How to Launch an ETF: The Number That Decides It Is Not the Launch Cost

Every conversation about launching an ETF starts in the wrong place. Somebody asks what it costs to get to market, gets an answer around ninety thousand dollars, and concludes the barrier is low.

The barrier is low. That is genuinely true and it is the least useful fact in the discussion, because the ninety thousand is a one-time number and the fund keeps costing money every year afterward, whether or not anybody buys it. The question that decides whether this works is not what it costs to launch. It is what it costs to still be here in year three, and what has to be in the fund by then.

Behind the Ticker has had four separate episodes with the people who actually run this process. Here is what they say, with the numbers.

The cost really did collapse

Garrett Stevens of ETC, who has been doing this since before the current regime existed, described the change in one comparison. Launching in the old world "cost us a million bucks," and by 2023, in his words, "90 grand to get to market is typical, right? That’s what it costs to our platform to get a fund" listed, a reduction of more than ninety percent in about fifteen years. Quotes move: Mike Venuto cited under seventy thousand in late 2024. Get a current number rather than working from either.

Two things caused that. The first was regulatory. Before 2019, launching an ETF meant applying to the SEC for individual exemptive relief, a process Stevens remembers from "2008. Back in those days, you know, you had to get exemptive relief from the SEC. And it took us" a long time. Rule 6c-11 ended that for standard funds. As the K&L Gates team put it, if you are "managing a fully transparent, passively managed ETF, you don’t have to get exemptive relief anymore." The rule covers actively managed ETFs too, provided they publish holdings daily. What it excludes is leveraged and inverse funds, unit investment trusts, non-transparent active funds, and share classes of multi-class funds. And relief is not the only hurdle: you still need an effective registration statement either way.

The second was the arrival of white-label platforms, which is really the arrival of a rental market for the infrastructure.

Rent the trust, do not build one

An ETF has to live inside a registered investment company, and the fund is a series of it. You can create your own, or join one that already exists.

Mike Venuto of Tidal gave the number that settles the question for almost everybody: it costs "somewhere between $250 [thousand] and $1 million to start that trust."

Against that, joining an existing series trust means the legal entity, the board, the compliance program, the auditor relationship and the service-provider contracts already exist. You are adding a series to a structure that works. The K&L Gates team confirmed the practical effect on timing: with an "existing series trust, then it’s much faster, because you don’t have to go through" the formation process.

Stevens put the economics of doing it yourself bluntly. For a fund with "10 or $20 million in it, you’re going to spend hundreds of thousands of dollars" running your own structure, which no fund that size can carry.

The clock

Once the paperwork is filed, there is a regulatory waiting period. Stevens gave the figure his firm plans against: "From the time you file the prospectus, the SEC has 75 days to declare that fund effective." The mechanism is slightly different from how he puts it. Under Rule 485(a)(2) the amendment goes effective automatically on the 75th day rather than being declared effective, and the same 75-day path applies to adding a series to a trust that already exists.

That is the floor, not the expectation. It assumes the filing is clean, the strategy is conventional, and nothing in it invites a second look. A novel strategy, an unusual asset, or anything requiring relief adds time that is not on anybody’s schedule.

Where the seed capital comes from

A fund cannot list with nothing in it. Somebody has to put up the initial capital that buys the first creation unit.

Frequently, that somebody is the market maker. Paul Weisbruch of GTS described the arrangement from his side: on a genuine launch rather than a conversion, "we provide the seed. And then we agree to be the lead market maker" for the fund. The firm funding the first shares also commits to quoting them.

That is a service with a price, and it is a relationship rather than a favour. It also means one of your earliest conversations is with a trading firm assessing whether your fund will be tradeable.

What actually recurs

Here is the part that gets left out of the ninety thousand dollar headline. Venuto listed the annual stack: "you hit year end and you’ve got listing fees, indexing fees, audits, your board, your annual board meeting."

Add the ones he did not name and the picture fills in: the administrator, the custodian, the transfer agent, the distributor, the chief compliance officer’s time, the fund’s own legal counsel, the annual and semi-annual report production, the tax work, and the cost of maintaining a website that satisfies the disclosure rules.

On a white-label platform much of this is bundled, which is the point of the platform. Bundled is not free. It is a recurring number that the fund’s expense ratio has to cover, and until assets arrive, the sponsor covers it out of pocket.

That is the real arithmetic. A fund with a 0.65% expense ratio generates $65,000 a year at $10 million in assets. Run the recurring cost against that number and you have your answer about what scale the fund needs, and how long you can fund the gap.

The honest counterweight

Will Rhind of GraniteShares gave the optimistic version, and it is true as far as it goes: "the barrier to entry in ETFs has never been as low as it is today, meaning that if you want to launch an ETF, you can do that," and "If you can write a check, you can launch an ETF."

Both statements describe the launch. Neither describes the outcome, and the people who run platforms are noticeably more measured about that part.

Venuto’s rule of thumb for how long to commit: "usually three years is enough to hit one of your cycles and get to the break even and things like that." Three years of recurring costs, not one, is the commitment you are actually making.

And Kyle Wiggs of UX Wealth named the failure mode he sees most in advisers who launch: "the mistake that I think they’re making, and we try to talk them out of this, is don’t be in such a hurry to launch an ETF and forget your core business." An adviser with a healthy book who spends two years on a fund that gathers eight million dollars has usually made a poor trade, and the cost is not the launch fee.

If you are converting instead

Converting an existing mutual fund or SMA book brings assets on day one, which solves the hardest problem in the business, and it brings a set of questions of its own.

Venuto has done one and was measured about it: "We have done one mutual fund conversion and it was thoughtful, meaning to convert a mutual fund, you want to make sure that the bulk of the clients have not paid loads." His summary afterward: "one of the learnings of that is the conversions, not that simple."

Shareholders in a mutual fund who paid a sales load bought something with a distribution arrangement attached. Moving them into a vehicle that trades on an exchange changes what they own and how their adviser gets paid, and it needs handling rather than announcing.

The order to do this in

Answer the recurring cost question first. Get the annual number from a platform, in writing, and work out what asset level makes the fund self-sustaining.

Then decide how long you will fund the gap. Three years is the number practitioners use. Write it down before you start, because the decision to stop is much harder to make later.

Then choose the wrapper route. Rent a series trust unless you have a specific reason and enough scale to justify your own.

Then, and only then, work on the strategy and the filing.

Most people do this list backwards, and the strategy is the enjoyable part, which is exactly why it is the part that gets all the attention.

This is educational content and not investment advice. It is not a recommendation regarding any security. Investing involves risk, including possible loss of principal. Past performance does not guarantee future results.