Launching an ETF

The chain, step by step

1. Legal structure and approval. In the US, an open-end fund registered under the '40 Act, operating within Rule 6c-11. In Europe, a UCITS fund authorised in a domicile (Ireland or Luxembourg, per Chapter 50) with a passport for cross-border distribution (Chapter 55). Timelines run to months rather than years for standard products; anything novel still requires bespoke engagement with regulators.

2. The index, or the absence of one. License a branded index (Chapter 39's ad valorem fees, paid from the TER), commission a challenger provider at flat rates, self-index under governance rules, or run an active mandate with no index at all (Chapter 76). This choice sets a recurring cost and, for branded indices, contributes to the fund's marketability.

3. Service providers. Custodian, fund administrator, transfer agent, auditor, legal counsel, and index calculation. These are the fixed costs that produce Chapter 88's viability threshold.

4. Exchange listing. Meeting listing standards and paying listing fees, plus arranging the exchange's market-maker programme where applicable (Chapter 32).

5. Authorized participants and market makers. At least one AP agreement is required to function at all, and quoting quality depends on how many market makers find the fund worth supporting: which depends on whether the basket is hedgeable and whether volume is expected. A fund nobody wants to make markets in trades badly from day one, regardless of its merits.

6. Seed capital. Someone must fund the first creation unit. Typically the sponsor, an affiliated party, or a seed investor negotiated in advance. Seeding is the moment a fund becomes real and the point at which many planned launches quietly die.

7. Distribution. Platform approval, adviser access, model-portfolio inclusion, and in Europe country-by-country registration and paying-agent arrangements. This is the step that determines survival, and the one first-time sponsors consistently underestimate.

What changed: Rule 6c-11 and white-labelling

The pre-2019 world. Every US ETF needed individual exemptive relief from the SEC, a process taking years and considerable legal expense, with terms varying by applicant (Chapter 70's custom-basket asymmetry being the notable consequence). This was a genuine barrier to entry and it protected incumbents.

Rule 6c-11 (2019) created a standard pathway: funds meeting its conditions: daily portfolio transparency, website disclosure of premiums, discounts and spreads, and written basket policies: can launch without bespoke relief. Time-to-market collapsed, and the product menu expanded rapidly afterwards.

White-label platforms completed the shift. A firm with a strategy but no infrastructure can rent a fund trust, an administrator, a listing and the operational machinery, launching under a platform's umbrella for a fee. The result is that an asset manager, a research boutique or an individual with a thesis can bring an ETF to market, and many have. Europe has analogous arrangements, though the UCITS process and multi-country registration keep the friction somewhat higher.

The honest consequence: a lower barrier produced both genuine innovation and a great deal of product that exists because launching became cheap.

Why launches fail

The survival statistics are unforgiving: a substantial minority of ETFs close within a few years (Chapter 67), and the pattern of failure is consistent:

No distribution. The dominant cause. Platforms have limited shelf space and adviser due-diligence teams have limited attention; a fund without a distribution partner gathers nothing regardless of merit.

No differentiation. Launching the twelfth fund on a well-covered index means competing with incumbents on liquidity, fees and track record simultaneously. First-mover durability (Chapter 88) makes this close to unwinnable.

Timing. Thematic launches cluster at the top of a theme's popularity (Chapter 47's launch clock), which means many arrive precisely as performance turns and flows reverse.

The seed never grows. A fund launched with $5m of seed capital and no pipeline sits below viability from day one, and providers rationalise it within a couple of review cycles.

Liquidity spiral. Small AUM means few market makers, which means wide spreads, which deters buyers, which keeps AUM small. Several perfectly reasonable funds have died in this loop.

What makes a launch work

The survivors share features that have little to do with investment quality:

• A committed distribution channel before launch — an adviser network, a model-portfolio allocation, or an institutional anchor investor.
• Meaningful seed capital, ideally with a pipeline behind it rather than a single sponsor cheque.
• A hedgeable basket, so market makers can quote tightly from day one.
• Genuine differentiation: first in a category, materially cheaper, or an exposure genuinely unavailable elsewhere.
• A range rather than a product: a fund inside a coherent family with a committed sponsor survives thin periods that a standalone experiment does not.

What this means for you as an investor

Read the launch pipeline as information about the industry rather than about markets:

• A new fund is a business decision first. Ask what problem it solves that an existing fund does not, and whether the answer is "none, but the category sells" (Chapter 88's margin game).
• Prefer funds past the survival threshold, with a committed sponsor and a coherent range (Chapters 59, 67).
• Early is not an advantage for you. Being an early investor in a fund carries closure risk and wide spreads with no compensating upside: unlike an early investor in a company, you gain nothing from the fund's growth beyond marginally better liquidity.
• Novel exposures deserve extra scrutiny, not extra enthusiasm: the newer the structure, the more of Part III's chapters apply at once.

A worked example

A research boutique with a well-regarded systematic strategy decides to launch a US ETF.

• Structure: white-label platform, live in four months, all-in setup cost in the low hundreds of thousands rather than the millions the pre-2019 route required.
• Index: self-indexed under written governance, avoiding six figures a year of licensing.
• Seed: $8m from the founders and a friendly institution.
• Market making: two firms quote it; the basket is liquid US large-caps, so spreads are acceptable from day one.
• Distribution: no platform commitments, no adviser network, no model-portfolio allocations. The plan is "performance will attract assets."

Eighteen months later: the strategy has performed adequately, AUM is $22m, spreads have widened as market-maker interest faded, the platform fee is consuming the management fee, and the fund appears on watch lists for closure.

Nothing about the investment thesis was wrong. The chain had six links solved and the seventh unaddressed, and the seventh is the one that decides.