probability bets, with a fixed upper limit of $1, similar to buying insurance or options—winner takes all, loser loses all.

The question is, when probability becomes a tradable asset, does it remain a pure information aggregation mechanism?

Mainstreaming, or Gamblification?

If these ETFs are approved, prediction markets will truly enter the mainstream financial view.

Currently, political prediction markets are still concentrated among crypto users or professional traders. Once ETFs are listed, the participation barrier for institutional capital and traditional investors will be significantly lowered. Companies might use them to hedge against policy change risks, and portfolio managers might view them as macro risk management tools. Liquidity will be amplified, and price signals might become more acute.

But the issues on the other side are equally apparent. The 2024 election already proved that prediction market prices are cited by media, amplified on social platforms, and even influence public sentiment. When probability is packaged as "market consensus," it is easily interpreted as an objective trend. If the scale of capital further expands, could there be deliberate price manipulation to influence public opinion? PredictIt was embroiled in legal disputes over compliance issues in its early years; such problems are not unfounded.

Regulation remains the biggest uncertainty. The SEC might worry that this is essentially the "gamblification" of finance, increasing the risk of manipulation or moral hazard. The approval process might come with conditions, such as trading limits or additional disclosures. Currently, the CFTC has allowed Kalshi to trade election futures, which is a positive signal, but the SEC's stance remains unclear.

Conclusion

From crypto-native markets to Wall Street ETFs, prediction markets are undergoing an identity transformation. However, before the regulatory framework is clear, the moves by issuers seem more like a probe—testing regulatory boundaries and the market's acceptance of "probability as an asset."