The Road to the Spot Bitcoin ETF

The 2017 Denial, and Its Arithmetic

On March 10, 2017 the SEC's Division of Trading and Markets, acting under delegated authority, disapproved the BZX proposal — Release No. 34-80206, File No. SR-BatsBZX-2016-30.

The stated ground was Section 6(b)(5) of the Exchange Act, which requires an exchange's rules to be designed to prevent fraudulent and manipulative acts and practices and to protect investors and the public interest. The order set out two requirements: the listing exchange must have surveillance-sharing agreements with significant markets for trading the underlying commodity or its derivatives, and those markets must be regulated. (Release No. 34-80206, at 2)

The order's numbers explain what “significant” meant in practice. In the six months preceding February 28, 2017, trading on the Gemini Exchange accounted for 0.07% of worldwide Bitcoin trading and 5.16% of the worldwide Bitcoin–US dollar market. Gemini's daily auction — the pricing mechanism for the proposed trust — had averaged 1,195.72 Bitcoin per day since its inception on September 21, 2016, with a business-day median of 1,061.99, against a proposed creation basket of 1,000 Bitcoin. On 88.2% of business days, auction volume was under two baskets. (Release No. 34-80206, at 32–34)

One methodological note belongs in the record. The Commission acknowledged in a footnote that because Bitcoin trading was dispersed across largely unregulated markets, there was no centralised regulatory data source for Bitcoin trading statistics, and that its own analysis therefore relied on unofficial sources. (Release No. 34-80206, n.127) The agency applying the standard was working from the same kind of data it was declining to trust.

The order also stated the terms on which the answer could change: Bitcoin was in the relatively early stages of its development, regulated Bitcoin-related markets of significant size might develop over time, and if they did, the Commission could reconsider. (Release No. 34-80206, at 38)

2018: The Test Gets a Name — and an Escape Hatch

The usual shorthand is that “the SEC denied the Winklevoss ETF in 2017.” The procedural history is more specific. BZX petitioned for Commission review. On July 26, 2018 the Commission set aside the delegated action, reviewed the proposal itself, and disapproved it again — Release No. 34-83723, published at 83 Fed. Reg. 37579. (Release No. 34-83723)

Two things in that order shaped everything after it.

First, it supplied the operative definition of a “market of significant size,” later restated as a two-prong test: a market (or group of markets) as to which (a) there is a reasonable likelihood that a person attempting to manipulate the ETP would also have to trade on that market to succeed, and (b) it is unlikely that trading in the ETP would be the predominant influence on prices in that market. (Winklevoss Order at 37594, cited at Release No. 34-99306, n.23)

Second — and this is the part usually left out — the order said the surveillance-sharing agreement was not the only route. A listing exchange could instead demonstrate that “other means to prevent fraudulent and manipulative acts and practices will be sufficient.” (Winklevoss Order at 37580, cited at Release No. 34-99306, nn.25–26)

That sentence sat unused for five and a half years and then became the legal basis for approval.

The Pattern, 2019–2023

The denials that followed applied the same framework:

• Bitwise (2019). NYSE Arca filed on January 28, 2019. The Division of Trading and Markets disapproved on October 9, 2019 under delegated authority (Release No. 34-87267, 84 Fed. Reg. 55382), and the Commission took that delegated action up for review on October 15, 2019. The standard account stops there and counts this filing among the rejections. The Commission's own records say otherwise: NYSE Arca withdrew the proposed rule change on January 13, 2020, while the review was still pending (withdrawal letter), and on November 13, 2020 the Commission formally set aside the delegated disapproval order, citing that withdrawal (Release No. 34-90431, 85 Fed. Reg. 73819). The 2019 Bitwise proposal was never finally disapproved by the Commission.
• United States Bitcoin and Treasury Investment Trust (2020). Release No. 34-88284, 85 Fed. Reg. 12595.
• NYDIG (2022), 87 Fed. Reg. 14932; One River (2022), 87 Fed. Reg. 33548; Bitwise (2022), 87 Fed. Reg. 40282.
• Grayscale (2022). 87 Fed. Reg. 40299 (July 6, 2022).
• VanEck (2023). Release No. 34-97102 (March 10, 2023), 88 Fed. Reg. 16055.

Chair Gary Gensler's own count, given on the day of approval: beginning under Chair Jay Clayton in 2018 and through March 2023, the Commission disapproved more than 20 exchange rule filings for spot Bitcoin ETPs. (Gensler statement, Jan. 10, 2024)

The reasoning did not meaningfully vary. Each order found that the proposing exchange had not carried its burden of showing consistency with Section 6(b)(5), and each found the significant-market test unsatisfied.

The Futures Asymmetry

In 2022 the Commission approved two exchange-traded products holding CME Bitcoin futures: the Teucrium Bitcoin Futures Fund on April 12, 2022 (87 Fed. Reg. 21676) and the Valkyrie XBTO Bitcoin Futures Fund on May 11, 2022 (87 Fed. Reg. 28848). For both, the listing exchange's surveillance-sharing agreement with the CME satisfied the significant-market test.

In both orders the Commission stated explicitly that approving futures ETPs did not mean approval of spot ETPs was imminent. (summarised in Grayscale v. SEC, slip op. at 6)

The same surveillance-sharing agreement with the same exchange was producing opposite outcomes. That asymmetry is what went to court.

Grayscale v. SEC: What the Court Held, and What It Did Not

Grayscale petitioned for review on June 29, 2022. The case was argued on March 7, 2023 and decided on August 29, 2023, No. 22-1142, before Chief Judge Srinivasan, Circuit Judge Rao, and Senior Circuit Judge Edwards. Judge Rao wrote for the court. (slip op.)

What the court held. The denial was arbitrary and capricious under the Administrative Procedure Act because the Commission failed to explain its different treatment of similar products. The court granted the petition and vacated the order.

The reasoning ran in two steps. Grayscale had shown its proposed product was materially similar to the approved futures ETPs across the relevant regulatory factors: the underlying assets were closely correlated, and the listing exchanges had identical surveillance-sharing agreements with the CME. And the Commission had not given a reasonable and coherent explanation for reaching opposite results.

Findings worth quoting precisely, because they are frequently garbled:

• Grayscale presented uncontested evidence of a 99.9% correlation between Bitcoin spot prices and CME futures prices. (slip op. at 10)
• The Commission “neither disputed” that evidence “nor suggested that market inefficiencies or other factors would undermine the correlation.” (slip op. at 14)
• The Commission had treated evidence of a lead/lag relationship as unnecessary when approving the futures ETPs but central when rejecting Grayscale's, and offered no reason for the difference. (slip op. at 16)
• Grayscale held 3.4% of outstanding Bitcoin, and its shares traded at a discount of as much as 30%, which Grayscale estimated left over $4 billion on the table for its investors. (slip op. at 6, 18)

What the court did not hold. It did not hold that the SEC must approve spot Bitcoin ETPs. It did not hold that Bitcoin markets are adequately regulated. And it expressly declined to reach Grayscale's argument that the significant-market test is contrary to law, stating in a footnote: “We do not reach this argument because we set aside the SEC's order as arbitrary and capricious.” (slip op. at 8 n.2)

The holding is narrow and procedural: an agency that articulates standards case by case must justify different results on similar facts. The mandate issued on October 23, 2023.

January 10, 2024: The “Other Means” Approval

Release No. 34-99306 approved eleven proposals across three exchanges on an accelerated basis: Grayscale Bitcoin Trust, Bitwise Bitcoin ETF and Hashdex Bitcoin ETF on NYSE Arca; iShares Bitcoin Trust and Valkyrie Bitcoin Fund on Nasdaq; ARK 21Shares Bitcoin ETF, Invesco Galaxy Bitcoin ETF, VanEck Bitcoin Trust, WisdomTree Bitcoin Fund, Fidelity Wise Origin Bitcoin Fund and Franklin Bitcoin ETF on Cboe BZX.

The most misreported fact about this order is what it did not find. The Commission did not conclude that the CME Bitcoin futures market is a market of significant size related to spot Bitcoin. In its own words, the surveillance-sharing agreement was with “a U.S. regulated market whose bitcoin futures market is consistently highly correlated to spot bitcoin, albeit not of ‘significant size’ related to spot bitcoin.” The approval came through the “other means” pathway opened in 2018: the Commission found “that sufficient ‘other means’ of preventing fraud and manipulation in this context have been demonstrated.” (Release No. 34-99306)

What changed was the quality of the evidence. The Commission began with a correlation analysis submitted in the ARK filing, covering January 20, 2021 to February 1, 2023, which reported correlations of no less than 92% hourly and 78% minute-by-minute — but which, the Commission noted, did not assess whether those results held consistently across the sample period.

So the Commission ran its own. Using stationary price-return series at hourly, five-minute and one-minute intervals for spot BTC/USD on Coinbase and Kraken against the closest-to-maturity CME Bitcoin futures contract, from March 1, 2021 to October 20, 2023, with data from the CME via the SEC's Market Information Data Analytics System and from Kaiko:

On that basis the Commission concluded that fraud or manipulation affecting spot Bitcoin prices would likely similarly affect CME futures prices, so the Exchanges' agreement with the CME could reasonably be expected to assist in surveillance. It added a caution in a footnote: correlation should not be interpreted as an indicator of a causal relationship or of which variable leads or lags. (n.39)

The 99.9% Figure Has a Second Act

The correlation number the D.C. Circuit found decisive — 99.9%, uncontested — did not survive the Commission's own scrutiny once it got around to performing it.

In the approval order, the Commission identifies that figure as coming from a comment letter filed by Coinbase in the Grayscale proceeding, covering November 4, 2021 to February 23, 2022. It then dismantles the methodology: the correlation appears to have been calculated on time series of price levels, which “are often non-stationary, which leads to results that indicate relationships that do not actually exist”; using only daily observations “provides no information on how prices in the two markets are associated—if at all—throughout the trading day”; and a single four-month sample “does not provide evidence of a consistently high correlation over time.” (n.30)

The number that won the case was, by the agency's own later account, not sound. The court's point had never been that the figure was right — only that the Commission had not disputed it. When the Commission finally did the analysis properly, it reached the same conclusion by better means.

Two further details from the order matter for anyone reading fund documents. The proposals contemplated cash creation and redemption only; in-kind processes were expressly “outside the scope of this order.” (n.77) And proposed bilateral surveillance-sharing agreements with Coinbase had been removed from the amended filings and “are not a basis for approval.” (n.41)

Commissioner Caroline Crenshaw dissented. Trading began the next day; Grayscale's trust uplisted from OTCQX to NYSE Arca on January 11, 2024. Whether these approvals changed Bitcoin's long-run price behaviour is a separate and contested question — examined in our review of the Bitcoin power-law debate.

The Fee War

Competition among the eleven was immediate and expressed almost entirely through fees and waivers.

BlackRock's is the clearest documented example. The iShares Bitcoin Trust's sponsor fee accrues daily at an annualised 0.25% of net asset value. For twelve months from January 11, 2024, the sponsor waived a portion so the effective fee was 0.12% on the first $5.0 billion of trust assets. For the period ended December 31, 2024 the sponsor's fee was $53,457,094 and the amount waived was $5,958,429. (IBIT Form 10-K FY2024, Note 3)

The direction of travel is visible in the newest entrant's filing: 0.14% for the Morgan Stanley product launched in April 2026, against 0.25% for the largest incumbent. Fee compression in this category has been real and one-directional.

What a fee difference does to BTC-per-share over a holding period is mechanical rather than mysterious — our fee-drag calculator computes it for every fund, and our Bitcoin ETF comparison tracks the current fee table across the category.

GBTC: The Outlier Fee

The Grayscale Bitcoin Trust entered the ETF era with by far the highest fee in the group and did not cut it.

Before conversion, the trust paid its sponsor 2.0% of the aggregate value of trust assets less liabilities. (GBTC Form 10-Q, Q2 2020) After conversion the figure is 1.5%, calculated the same way, accruing daily in US dollars and payable in Bitcoin daily in arrears. (GBTC Form 10-Q, June 30, 2026)

The decline in the trust's position is documented in its own filings. Holdings are audited; flow estimates generally are not, so the holdings are given here instead:

Audited figures quoted from the filing linked on each row; the 2022 row is as recited by the D.C. Circuit.

Grayscale's competitive response was structural rather than a fee cut. On July 31, 2024 the trust completed a pro rata distribution of 303,690,100 shares of the Grayscale Bitcoin Mini Trust ETF to holders of record as of July 30, 2024, contributing to that trust Bitcoin equal to approximately 10% of GBTC's total holdings as of the record date. (GBTC Form 10-Q) The result was a low-fee Grayscale product alongside, rather than instead of, the 1.5% one. Whether a GBTC holder should switch — and the tax math that decides it — is worked through in our GBTC vs Mini Trust comparison.

One structural fact from the filings bears on why an investor might hold rather than switch, though it should not be asserted as the reason: assuming grantor-trust treatment, each sale or delivery of Bitcoin by the trust is a taxable event for shareholders, and the trust has a pre-ETF operating history stretching to its formation on September 13, 2013. No Grayscale filing states a rationale for holding the fee at 1.5%. The explanations in circulation are attributed to analysts, not to the sponsor.

One measure improved sharply on conversion. From the January 11, 2024 uplisting through December 31, 2025, the maximum premium of GBTC's closing price over NAV per share was 1.68% and the maximum discount 1.56%, with averages of 0.06% and 0.08%. (GBTC Form 10-K FY2025) The 30% discount the D.C. Circuit described is gone. The creation-and-redemption mechanism did what an ETP structure is supposed to do.

Plumbing: In-Kind Transfers and Generic Listing Standards

Two regulatory changes after 2024 mattered more to the category's structure than any single fund launch.

In-kind creation and redemption. On May 15, 2025 the staff of the Division of Trading and Markets stated that broker-dealers are permitted to facilitate in-kind creations and redemptions for spot crypto ETPs. BlackRock's trust reported that SEC actions on July 29 and 31, 2025 authorised in-kind activity for IBIT. (IBIT Form 10-Q)

Generic listing standards. On September 17, 2025 the Commission granted accelerated approval to substantially identical proposals from Nasdaq, Cboe BZX and NYSE Arca adopting generic listing standards for Commodity-Based Trust Shares — Release No. 34-103995, 90 Fed. Reg. 45414. (Release No. 34-103995)

The practical effect: a qualifying product no longer needs its own Section 19(b) rule filing and Commission order. The eleven-year bottleneck described in sections 2 through 6 was, for products meeting the standards, removed.

2026: A Bank Issues Its Own

The Morgan Stanley Bitcoin Trust was formed as a Delaware statutory trust on December 16, 2025. Its final prospectus is dated April 6, 2026 (Registration No. 333-292586); shares list on NYSE Arca under the ticker MSBT, with an inception date of April 7, 2026. (MSBT prospectus)

From the prospectus:

• Fee: a unitary Delegated Sponsor Fee accruing daily at an annualised 0.14% of net asset value, out of which the Delegated Sponsor pays all operating expenses other than litigation and extraordinary expenses.
  FILED · PROSPECTUS

  Source

  MSBT final prospectus (Form 424B3), dated April 6, 2026

  As of

  2026-04-06

  Figure quoted from the filing itself — a primary source, not an aggregator or estimate.

• Delegated Sponsor: Morgan Stanley Investment Management Inc.
• Benchmark: the CoinDesk Bitcoin Benchmark 4PM NY Settlement Rate.
• Custodians: The Bank of New York Mellon and Coinbase Custody Trust Company, LLC.
• Creations and redemptions: both cash and in-kind — a direct consequence of the 2025 changes above.

The governance structure is unusual and worth noting for anyone building fund-comparison data: the trust has two trustees, CSC Delaware Trust Company and AGS Trustees Limited (Cayman Islands), with the Cayman trustee having delegated substantially all day-to-day management to Morgan Stanley Investment Management under agreements dated March 25, 2026.

At 0.14%, this is the lowest sponsor fee among US spot Bitcoin ETPs verifiable from filings.

The Income Category

On June 12, 2026 the SEC declared effective the registration statement of the iShares Bitcoin Premium Income ETF, and the shares listed on Nasdaq under the ticker BITA on June 16, 2026. The filing sequence: Form S-1 on January 26, 2026 (Registration No. 333-292938), Amendment No. 4 on June 9, 2026, Form 8-A12B on June 11, 2026. (BITA Form 10-Q, period ended June 30, 2026)

The trust seeks to reflect generally the performance of the price of Bitcoin while providing premium income by writing call options primarily on iShares Bitcoin Trust shares and, from time to time, on Exchange-Traded Products Indices. Its sponsor's fee accrues daily at an annualised 0.65% of net asset value.

Its first balance sheet shows the hybrid construction plainly:

This is a different animal from the 2024 cohort. Those funds were passive vehicles whose only job was to track a price. This one holds spot Bitcoin and shares of another fund and writes options against them, and its return profile depends on realised versus implied volatility as much as on Bitcoin's direction. Comparing it to IBIT on fee alone compares two unlike things — our BITA analysis works through the mechanics and what the income costs in BTC terms. The sharpest version of that unlike-things lesson is a company rather than a fund: our IBIT vs MSTR comparison shows why an operating company's bitcoin-per-share is a different object from a trust's.

A caution on the “first” claims circulating about BITA: it is not the first US Bitcoin covered-call ETF. Roundhill's YBTC, a 1940 Act fund that does not hold Bitcoin directly, describes itself as the first US-listed Bitcoin covered-call ETF. What is distinctive about BITA is the combination — a 1933 Act commodity trust holding spot Bitcoin that also runs an options overlay — and the size of the issuer.

The First Liquidation

On August 3, 2026 Hashdex Asset Management Ltd. announced a plan to liquidate the Hashdex Bitcoin ETF (NYSE Arca: DEFI), terminate its continuous offering and deregister its shares. Assets under management were approximately $14.7 million as of July 30, 2026. (Form 8-K, Aug. 3, 2026)

No filing designates it the first. The record does that work instead: no US spot Bitcoin ETP other than DEFI has liquidated since the category began trading in January 2024, so the designation holds by enumeration. What shareholders receive, and when, is covered in our liquidation explainer.

The key dates: no creation orders after August 17, 2026; trading on NYSE Arca suspended after the close the same day; liquidation of the trust's Bitcoin beginning August 18, 2026; proceeds to shareholders on or about August 24, 2026.

That date requires a note. The original 8-K filed August 3 contained an internal conflict — the body gave August 24, while the press release attached as Exhibit 99.1 gave August 28. Most coverage reported August 28. An 8-K/A filed August 12 resolved it: the Plan of Liquidation gives a liquidation date of August 18 and proceeds on or about August 24. (8-K/A, Aug. 12, 2026)

DEFI's history is more complicated than “one of the January 2024 eleven,” and it matters for anyone maintaining a fund registry. The predecessor fund, the Hashdex Bitcoin Futures ETF, was a series of the Teucrium Commodity Trust sponsored by Teucrium Trading, LLC, and commenced operations on September 15, 2022. Its initial Form S-1 was declared effective on January 2, 2024; the merger into the surviving entity closed January 3, 2024; the fund was renamed the Hashdex Bitcoin ETF on March 26, 2024. Its sponsor was Tidal Investments LLC until January 16, 2026, when Hashdex Asset Management Ltd. became successor sponsor. (Form 10-K FY2024)

The sponsor's stated reasons were assets under management, trading liquidity, operating costs, investor interest and product-line fit — not any judgment about Bitcoin.

What the Record Shows

The legal question was never whether Bitcoin is a good investment. The 2019 Bitwise order says the disapproval does not rest on an evaluation of whether Bitcoin or blockchain technology has utility or value. The 2024 approval says the Commission does not apply a “cannot be manipulated” standard and does not understand the Exchange Act to require that a product or market be immune from manipulation. (Release No. 34-99306, n.61) The dispute was about surveillance, and about the burden a listing exchange carries under Section 6(b)(5).

The unlock was procedural, not substantive. The D.C. Circuit did not find Bitcoin markets adequately surveilled. It found the SEC had not explained itself. The Commission then approved on a theory it had written into its own 2018 order and left dormant. What changed empirically was the quality of the correlation evidence, not the structure of the Bitcoin market.

The bottleneck that defined the decade no longer exists. The generic listing standards adopted in September 2025 removed the requirement for a product-specific Commission order. The process that took eleven years for the first product now takes none for a qualifying one.

FAQ