The first wave of filings can be sorted into two categorically different products: ETFs that hold direct economic exposure to event-contract payoffs via swaps, and ETFs that hold equity in the operating businesses that run prediction markets and adjacent infrastructure.

Direct-exposure funds (Roundhill, Bitwise, GraniteShares)

Roundhill's slate is the most concrete and the most aggressive. Six funds, three pairs across the three federal chambers, with full ticker disclosure in their April 29, 2026 prospectus:

• BLUP — Roundhill Democratic President ETF (2028 race)
• REDP — Roundhill Republican President ETF (2028 race)
• BLUS — Roundhill Democratic Senate ETF (Nov 3, 2026 midterm)
• REDS — Roundhill Republican Senate ETF (Nov 3, 2026 midterm)
• BLUH — Roundhill Democratic House ETF (Nov 3, 2026 midterm)
• REDH — Roundhill Republican House ETF (Nov 3, 2026 midterm)

Each fund gains exposure through total-return swaps that reference a single binary event contract on a CFTC-designated contract market (primarily Kalshi). The prospectus is blunt about the asymmetry: in all-capital letters it warns that if the targeted party does not win the targeted election, the fund will lose substantially all of its value. There is no diversification within a fund — each fund is a vehicle to express conviction on one outcome.

Bitwise Asset Management (sponsor: Bitwise Investment Advisers, LLC) filed a similar slate of event-contract-linked funds in February 2026. GraniteShares ETF Trust filed a parallel set in March under File Nos. 333-214796 (1933 Act) and 811-23214 (1940 Act). Specific ticker disclosures across the Bitwise and GraniteShares filings have been more limited than Roundhill's; collectively the three sponsors account for the 24 funds the SEC put on pause in May.

Operator-equity funds (Tema)

Tema ETF Trust's filing is structurally orthogonal. The Tema Trading & Prediction Markets ETF is an actively managed equity fund that commits at least 80% of net assets to companies operating in the prediction-market and trading ecosystem: financial exchanges, event-contract platforms, regulated prediction markets, crypto trading venues, and market-data providers. It does not hold event contracts. It holds the firms that list, route, clear, and provide infrastructure for event contracts.

Tema is the closest analog to a thematic equity ETF — the way one might own a payments ETF rather than payment volume. Volatility profile, fee structure, and return drivers all look more like traditional equities than anything binary.

SEC posture

On May 5, 2026 the SEC delayed launches across all 24 direct-exposure ETFs, opening a public-comment window and requesting additional disclosure on product mechanics. Industry counsel and ETF analysts have largely framed the pause as a Bitcoin-ETF-style review period rather than a rejection. The comparison is apt: it took roughly a decade for the SEC to approve a spot Bitcoin ETF after the first filing, with the substantive concerns centered on market-manipulation surveillance and underlying-market integrity. For event contracts the surveillance picture is cleaner (Kalshi and Polymarket are CFTC-regulated DCMs with full order-book transparency), so the timeline is likely to compress. The dominant near-term question is whether disclosure language adequately conveys the all-or-nothing payoff to retail.

A 1940 Act open-end fund cannot directly hold event contracts. The Investment Company Act constrains what counts as a permissible investment, and CFTC-regulated event contracts — even though they're "swaps" under the Commodity Exchange Act post the Third Circuit's April 2026 Kalshi ruling — don't fit cleanly into the 40 Act's eligibility regime. The structural workaround is a total return swap (TRS) with a creditworthy counterparty:

• The ETF enters a swap that references the price of a specific event contract (or a small basket).
• The ETF pays the counterparty a financing rate — typically SOFR plus a spread of roughly 50 to 200 basis points, depending on the counterparty's hedging cost and the contract's liquidity.
• The ETF receives the total return on the referenced event contract: if the event resolves YES, the fund receives $1 per $1 of notional referenced; if it resolves NO, it receives $0. Between today and resolution, the swap marks to the event contract's market price.
• The counterparty — generally a market-maker or bank — holds the underlying event contracts on Kalshi or Polymarket and hedges its book. The ETF never holds the contracts directly.
• The ETF posts cash or T-bills as collateral. Because the swap has no built-in leverage, the collateral approximates the swap notional. Unused cash earns the T-bill rate, which materially offsets the financing leg in the current rate environment.

The economic result is direct exposure to a binary outcome, wrapped in a structure that is 1940 Act compliant. The cost to the end investor is the financing spread, fund expenses, and counterparty credit risk — none of which exist if you trade the underlying event contract directly on Kalshi. The benefit is qualification for tax-advantaged accounts, intraday liquidity, and access through standard brokerage rails.

Margin and economic exposure

Importantly, none of these ETFs are leveraged at the fund level. A TRS referencing $100M of event-contract notional requires the ETF to post approximately $100M in collateral. The counterparty, on its book, may run tighter margin against the underlying Kalshi position (Kalshi requires full pre-funding, so 100% collateralization in that direction) but also against its hedge book. The financing spread compensates for that hedging cost and for warehousing event-resolution timing risk.

For an ETF investor, the take-away is that one share of BLUP or REDP corresponds to a fixed dollar amount of binary payout exposure. There is no margin to call, no liquidation cascade — only a path-dependent NAV that tracks the event-contract price between today and resolution.

The expected return on a direct-exposure event-contract ETF is mechanical. Buy at price p, and the contract resolves to either $1 or $0. The expected payoff is exactly p dollars per contract, which is what you paid. Net expected return is zero, before fees and counterparty spread. Realized return is binary — either a multiple of roughly (1/p − 1) if you win or a total loss if you don't.

For BLUP at, say, $0.45 (the contract pricing a 45% probability of a Democratic 2028 win): a winner returns roughly 122% on the position. A loser returns -100%. That asymmetry is the entire product. Sharpe ratio in the conventional sense doesn't really apply — the realized distribution is two-point with no time-series variance to speak of (just a single binary shock at resolution). What investors actually experience between purchase and resolution is the NAV path tracking the underlying event-contract market price, which can move sharply with polls, debates, or political news.

Three things make this category interesting despite the zero-EV mechanics:

• Information asymmetry. Sophisticated investors who can forecast outcomes better than the prediction-market consensus have positive expected return at the margin. Whether enough of those investors exist to make the strategy scalable is a separate question.
• Tax wrapping and access. Owning binary exposure in a tax-advantaged account, or expressing a directional view through an accounted-for ETF, has structural utility separate from the EV of the underlying. This is most of the use case for early adopters.
• Hedging utility. An event-contract ETF that pays off in a tail scenario for a correlated portfolio can be sized as an overlay, with the EV mechanics partly offset by the marginal utility of cash in the bad state. This is the framework Section 5 develops.

The return-profile section sets up a more structural question: even granting that the SEC eventually clears the binary direct-exposure cohort, is the ETF wrapper actually the right container for these products? The ETF format carries a set of assumptions inherited from its equity-basket origins — intraday-liquid underlyings, in-kind create/redeem mechanics, tax efficiency through equity transfer, and diversification as the default investor benefit. The direct-exposure prediction-market ETFs satisfy none of these.

Where the wrapper fights the binary product

• No diversification by design. Each Roundhill fund files non-diversified status because a single-event ETF cannot, definitionally, diversify. The 1940 Act permits this, but the wrapper's most familiar investor benefit is forfeit before the trade is on.
• The tax efficiency story doesn't carry over. ETFs derive most of their tax advantage from in-kind create/redeem on equity baskets. A swap-only fund pays ordinary income on the financing leg, and binary resolution creates a discrete taxable event for the fund and a pass-through for shareholders. The wrapper's most quantitative selling point is muted.
• Intraday NAV references a thin order book. Equity-index ETFs work because the underlying basket is continuously priced. An event-contract ETF's intraday NAV tracks a thin Kalshi or Polymarket order book whose marginal liquidity is news-driven, not fundamentals-driven. Intraday spreads on the ETF can dislocate sharply from the swap's reference price during news events — precisely when retail will be most active.
• Capacity is hard-capped by event-contract open interest. A fund that gathers $100M of AUM may quickly become the dominant participant in its own reference market. This is not a problem equity ETFs ever have to solve.
• The -100% binary case lives in IRAs and 401(k)s. The wrapper's most common downstream allocation is a tax-advantaged retirement account, optimized for long-horizon compounding. A product whose worst-case is total loss in a 12-24 month window sits poorly inside that container, and the SEC's pause likely reflects this concern more than any other.

None of these are fatal. Single-name leveraged ETFs exist; non-diversified status is legitimate; disclosure language can address risk profile. The argument isn't that the binary funds shouldn't exist — it's that the wrapper is doing meaningful structural work against the product while the product borrows the wrapper's credibility without inheriting its strengths.

Where signal and overlay structures fit cleanly

Both alternative concepts hold conventional securities as their core. The 1940 Act constraints that necessitate the TRS workaround don't apply. In-kind create/redeem and equity-basket tax efficiency are fully available. Diversification is structural — a signal ETF rebalances a diversified basket, an overlay ETF dedicates 95-99% of NAV to a conventional core with the prediction-market exposure as a small sleeve. Capacity scales with the underlying equity or fixed-income market, not with a single event contract's depth. The retail risk profile is bounded — overlay sleeves typically cap catastrophic loss at a few percent of NAV, not 100%.

The category pattern is familiar from adjacent ETF segments. Managed futures, tail-risk hedge, and volatility-targeting ETFs all use small sleeves of derivatives to modify a conventional portfolio. The overlay concept slots into that taxonomy directly — sponsors who already run tail-risk products understand both the wrapper mechanics and the advisor narrative. The signal concept slots even more cleanly into the rules-based and quantitative ETF category that comprises a meaningful fraction of asset-management AUM. The conceptual lift for advisors and allocators is minimal because the wrapper is doing what it's designed to do.

The first wave of filings reflects what sponsors could ship quickly inside an existing regulatory framework, not what the optimal product looks like. The natural shape of prediction-market ETFs — the version that scales, fits retirement-account use cases, and matches investor expectations of the wrapper — looks more like a signal or overlay structure than a binary-bet vehicle. The next two sections sketch what each would look like.

None of the registered products solve a problem that the asset-management industry actually has — namely, how to translate the information content of prediction markets into systematic exposure on traditional financial assets. A predictive signal ETF would hold a conventional equity or fixed-income basket and dynamically rebalance based on prediction-market state.

The intellectual ancestry here is sturdy. Snowberg, Wolfers and Zitzewitz (2006) showed that equity, bond, and FX markets at intraday frequency reflected anticipated partisan policy impact during the 2004 US election — using prediction-market price moves as a real-time partisanship instrument. Subsequent literature has consistently found that prediction markets incorporate new information faster and with lower forecast error than polls or professional forecasters. The natural question is whether that information signal can be productized.

Three implementation flavors

• Macro-overlay basket. The ETF holds a balanced 60/40-style basket. When Kalshi's "US recession within 12 months" probability crosses a threshold (say 50%), the fund shifts duration longer and underweights equities. The signal is the prediction-market level; the trade is in underlying liquid assets.
• Sector-rotation by political probability. The fund holds a sector overlay that tilts to defense, energy, and traditional finance when Republican congressional control probability exceeds polling consensus, and to renewables, healthcare, and tech when Democratic control probability exceeds consensus. The signal is the divergence between prediction markets and polls — using the market as the de-noising estimator.
• Event-driven equity sleeve. The fund maintains long positions in single names whose earnings or product-cycle outcomes have corresponding event contracts on Polymarket or Kalshi. The fund weight scales with the prediction-market implied probability of the favorable outcome, less a threshold for transaction cost. Closest spiritual analog: merger-arb ETFs.

None of these structures hold event contracts. The 1940 Act constraints that make the direct-exposure funds need a TRS workaround don't apply — these funds hold conventional securities, with prediction markets only informing the trading rule. The regulatory bar is materially lower; the data infrastructure (real-time prediction-market feeds, calibrated trading signals) is the harder part of the engineering.

For an issuer interested in differentiated active management with a clean quantitative thesis, this is the more interesting white space than packaging binary outcomes for retail.

The overlay concept sits between the direct-exposure ETFs and the signal ETFs. An overlay product would hold a base portfolio (equity, fixed income, or balanced) and apply event-contract positions as a small-allocation overlay sized to either hedge specific tail risks or to generate return enhancement from prediction-market mispricings.

The product structure could be a 1940 Act ETF using the same TRS plumbing as the direct-exposure funds, but with the overlay typically constrained to 1-5% of NAV. The remainder is held in conventional assets. Two sub-strategies define the use case:

5.1 Hedge overlay — using prediction markets as tail insurance

Equity portfolios are exposed to a small set of large tail scenarios: recession, geopolitical escalation, policy shocks, sector-specific regulatory events. Each of these has a corresponding event contract on Kalshi or Polymarket. The hedging logic is straightforward — buy YES on the event contract sized so that its payoff in the bad state offsets the portfolio's expected drawdown.

The sizing problem is the interesting part. Assuming a long-only portfolio with conditional drawdown d in the event scenario, expected drift μ outside the event scenario, market-implied event probability p, and target hedge ratio h:

• Notional event-contract exposure needed: N = AUM × d × h
• Upfront cost: C = N × p + fees
• Steady-state drag in the no-event scenario: −C / AUM per horizon
• Net PnL in event scenario: −AUM × d + N − C = N(1 − p) − AUM × d

At market-implied probability and a hedge ratio of 1.0, the expected value of the overlay is zero. Three sources of edge remain: (a) your conditional drawdown estimate exceeds market-implied (you're more bearish on the joint distribution than consensus); (b) the event itself is mispriced relative to a fundamental probability you trust; (c) the marginal utility of cash in the bad state exceeds linear EV (the classic risk-aversion argument for buying insurance).

5.2 Return enhancement — Kelly sizing on probability divergence

The return-enhancement variant uses prediction-market mispricings as a generalized statistical-arbitrage signal. The fund maintains a fundamental probability estimate for each tracked event — derived from polls, related financial markets, models, or analyst consensus — and takes a position when the prediction-market price diverges from that estimate.

Position sizing follows fractional Kelly. For a binary contract bought at YES price p_y when your estimated win probability is p_f:

• Payoff per $1 staked: b = (1/py) − 1
• Full Kelly fraction: f* = (b · pf − (1 − pf)) / b
• Recommended deployment: quarter-Kelly (25% of f*) or less, capped at a small percentage of capital per position

Quarter-Kelly is the industry-standard deployment fraction for production allocators. Full-Kelly is mathematically optimal for long-run geometric growth but unforgivingly path-volatile when probability estimates drift — most production books run something between 10% and 30% of full Kelly, paired with hard position caps to control concentration.

Critically, the operative metric is net edge after fees, not raw probability divergence. A 5-point edge at 100 bps round-trip cost is structurally different from a 5-point edge at 30 bps. Kalshi and Polymarket fee schedules differ materially; production overlay strategies need explicit fee accounting in the position-sizing logic.

Two modes. Hedge overlay sizes an event-contract overlay to offset a conditional portfolio drawdown, with steady-state carry drag and full scenario PnL breakdown. Return enhancement takes a market YES price and a fundamental probability estimate and produces a Kelly-sized position with net edge after fees. Both modes ship with preset scenarios you can adjust.

Three questions decide whether prediction-market ETFs become a durable category rather than a regulatory curiosity:

• Does the SEC clear the direct-exposure cohort, and on what disclosure terms? If the disclosure language becomes too restrictive — explicit gambling-style warnings, prohibition from tax-advantaged accounts — the structural utility collapses and the product is just a worse way to access Kalshi. If the disclosure converges on the same risk-factor language as leveraged or single-stock ETFs, the category likely scales.
• Do prediction-market liquidity profiles support institutional size? Kalshi has scaled significantly through 2025 and 2026, and Polymarket's CFTC-regulated re-launch has materially deepened US liquidity. Roundhill's 2026 midterm products (BLUS, REDS, BLUH, REDH) are the first stress test of whether a fund-size NAV can be supported by prediction-market depth.
• Do signal and overlay structures emerge? The most interesting product economics aren't in the direct-exposure funds — they're in vehicles that productize prediction-market information without constraining themselves to binary payoffs. That requires sponsors with quant infrastructure rather than thematic-marketing infrastructure. The field is currently open.

On the regulatory side, the most consequential ruling of 2026 has already happened. The Third Circuit's April 6 decision in Kalshi v. New Jersey held that Kalshi's sports event contracts are "swaps" under the Commodity Exchange Act, which substantially reinforces CFTC primary jurisdiction over the underlying market. That clarification removes a meaningful state-law uncertainty that had been dragging on the sponsor outlook, and helps explain the SEC's willingness to engage substantively rather than reject outright.

The narrower the disclosure conditions and the deeper the underlying market liquidity, the more attractive the structural play becomes. We'll update this piece as the SEC's response window closes and the first wave of approvals or denials lands.