What Is a 351 Exchange? Moving a Concentrated Portfolio Into an ETF

Every adviser has this client. Twenty years of holding, a handful of positions that did extremely well, and a cost basis so low that selling anything triggers a bill large enough to freeze the whole conversation. The portfolio is dangerously concentrated and everybody knows it. Nobody can afford to fix it.

The 351 exchange is a route out of that room. It moves the portfolio into an ETF without a taxable sale, and it has gone from an obscure corner of the tax code to something a great many advisers are now asking about.

It is also badly explained almost everywhere, usually as a loophole or a trick. It is neither, and the two qualification tests are arithmetic, which means you can screen a portfolio for eligibility in an afternoon before anyone spends real money on the question.

It is not new, and it is not an ETF rule

Section 351 has been in the Internal Revenue Code for decades, and it was not written with funds in mind at all.

Raymond Holst, a tax attorney at Practus who has worked on these transactions, gave the definition on Behind the Ticker in one sentence: 351 "is simply the provision in the Internal Revenue Code, which is used to create a corporation."

That is the whole of it. When people form a company and contribute property to it in exchange for shares, Section 351 is what stops that contribution from being treated as a sale. Two founders put in a building and a patent, they receive stock, and neither of them recognises gain on the way in.

Holst walked through the ordinary case: "We both contribute that into the corporation exchange for corporate shares," and if the contributed property carries embedded appreciation, "you have $25 of built-in gain in that property that you’re contributing." The gain does not disappear, it travels with the shares you receive in the form of carried-over basis.

An ETF is usually a series of a statutory trust, and it is treated as a corporation for federal income tax purposes. Contributing securities to it in exchange for its shares is the same transaction the code has always described. What changed was not the law but the willingness of fund administrators, custodians and issuers to actually run it, and the arrival of enough new ETF launches to give people somewhere to go.

Holst summarised the application carefully, and the hesitation in his phrasing is worth preserving: "I wouldn’t use the word play, but it’s a mechanism by which you can seed an ETF without the recognition of gain for your clients."

Test one: control

The first requirement is the one that makes this a group activity rather than a private arrangement.

Holst stated it directly. The rule "says the transferors, meaning the people putting property into the corporation, have to" "own 80% vote and value of the corporation after the transfer." The statute itself, at Section 368(c), measures control as 80% of the combined voting power and 80% of the number of shares of every other class, rather than 80% of value. Vote and value is the working shorthand.

Immediately after the exchange, the people who contributed have to control at least eighty percent of the entity. In practice this shapes the whole mechanic. These transactions happen at a fund’s launch, or at a defined contribution date, with a group of contributors coming in together against a small amount of outside seed capital. You cannot walk up to a five year old ETF with a billion dollars in it and contribute your portfolio, because you would be nowhere near eighty percent of it afterward.

That is why every 351 exchange you read about is tied to a specific fund and a specific date.

Test two: diversification

The second test is where most portfolios actually fail, and it is pure arithmetic.

Section 351 will not shelter a contribution to an investment company where the transfer results in diversification. The practical reading is inverted from how it sounds: if the portfolio you are contributing is already diversified by the code’s definition, the transfer does not achieve diversification for you, and the exchange qualifies. If your portfolio is concentrated, contributing it into a diversified pool would achieve diversification, and the shelter is unavailable.

The definition has two limbs, both measured on the portfolio being contributed:

- No more than 25% of the value in any single issuer.

- No more than 50% of the value in the five largest issuers combined.

Holst gave the second one on air: "the top five holdings of the securities that the client has can’t be greater than 50%."

This is the part worth internalising, because it inverts the intuition every adviser brings to the conversation. The client with ninety percent of their wealth in one stock, the client this sounds designed for, is the client who does not qualify. A portfolio has to already be reasonably spread out to make the trip.

Government securities do get special treatment in the test. They count toward total assets in the denominator but are not treated as an issuer’s securities in the numerator. That treatment is expressly withdrawn if the government securities were acquired in order to meet the tests, and cash and cash items are excluded from total assets altogether. So pre-positioning a portfolio to pass is not a route the regulation leaves open. Any restructuring ahead of a contribution is a conversation for a tax adviser working with the specific holdings, and it is not something to plan from an article.

What you gain on the other side

Deferring one tax bill is worth something. What makes the destination worth reaching is what happens afterward.

Holst made the point that most coverage skips: "With respect to ETFs, ETFs can generally rebalance their portfolios using technology in the marketplace that it isn’t a taxable event."

Once the securities are inside an ETF, the in-kind creation and redemption machinery that makes ETFs tax efficient applies to them. The manager can reshape the portfolio over time without generating distributions the way a taxable brokerage account would. So the exchange is not only a one-time deferral. It moves the assets into a structure where ongoing management is far cheaper in tax terms than it was outside.

The gain itself is still there. Basis carries over into the ETF shares you receive, and when you eventually sell those shares you will recognise it. You have deferred, and you have bought the ability to diversify and to be managed in the meantime.

What it costs to actually do

The tests are arithmetic. The execution is not.

Holst was blunt about the operational side: "you have to make sure your custodians are at the plate, the fund administrators are at the plate, your advisors are at the plate, because it is a huge information exchange."

Every contributed lot needs its cost basis documented and transferred. Multiple custodians have to deliver securities on the same date. The fund’s administrator has to accept and record all of it. Any one party who is not organised can delay or break the transaction, and the deadlines are real.

There is a further constraint on what can be contributed at all. Holst noted that "what assets the ETF can hold will depend upon the market maker for the ETF because you do in-kind create redeems." A security that cannot move through the creation and redemption process cleanly is a problem for the fund regardless of what the tax code permits.

The practice has spread beyond specialists. Raymond Bridges, describing his own conversations with advisers, noted that "if you’re just an RIA out there and you have these concentrated positions in Nvidia or AMD or any of this stuff, you can do 351 exchanges into an ETF."

The honest summary

A 351 exchange contributes appreciated securities into a newly formed or newly seeded ETF in return for its shares, with no gain recognised on the way in, provided the contributors control eighty percent of the fund afterward and the contributed portfolio already satisfies the twenty five and fifty percent diversification limits.

It defers rather than forgives, it requires a specific fund and a specific date, and it demands more operational coordination than most transactions an adviser runs in a year.

Run the two percentages against the client’s holdings before you do anything else. That screen takes ten minutes and it tells you whether the rest of the conversation is worth having. It is a screen and not the answer: the diversification test is applied transferor by transferor, boot rules apply, and whether the receiving fund is an investment company for this purpose is its own question. Those are for counsel. The percentages are for you.

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