The ETF share class proves the staying power of the mutual fund
Advisers will have more ways to invest retirement account assets and lower transaction costs.
Mutual funds are dead, or dying, or maybe actually growing?
Even though more than 50 mutual fund sponsors have petitioned to add an ETF share class, the real story might be the dozens of ETF sponsors who have petitioned to add mutual fund share classes.
Why would they do that? Well…
Share-class flexibility is to 401(k)s what tax efficiency is to ETFs
Tax Alpha Insider
I’ll return to that point later, but for now, why should advisers care?
Advisers should be able to access previously ETF-only strategies in qualified accounts like 401(k)s and 403(b)s (e.g. F/m Invest’s strategies)
Pricing for some ETFs might be lower in qualified plans via mutual fund institutional and retirement share classes
ETF in-kind redemption should lower transaction costs for mutual fund share class holders
There is some “tax contagion” risk for taxable investors in mutual fund/ETF dual-class funds
The backstory
“The ETFs adding mutual fund share classes will NOW have a ‘real’ tool to drive distribution,” Bill Poulin, a 35-year fund executive, and consultant to fund company boards, told me on a call last week.
By “real” tool, he’s talking about how mutual funds are sold via different pricing structures through various mutual fund share classes (e.g. “I” institutional shares, “R” retirement shares, etc.).
“ETFs primarily rely on banner ads and paid media” for distribution, Poulin wrote on Twitter recently. At the same time, mutual funds are sold through distribution networks. ETFs may have revenue share arrangements but nothing as effective as mutual fund share classes.
This means mutual funds are sticking around.
Not serious investment products
The mutual-funds-are-getting-replaced-by-ETFs narrative gets shakier if we consider that some recent ETF growth is pure speculation. Two insiders, who wished to remain anonymous, said that a large slice of the new ETF AUM growth is trash, “not investment” products. They both pointed to various leveraged, inverse, high-cost, single-name ETFs that are silly compared to the buy-and-hold bastions that will still be around decades from now. “What happened to fiduciary duty!?” one asked, slightly exasperated.
Why the ETF share class?
The ETF share class solves the “401k [mutual fund] but tax efficiency for etf buyers” problem, Wes Gray at ETF Architect told me, essentially allowing fund sponsors to have their cake and eat it too, but let’s start from the beginning.
Mutual funds have two major (and several minor) drawbacks:
They redeem in cash, which means…
Possible forced selling
Capital gains distributions (even to investors who did not sell)
They hold a cash buffer to satisfy redemptions (cash drag)
This is bad even for non-taxed investors because selling means transaction costs.
Equity ETFs don’t have these problems; they redeem assets in-kind through heartbeat trades and usually hold minimal cash (though they may get front-run1).
How does an ETF share class solve a mutual fund’s problems?
Some context: In 2000, Vanguard received exemptive relief allowing its index-tracking mutual funds to add an ETF share class (source)
This was a big deal at the time. ETFs were dominated by State Street and iShares (now BlackRock)
How does the ETF share class solve this?
The ETF and mutual fund shareholders have an interest in the same pool of assets
“Perhaps the largest benefit of the dual class structure is the ability for the fund to distribute low-basis assets to APs when they redeem their ETF shares”
For example, in 2018, Vanguard (in)famously used its ETF share class and a heartbeat trade to remove highly appreciated shares of Monsanto from the fund via the ETF.
Vanguard filed a patent application in 2001, approved in 20052, blocking the industry from creating similar dual-class structures for nearly 20 years.
ETFs don’t have to hold cash like mutual funds do, and they avoid distributing capital gains and transaction costs (aside from front-running).
Why now?
Rule 6c-11 (2019) paved the way for more ETF launches
Vanguard’s patent expired3 in 2023 and…
New SEC leadership seems amenable
I have directed the Commission staff to prioritize their careful review of the many applications filed for this relief, and I look forward to considering their recommendations.
Is the ETF share class the only way to make a mutual fund more efficient?
No.
A mutual fund complex has several options:
Convert the entire mutual fund to an ETF via Section 351 (I wrote a short book about this, you should buy a copy)
Reorganize the entire mutual fund as an ETF via Section 368
Attach an ETF share class
Use a liquidity buffer like ReFlow or something functionally similar
Dimensional Fund Advisors has done all of these things and is acting like a petri dish. As an example, an insider mentioned that Dimensional previously had “tax managed” mutual funds that were distributing large capital gains “and it was a bad look,” so these were ideal candidates for a wholesale conversion to ETF.
The liquidity buffer is probably the most mysterious option to readers, so let’s talk about that.
What is a liquidity buffer?
Mutual funds have two defects investors despise: cash drag and tax drag.
Since many mutual funds have been in net redemption for several years4, some are in a downward spiral, where they get redeemed, must sell and distribute gains, and then get redeemed again. All the while, they’re holding a cash buffer to process the redemptions.
ReFlow was founded “to implement proactive liquidity solutions to reduce friction and performance issues due to disruptive shareholder flow.”
Here’s a short summary of their services using Dimensional as a case study:
Purpose: ReFlow provides cash to mutual funds facing net redemptions
Mechanism: ReFlow buys fund shares daily, up to net redemptions, then redeems later when:
The fund has net sales, or…
The fund requests that ReFlow redeem, or…
ReFlow hits its max holding period of 8 days
Redemption: Typically in-kind, per fund policy, using the same section of the IRC that allows an ETF’s authorized participant to redeem shares in-kind (§852(b)(6))
Fees: Fund pays ReFlow a fee (~0.14%) via daily auction; cost is spread across share classes
Blind Purchases: ReFlow buys without considering fund strategy or performance
No Load: ReFlow buys at NAV, with no sales charges or investment minimums
Limits: Cannot exceed 3% ownership of a fund
ReFlow’s CEO, Bill White, told me they act as “a benevolent shareholder, buying and selling shares to offset shareholder flow.”
ReFlow’s position as a liquidity provider is supported by a 2002 no-action letter5 from the SEC, which requires:
They receive no special treatment
Are paid a fee
“A gigantic gains overhang makes it really hard to market a mutual fund to taxable and non-taxed investors... distributing gains still causes asset liquidation, which includes transaction costs,” White explained. “The trading costs are definitely material.”
When a fund is in this position, conversion to ETF might work, but if it has qualified plan shareholders, they're in a tricky position, and ETF share class may be the only answer.
"We always look at the least-advantaged shareholder," including taxed and non-taxed investors, White explained.
Does the ETF share class threaten ReFlow’s business model?
“We see them as complements,” White suggested.
When a mutual fund adds an ETF share class, it’s not a panacea.
The ETF share class may be small versus the overall fund. While using a small ETF to do big tax work is possible, it’s less “natural” and may prompt scrutiny.
Moreover, mutual fund shareholders still expect redemptions in cash, and an ETF is set up to create and redeem securities in kind and, therefore, isn’t much help. This means the mutual fund will need liquidity, and ReFlow will have customers.
ReFlow is positioned to scale. “We had a chicken and egg problem for a while as more mutual fund complexes learned about our solutions,” White continued that they are now doing a “couple billion dollars a month in transactions” and getting more interest.
The tradeoffs of an ETF share class
All wholesale conversions into the ETF have the same major drawback: The ETF might fail to “stop the bleeding” of mutual fund net redemptions.
There are other considerations, too:
Not all investors hold their mutual fund shares in brokerage accounts and may be forced to liquidate.
Small or esoteric ETFs might not attract authorized participants who are themselves attempting to optimally deploy capital
Plus, simple things like a ticker change and track record won’t transfer cleanly
The ETF share class solves many of these problems. But it also has tradeoffs.
Every ETF share class petition mentions some chance of “cross-subsidization6” risk, where “cash flows associated with Mutual Fund Classes could impact a Multi-Class ETF Fund’s portfolio, generating costs that shareholders of all classes, including the ETF Class, would share.”
For example, in 2009, the Vanguard Extended Duration Treasury Index fund distributed a giant capital gain to ETF shareholders following outflows from the institutional mutual fund share class. “This risk isn't present in ETFs that aren't linked to a mutual fund,” said Morningstar’s Ben Johnson in a 2020 piece on ETF share classes.
“Investors can assess the level of risk in a[n] ETF by looking at its size relative to the fund and trying to understand the nature and objectives of shareholders in the mutual fund share classes,” Johnson said in the same article. “The greater the portion of assets invested in the ETF, the lower the risk is likely to be.”
I asked around, and nobody seemed able to find another instance of tax contagion like the case above. Still, it’s worth adding to the tax diligence checklist as advisers consider new products.
A liquidity buffer might help. “[I]t is the only game in town,” if for whatever reason a mutual fund cannot convert to an ETF, Wes Gray suggested.
The ball is in the SEC’s court.
If the 50+ petitions for new mutual fund and ETF share classes are approved, we’ll see if it makes a meaningful difference in fund flow patterns.
Aside from the tax contagion case, dual-class funds seem like a win for investors.
Thanks for reading.
See Elisabeth Kashner’s “The Heartbeat of ETF Tax Efficiency” for more on this.
Side notes about the patent…
Vanguard allowed investors to seed the new ETF share class in-kind, but the assets had to be held at Vanguard
The shares were called VIPER Shares (“Vanguard Index Participation Equity Receipts”)
From the patent application: “Unlike conventional mutual fund shares, an organized secondary market is expected to exist for VIPER Shares…” (describing that ETFs are meant to trade)
Interestingly, “The validity of Vanguard’s patent [was] open to doubt in the light of Alice Corp. v. CLS Bank International, 573 U.S. 208 (2014).” The argument was that:
“…an abstract idea could not be patented just because it was implemented using a computer.” and…
Vanguard’s patent was a “business method patent,” which was “routinely invalidated” following Alice.
Despite some licensing discussions and one competing petition (Van Eck, “which was never acted upon by the SEC”), Vanguard remained largely unchallenged until their patent expired in 2023.
This overstates the migration from mutual funds into ETFs. Bill Poulin would say that some mutual fund share classes have migrated aggressively to ETFs. For example, the equivalent of “A” share classes, which usually include large front-loads, are disappearing rapidly. But “R” shares are still growing because they provide qualified plan administrators with various payment solutions.
The 2002 no-action letter is for ReFlow Fund, LLC, which is pooled investment vehicle and the source of liquidity financing. Investors in the fund receive fees paid by mutual fund complexes. Since ReFlow serves as a liquidity provider for many mutual fund complexes, “we essentially hold the market,” Bill White mentioned, which they hedge with equity futures.
Editor’s note: I hate euphemisms like this. Let’s just be straightforward and call it what it is: “cross-contamination” or, as an insider put it, “tax contagion.”
excellent