I keep meeting families who treat a pile of appreciated stock like a locked room. The door works. They just refuse to turn the key, because the tax bill on the other side looks ugly. That hesitation is rational. What is less rational is assuming a clever wrapper can open the door, hand them a diversified portfolio, and leave the tax bill in the hallway. This week the Treasury Department and the IRS made it plain they have been watching that hallway.
The warning is aimed at a narrow but expensive habit: seeding a freshly formed exchange-traded fund with a basket of highly appreciated securities, then watching that basket get swapped, redeemed, or reshuffled so the original owner walks away diversified and still untaxed. Officials described some of these setups as conduits. The Treasury secretary put it more bluntly on social media, saying the government is serious about transactions built to dodge tax, and that a companion revenue ruling on these conversions carries a simple message. Under existing law, they do not work.
Why a Legitimate Code Section Became a Wealth Shortcut
Section 351 of the tax code is not a loophole invented on a trading desk. It is an old corporate rule. If you hand property to a corporation and take back stock, and if the transferors control that corporation right after the exchange, you generally do not recognize gain. The idea is mechanical. You changed the form of ownership, not the economics. A factory owner who incorporates a factory should not owe tax just for printing shares.
ETFs are corporations for this purpose, or they sit inside structures that can use the same nonrecognition rule. That is where the creativity started. A wealthy investor, usually through an advisor and a sponsor, contributes a portfolio of stocks into a new fund. In return they receive ETF shares. On paper, nothing was sold. The built-in gain stays attached to the contributed shares, waiting for a later sale by the fund or by the shareholder.
Done cleanly, that can be ordinary fund formation. Done as a costume change, it starts to look like a sale that refuses to call itself a sale.
The Diversification Tests People Quote at Dinner
Advisors who pitch this idea almost always mention two numerical guardrails. No single asset can exceed 25 percent of the contributed portfolio, and the five largest holdings cannot exceed 50 percent. Those figures come from the diversification requirements that sit alongside Section 351 when investment companies are involved. Miss them, and the nonrecognition promise can fail before the conversation even gets interesting.
I have sat through versions of this pitch where the percentages are treated like a magic spell. Hit 24 percent on the biggest name, keep the top five under half, and the tax disappears. That is not how the statute reads, and it is certainly not how the new guidance reads. The percentages are a floor, not a hall pass.
A portfolio can satisfy the math and still fail the purpose test if the fund was never meant to hold what it received. That distinction is the whole argument.
What the Authorities Actually Object To
The revenue ruling walks through a fact pattern that tax lawyers will recognize instantly. An investor transfers a portfolio into a newly formed ETF. Soon afterward the fund distributes those contributed securities, or otherwise sheds them, and the investor is left holding something materially different. No gain was reported on the original shares. The economic result is diversification without a taxable sale.
This is really about getting diversification without paying tax.
A tax partner describing the challenged pattern
Authorities called that result a problem. They also flagged cases where the ETF is merely a conduit for moving securities around. A notice, broader than a ruling, lays out the kinds of facts that make a transfer look abusive rather than like genuine seeding. A ruling interprets one set of facts. A notice tells the market which neighborhood the agency intends to patrol.
Both landed in the same week, which is rarely an accident.
The Scale That Made This Impossible to Ignore
A market analysis published last summer put a number on the habit. Roughly $22 billion of ETFs had been created for this kind of contribution, with deferred gains estimated as high as $6.5 billion. Activity picked up sharply after 2024. Those are not retail-account figures. They are family-office figures, the kind that show up in compliance meetings and, eventually, in revenue estimates.
Back in July the Treasury secretary had already posted that tax rules should reward investment, not abusive financial engineering, and that a pitch sounding too good to be true probably is. This week’s messages pointed back at that warning. Officials had also sat down over the summer with a Wall Street tax group to talk through the transactions drawing interest, including questionable Section 351 exchanges. The notice tracks many of the strategies raised in that room.
When a cabinet official repeats a warning and the agency publishes both a ruling and a notice, the “wait and see” posture stops being conservative. It starts being hopeful.
Who Actually Uses These Exchanges
Not your neighbor with a brokerage account and a handful of tech shares. Fees alone push this into rare air. Sponsors and counsel often quote $200,000 to $300,000 just to stand up the fund. Some firms tell clients the contributed stock should be at least $25 million before the economics make sense. One Boston wealth counsel I would trust on this sets the bar higher still, arguing it is hard to justify below $100 million of stock going in.
That cost structure explains the clientele. Legal work, seed mechanics, board formation, exchange listing, authorized-participant arrangements, and ongoing compliance do not shrink just because the investment thesis is “please diversify my founder shares.” The fixed cost is the filter.
- Formation costs often land between $200,000 and $300,000 before a single share trades.
- Practical minimums cited by advisors start around $25 million of appreciated stock.
- Some counsel will not entertain the idea under $100 million.
- The users are almost always higher-income families, not ordinary ETF buyers.
- Intermediaries, not the investor alone, usually assemble the vehicle.
If someone is selling a “351 ETF” to a household with a seven-figure brokerage account, the fee math should be the first red flag, not the last.
The Sentence That Keeps Legitimate Seeds Alive
Here is the line practitioners circled first. The notice says it does not address, and expresses no view on, transactions that seed a new ETF with assets consistent with the fund’s investment thesis, where those assets are intended and expected to stay unless circumstances change in a substantial way.
Read that twice. The government did not outlaw contributing securities to a new fund. It drew a circle around contributions that match what the fund claims it wants to own, and that the fund actually plans to keep. That is a purpose test dressed up as a non-ruling. It leaves real seeding on the table and puts costume changes outside.
A few ETF managers read the same sentence as an invitation. One founder argued, against the consensus, that regulators had opened the door for well-designed 351 transactions to move toward the mainstream. I understand the optimism. I do not share it without a long holding record and an investment process that would make sense even if no one had a tax problem.
Valid Reasons Families Still Bring Up
Not every contribution is a costume. A principal at a multi-family advisory firm pointed to situations that can still be coherent. A parent may want to give ETF shares to someone who has no interest in a pile of individual stocks. Converting separately managed accounts into a fund can also cut the ongoing tax drag that comes from constant rebalancing inside a taxable sleeve, which improves after-tax results if the fund is run as a fund.
Those motives survive the notice, provided the assets fit the thesis and are expected to remain. Gifting a real portfolio product is different from parking concentrated winners in a shell that empties itself in a quarter. The difference is facts, not branding.
Perhaps the most interesting aspect of the notice is how carefully it preserves ordinary ETF plumbing. Creation and redemption of shares, the in-kind mechanism that makes the whole industry tax-efficient, is not the target. One counsel put the scale in perspective: this is a multi-trillion-dollar industry. Nobody in Washington is about to break the daily create-redeem cycle that lets large funds hand out low-basis shares to authorized participants instead of selling them in the market.
The pressure is aimed lower, at smaller vehicles that create units and then diversify immediately for a tax result. Guardrails, not a funeral.
The Gray Zone Everyone Will Argue About
Timing is the trap. The notice treats transactions that happen shortly after appreciated securities are contributed as suspect. It does not define shortly. Tax lawyers hate undefined adverbs, and they are right to.
A tax analyst who writes about portfolio tax strategy listed three markers of aggressive planning: evidence of a plan, a quick redemption after the seed, and a portfolio that looks very different from what was contributed. He also noted, correctly, that this leaves a wide band of facts-and-circumstances fights. Expect those fights. The comment deadline on the notice is October 28, which is close enough that sponsors still shaping a seed should assume their memos may be read by someone other than the client.
How long is long enough? I do not have a number the IRS will bless, and neither does anyone else who is being honest. A week is theater. A month is still thin if emails from March describe the April redemption. A year of genuine holding, with trades that match a published process, looks different. Between those poles sits the litigation.
Markers that make a seed look planned: 1. A written diversification timetable 2. Redemptions clustered right after launch 3. End portfolio unlike the contribution 4. Investors who exit once the swap is done 5. A thesis that exists only in the offering memo
What “Materially Different” Is Likely to Mean
The ruling’s phrase is awkward on purpose. The investor ends up with a materially different portfolio and has not recognized built-in gain. Materially different is not a spreadsheet cell. It is a story. If you contributed a single-sector stack and six weeks later you hold a broad market blend, the story tells itself. If you contributed quality compounders and the fund still holds most of them, trimmed at the edges for risk, the story is harder to attack.
Courts and agents look at intent, timing, and result together. An email that says “we will rotate out of the contributed names after the quiet period” is worse than a bad quarter of performance. I have found that the paper trail, not the performance chart, is what decides these questions once an audit starts.
Sponsors who want the strategy to remain usable should write the investment process as if a revenue agent will read the investment committee minutes. Because one might.
Adjacent Trades Now Sitting in the Same Spotlight
The notice does not stop at the plain contribution. It points at transfers into partnerships connected with Section 351 conversions, and at ETFs using box-spread strategies. A box spread pairs options so the payoff resembles a loan or a deferred sale, depending on how it is built. Used inside a fund, it can be marketed as another way to push capital gains into a later year.
A tax partner at an accounting firm wrote that practitioners need to understand these highlighted strategies so they can warn clients about emerging audit exposure. That is the correct professional tone. This is no longer a cocktail-party idea. It is an exam topic.
Partnership detours deserve their own caution. Moving assets through a partnership on the way into a corporate ETF can look like an attempt to launder the tax character of the contribution. If the only reason the partnership exists is the conversion, the substance-over-form arguments write themselves. Extra entities are not sophistication. Sometimes they are a map.
Congress Could Still Change the Plumbing
A law professor who studies tax policy noted that lawmakers could narrow the benefit ETFs get from distributing appreciated securities in kind. That benefit is older and broader than any boutique 351 seed. It is also politically easier to describe than it is to repeal, because the same mechanism keeps taxable distributions low for ordinary shareholders in giant funds.
I would not bet a portfolio on a repeal this session. I would bet on more administrative guardrails, and on audit selection that prefers small, custom funds over household names. Customization is the tell. A fund with one family as the economic engine and a generic index as the costume is easier to explain in a Senate hearing than a fund with a million shareholders.
If Congress does step in, the change is more likely to limit in-kind distributions of contributed property, or to impose a holding period before low-basis lots can leave the fund, than to scrap ETF tax treatment wholesale. That is a guess, not a forecast. It is also the version that raises revenue without blowing up retirement accounts.
Exchange Funds, the Older Cousin
Advisors who still need a diversification tool are pointing clients toward exchange funds. These are private vehicles, typically limited partnerships, that pool concentrated stock from many investors into a broader book and defer the gain. The classic trade-off is time. Participants generally need a seven-year hold before they can redeem units and receive a diversified basket.
Seven years is not a slogan. It is illiquidity, manager risk, contribution minimums, and a portfolio you do not control. In return you get something the challenged ETF pattern was trying to imitate: diversification with deferral, inside a structure the market has used for decades. It is slower, more expensive to exit, and harder to explain at a dinner. It is also less likely to be described as a conduit in a revenue ruling published this week.
The comparison is worth making explicitly.
| Feature | Custom ETF seed | Exchange fund | Charitable remainder trust |
| Typical user | Very large taxable portfolios | Concentrated stock holders | Philanthropic families |
| Liquidity | Market trading, if listed | Often seven years | Income interest, not principal |
| Control | High if you seed it | Low, pooled | Trustee-directed |
| Current scrutiny | High on quick rotations | Established, still technical | Established, gift rules apply |
| Main trade-off | Audit and purpose risk | Lockup and fees | You give up the remainder |
None of these is free. Each moves the tax result by giving something up: time, control, or the asset itself. The pitch that gives up nothing is the one the ruling is aimed at.
Charitable Remainder Trusts as a Different Bargain
The other suggestion floating around advisor notes is a charitable remainder trust. You contribute appreciated stock, the trust can sell without immediate tax to you, you keep an income interest for life or a term of years, and a charity receives what is left. Capital gains are not erased. They are spread, and part of the economics is a gift.
People who want every dollar back should not use this. People who were going to give anyway, and who need income, sometimes find the math kinder than a outright sale. Basis, payout rate, and the charitable deduction all matter, and a sloppy trust is an expensive mistake. Still, it is a statute written for this problem, not a fund costume pulled over it.
I prefer tools whose purpose matches their paperwork. That preference is not moralizing. It is audit hygiene.
How Ordinary ETF Tax Efficiency Is Different
It is easy to smear every ETF with this story. That would be a mistake. The everyday tax advantage of a broad fund comes from in-kind redemptions. When an authorized participant wants shares back, the fund can deliver a basket of securities instead of cash. The fund chooses lots. Low-basis lots leave. The fund’s remaining shareholders are less likely to see a capital-gain distribution. No custom seed is required. No family is the secret limited partner.
That mechanism is why so many taxable investors prefer funds over mutual funds that must sell inside the portfolio to meet redemptions. It is also why officials keep saying they are not trying to kill the industry. They are trying to stop a personal diversification trade from borrowing the industry’s vocabulary.
If your ETF is a widely held fund you bought on an exchange, this week’s notice is background noise. If your ETF was born to hold your stock and then forget it, the notice is the subject line.
Questions Advisors Should Be Asking Before the Next Seed
A serious review is shorter than a pitch deck and less flattering. Start with purpose. Would this fund exist if the contributor had a high basis? If the answer is no, stop. Then look at the portfolio the fund says it wants. If the contributed names are a poor fit, the “consistent with the investment thesis” sentence in the notice is already working against you.
Then read the emails. Plans to redeem, rotate, or “clean up the book” after launch are exactly the evidence the ruling cares about. A compliance memo that says the opposite of the client memo is not clever. It is a exhibit.
- Write the investment thesis before selecting the contributed lots.
- Test the 25 percent and 50 percent diversification limits on actual values, not targets.
- Assume any redemption inside the first several quarters will be examined.
- Keep contribution, launch, and later trades in one file a stranger could follow.
- Price the legal cost of being wrong, not just the cost of formation.
- Compare the after-tax result with an exchange fund and with a staged sale.
Step six is the one pitches skip. A staged sale over two or three tax years, paired with charitable gifts of the highest-basis lots or of the shares themselves, is boring. Boring sometimes wins once you model state tax, Medicare surtax, and the fee drag of a custom fund that never gathers outside assets.
The After-Tax Math People Skip
Deferral is not the same as forgiveness. The built-in gain remains inside the contributed securities. If the fund later sells them, the fund recognizes it. If the shareholder later sells the ETF shares, the shareholder’s own basis, carried over from the contribution, determines gain. A 351 exchange that works still leaves a basis problem for someone. The fantasy version pretends the gain evaporated because the ticker changed.
Run a simple comparison in your head. Sell $10 million of stock with $8 million of gain, federal long-term rates plus the net investment income tax, plus state. Painful, immediate, done. Or contribute, pay $250,000 to build a fund, hold ETF shares with a carryover basis, and hope the fund never trips the gain in a way that lands on you. The second path wins only if time, stepped-up basis at death, or genuine fund economics do the work. If you need the money in three years, the path is a postponed invoice with legal risk stapled to it.
Stepped-up basis at death is the quiet third option. Hold the stock, diversify around it, borrow modestly if cash is the issue, and let the basis reset if the estate plan supports that. It is not available to everyone, and borrowing against a concentrated name has its own margin risk. It is still more coherent than a fund designed to be abandoned.
Rough decision filter: if the only payoff is tax deferral, and the portfolio will not be held, treat the seed as a sale until counsel proves otherwise.What Sponsors May Change in Offering Documents
Expect new risk-factor language, and expect it to be specific. Funds that accept in-kind seeds will describe holding intentions, diversification limits, and the possibility that the IRS challenges nonrecognition. Some will refuse custom seeds entirely until comments are digested. Others will accept them only when the contributed basket already looks like the target index, which rather defeats the client who wanted out of a single stock.
That last point is the commercial tension. The client’s goal and the notice’s safe sentence pull in opposite directions. The client wants out of the appreciated names. The notice is comfortable when the fund wants those names. A structure that satisfies the client and the notice at the same time is possible, but it will not feel like the product that was sold in 2024 and 2025.
Seed investors should also ask who else is in the fund. A vehicle with outside capital, a real board, and a process that would embarrass nobody in a public filing is a different animal from a family sleeve with a ticker. Outside capital is not a shield by itself. It is evidence of a business.
Audit Exposure, in Plain Language
Revenue rulings do not assess tax by themselves. They tell you how the agency will treat a pattern, and examiners use them. If your facts rhyme with the ruling, expect the nonrecognition claim to be denied, penalties discussed, and interest running from the year of the supposed exchange. If your facts rhyme with the notice’s untouched sentence, you still want a file that proves the rhyme.
Penalties are the part clients underweight. A substantial understatement, or a negligence penalty, changes the after-tax victory into a loss even before you count professional fees for the exam. Reasonable-cause arguments depend on advice that considered the actual plan, not a generic memo about Section 351 written for a different portfolio. If the advisor’s slide said “rotate after launch” and the opinion said “we will hold,” you do not have reasonable cause. You have a contradiction.
In my experience, the families who weather these exams are the ones who can explain the investment in a paragraph without using the word tax. If tax is the first sentence, the second meeting with the agent goes poorly.
A Note on Box Spreads and Deferred Gains
Box spreads have a legitimate life in options markets as a way to borrow or lend at implied rates. Inside a tax-motivated ETF, the pitch shifts. The option package is arranged so economic exposure changes while a gain recognition event is postponed. The notice puts that pitch on the list. Anyone running it should assume the agency wants comments, and then examples.
I am not interested in walking through strike selection. The practical point is smaller. If a fund’s marketing mentions tax deferral before it mentions a benchmark, read the notice before you read the fact sheet. Products designed around a code section age badly once the section is interpreted in public.
The same skepticism applies to partnership wrappers stapled to a conversion. Extra steps that have no business purpose are not planning. They are a narrative the other side will enjoy telling.
How This Lands on Portfolio Construction
Concentrated stock is still the real problem. Founders, executives with vesting schedules, and heirs sitting on a family holding company all face the same asymmetry. The position made them wealthy. Selling it feels like betrayal and a tax event at once. Good planning respects both feelings and then does math.
Tools that remain coherent after this week include trimming into strength over multiple tax years, donating shares directly so the charity’s sale does not hit your return, using a charitable remainder trust when the gift is real, pooling into an exchange fund when the seven-year lock fits the rest of the balance sheet, and building a completion portfolio that buys everything you do not own so the single name becomes a smaller slice without a forced sale. None of these is glamorous. All of them have a statute or a market practice that predates the custom-ETF boom.
Direct indexing gets mentioned in the same meetings, sometimes carelessly. Owning the stocks yourself and harvesting losses can improve after-tax returns in a diversified book. It does not erase a giant embedded gain in one name. Harvesting needs losses. A winner does not become a loss because it sits next to an ETF.
Borrowing can bridge spending needs so a sale can wait for a lower-income year or for a step-up. The rate, the collateral call, and the concentration risk have to be sized by someone who has seen a drawdown, not only a pitch. Leverage on the stock you are afraid to sell is a second bet on the same company.
What I Would Tell a Family Sitting on the Fence
If a sponsor already launched your seed, gather the record now. Thesis, contribution list, any plan to rebalance, redemption history, and the opinion letter should live in one place. Do not backfill a story. If the facts are clean, the file will show it. If they are not, you want counsel before an information document request, not after.
If you are considering a seed, pause until comments land and until someone who does not earn a formation fee has read the notice against your basket. Ask them to argue the government’s side first. A memo that only explains why you win is a brochure.
And if the goal is simply to stop being a one-stock household, say that out loud and price the tax. Sometimes the bill is the cheapest way out. Sometimes the seven-year pool is. Rarely, a fund that actually wants your shares is. The version where the fund pretends, briefly, is the version officials just described as not working.
Tax rules should reward investment, not abusive financial engineering.
Treasury secretary, restating the July warning
That line is political and legal at once. Families do not have to adopt the politics to respect the legal half. A transaction whose only cleverness is the tax result is a transaction the ruling was written to catch.
The Industry Incentive That Produced the Product
It is worth saying why this trade multiplied. ETF sponsors compete on assets. A single seed of $50 million or $200 million is a launch. Advisors compete on after-tax narratives. A slide that shows diversification with no gain is an easy meeting. Lawyers, most of them careful, were asked to map an old corporate rule onto a new product, and some maps were tighter than others. Volume followed the loosest map, which is how $22 billion accumulates.
Crackdowns follow volume. They also follow simplicity of explanation. “They swapped the stocks and skipped the tax” fits in a hearing. Complex but genuine seeding does not. Sponsors who want to keep the genuine version should make their funds boring in the best sense: published process, outside holders, holdings that match the name on the door.
Investors should assume marketing language will lag the notice by a quarter. The careful shops will already be rewriting. The shops that led with tax alpha may keep the old deck until someone asks a harder question.
State Taxes and the Quiet Extra Bill
Federal guidance is the headline. State conformity is the footnote that writes checks. Some states follow federal nonrecognition automatically. Others require their own analysis, or impose tax on gains the federal return deferred. A family that “solved” a federal problem and ignored a high-tax domicile has not solved the problem. They have relocated it.
Residency planning gets dragged into these conversations, often too late. Moving before a sale can change the state result. Moving as part of a prearranged seed, with the sale effectively baked into the fund’s redemption plan, invites the same substance arguments at the state level. I would not build a relocation around a transaction the IRS just called ineffective.
Ask the state question in the first meeting. It is less exciting than the federal deferral slide, and it is frequently the larger surprise.
Documentation Habits That Age Well
Good files are dull. An investment committee note that predates the contribution, a diversification worksheet with timestamps, a rebalancing policy that would apply to any shareholder, and broker statements showing the fund still holds the seed months later: that is the aesthetic the notice rewards. Slide decks titled “tax-free diversification” are the other aesthetic.
If you already used words like conduit, cleanup, or exit basket in client correspondence, do not delete them. Preservation matters more than cosmetics, and cosmetics are how people turn a civil exam into a worse conversation. Counsel can explain context. Counsel cannot explain a missing archive.
Comments on the notice are due October 28. Trade groups will ask for a brighter line on timing and for examples of acceptable seeds. Until those examples exist, the prudent reading is narrow. Match the thesis. Intend to hold. Do not schedule the unwind.
A Practical Close for Anyone Holding Appreciated Shares
The ruling does not raise your tax rate. It refuses a shortcut that tried to borrow a corporate nonrecognition rule for a personal diversification trade. Everyday fund shareholders are not the audience. Families and sponsors who built custom vehicles around contributed winners are.
Legitimate seeding of a fund that wants the assets, and means to keep them, is expressly left alone. Quick rotations into a materially different book are not. Between those poles, facts will govern, and the paper you already wrote will matter more than the paper you wish you had written.
If a pitch still promises diversification, liquidity, and no capital gain, ask which of the three it is willing to drop. The market versions that have survived for years each drop one. The version that drops none is the version officials just said does not work.
