The Lockup You Sign For

There is a particular meeting that happens, with some frequency, in the year after a founder's first liquidity event, or in the years a senior executive's restricted stock concentrates faster than it can reasonably be sold. The portfolio statement now shows a single ticker accounting for sixty, seventy, sometimes eighty percent of the household's liquid net worth. The advisor opens with the obvious problem: the position is too large. Then the advisor presents two products as though they were the same kind of answer.

They are not. An exchange fund and a direct-indexing account address overlapping concerns through structurally different mechanisms, and they buy the holder different things at different costs. Most of the marketing material in circulation treats them as alternatives within a single decision frame — pick one, depending on minimum and preference — when in many cases they are tools for different problems entirely, and the cleanest answer is sometimes both.

This piece is about the choice as it actually presents itself, not as the deck does.

What an exchange fund actually does

An exchange fund — properly, a §721 fund, after the section of the Internal Revenue Code that governs its core mechanic — is a partnership that accepts contributions of appreciated stock without triggering recognition of gain at the moment of contribution. In exchange for the contributed stock, the holder receives a limited-partnership interest. The partnership, which holds the contributions of dozens or hundreds of similarly situated holders, presents to its participants as a diversified basket. The original concentrated position has been, for accounting and economic purposes, exchanged for a piece of a portfolio.

The key word in that arrangement is recognition, not taxation. The deferred gain has not been forgiven. It has been pushed forward in time, attached to the partnership interest the holder now owns, and ultimately to the diversified basket of securities the holder receives at the end of the lockup. Cost basis carries from the original contributed stock to the redemption package. A holder who contributed a single ticker at a fifteen-dollar basis and receives, seven years later, a basket of forty securities will hold those forty securities at the original carryover basis, allocated across the basket. Selling them will produce gain at that basis. The tax has been deferred, redistributed across more positions, and decoupled from the timing of any single sale — but it has not been resolved.

Two structural features matter for any holder evaluating the product. The first is that the partnership must hold at least twenty percent of its assets in qualifying illiquid investments — typically real estate, infrastructure partnerships, or similar — for the §721 treatment to apply. This is a requirement of the partnership's structure, not of the holder's portfolio, but it shows up in the holder's economic experience as a drag against the part of the basket that resembles the public-equity index the holder believes they bought into. The drag is built in. It is not negotiable. The second feature is the lockup: holders must hold their partnership interest for a minimum of seven years before redeeming it for a diversified basket of public securities, after which gain on those securities, when realized, will be at the original carried-over basis. Earlier redemption is generally permitted, but it produces, by design, the original concentrated position back, with the original gain intact and the original problem unsolved. The lockup is not a detail. It is the central thing the holder has agreed to.

Most exchange funds also require participants to qualify as Qualified Purchasers under the Investment Company Act, which generally means five million dollars or more in investments, exclusive of primary residence and certain related-party holdings. A handful of newer entrants — Cache Exchange Fund, launched in the early 2020s, is the most visible — operate under different exemptions and admit accredited investors at lower thresholds, with different fee structures and different qualifying-asset profiles. The mechanics are consistent across the category. The terms are not.

What direct indexing actually does

A direct-indexing account is, structurally, the opposite kind of solution. The holder does not contribute their concentrated position; the holder adds a separate sleeve of capital, often in cash, to a separately managed account that purchases the individual stocks comprising a target index — typically a hundred to four hundred names selected to track the S&P 500 or the Russell 1000. The concentrated stock remains where it was, untouched. The new sleeve, parallel to it, is the active piece.

What the new sleeve does, on an ongoing basis, is harvest losses. As individual positions in the index portfolio decline below their cost basis, the manager realizes those losses and replaces the stock with a similar-but-not-identical security that maintains overall index exposure without triggering wash-sale recognition. The realized losses accumulate as carryforwards on the holder's tax return. They become available to offset other capital gains the holder realizes — including, importantly, gains realized by selling pieces of the still-concentrated position elsewhere in the holder's portfolio.

The estimated tax alpha from this mechanism, in the academic and practitioner literature, generally lands somewhere between thirty and a hundred basis points per year, with the higher end reserved for the early years of the strategy and a meaningful decay over time as the portfolio's basis converges toward market value. By year ten of a single direct-indexing account, the harvestable losses have largely been mined out; the strategy's residual value is closer to the lower end of that range, sometimes lower. This is not a critique of the strategy. It is a feature of how the math works. The first dollar of the sleeve is the most productive; the last dollar, by definition, is not.

The structural point worth dwelling on, for the concentrated holder, is that direct indexing does not, on its own, solve concentration. It solves the cost of reducing concentration. A holder with a forty-percent position in a single ticker who adds a five-million-dollar direct-indexing sleeve will, after a year, have produced perhaps two or three hundred thousand dollars of harvested losses. Those losses can be applied against gain realized by selling, say, two or three percent of the concentrated position. Concentration falls; net tax does not, on the year, increase. The pace at which this works is gradual. It is not a one-step diversification.

A practical complication that the wider commentary tends to underweight: the wash-sale rule applies across the holder's entire household, including spousal accounts and retirement accounts. A direct-indexing manager harvesting losses on a stock the holder also holds in a 401(k), or that the spouse holds in a separate brokerage, may have those losses disallowed under Rev. Rul. 2008-5 and successor authority. Coordination across accounts is the unglamorous operational requirement on which the strategy's value, in practice, depends.

The lockup, looked at honestly

The seven-year lockup of an exchange fund is the feature most often discussed and the feature most often misunderstood. It is described, in the marketing material, as the cost of admission. It is also, in another light, a substantial illiquidity assumption that the holder may or may not be able to honor across the time horizon in question.

Seven years is long. It crosses a recession with a high degree of probability. It crosses, often, a child's college funding window, a down payment on a different house, a divorce, a serious illness, or a separate liquidity event the holder did not anticipate at the time of contribution. Early redemption is permitted in most fund structures, but it returns the original concentrated position with the original gain intact — a non-result. A holder who contributes ten million dollars of concentrated stock to an exchange fund in May of 2026 and finds, in December of 2030, that they need eight million dollars in cash for a real-estate purchase has not solved the problem; they have moved the timing of the problem and, in many fund structures, paid four years of management fees for the privilege.

The lockup also has a less-obvious consequence on the tax side. A holder whose circumstances change during the seven-year window — a move to a different state, a marriage or divorce, the death of a spouse — finds that the partnership interest sits in the same place regardless. State tax planning that depends on residence at the time of recognition, for example, becomes harder to deploy when the recognition event is fixed at the partnership's choosing rather than the holder's. Certainty on diversification is purchased with flexibility on timing, in both directions.

None of this is an argument against exchange funds. It is an argument for understanding what the lockup is purchasing and what it is, by definition, not.

The wrinkle most decks skip

The single largest analytical gap in the wider commentary on this comparison is what happens at the holder's death. The estate-planning treatment of the two structures diverges in ways that can outweigh, by an order of magnitude, the differences in tax alpha, fee, and lockup that the marketing material centers.

A direct-indexing portfolio held until the holder's death receives a step-up in basis on every individual security. Each of the three hundred-odd lots that the manager has accumulated and harvested over the years resets to its date-of-death value in the hands of the heir. The harvested losses, the carried-over basis from earlier sales, the carefully sequenced replacement-security trades — all of it resolves, cleanly, in a single moment, at zero remaining federal capital-gains liability on the appreciation that preceded death.

An exchange-fund partnership interest held until the holder's death also receives a step-up at the partnership-interest level. But the inside basis of the partnership — the basis the partnership itself carries on the underlying contributed stocks and on the diversified holdings it produces at distribution — is generally not adjusted automatically by the holder's death, absent a §754 election by the partnership and the §743(b) basis adjustment that follows. Whether and how that election is in place, and whether it benefits the heirs uniformly, depends on facts the holder typically does not control. Heirs who receive the partnership interest at step-up and then redeem shortly thereafter may discover that the step-up applies to the interest they sold, but not, in the same clean way, to the underlying gain that flows through to them on the distribution side. The wider commentary tends to glide past this. It is not a small detail.

The implication is that, for an older holder, or a holder with a meaningful probability of mortality inside the lockup window, the two structures may not produce the same after-tax outcome to heirs. Direct indexing, which on a pre-mortality basis offers more modest tax alpha, may produce a substantially better post-mortality outcome. For younger holders with multi-decade horizons, the calculus runs the other way. Age is, in this comparison, often a more decisive variable than size of position. The deck does not usually mention that.

A question worth asking

If the holder were to die before the seven-year lockup ended, who inherits the partnership interest, and on what basis do they sell?

The hybrid case

It is worth saying directly: the choice between an exchange fund and a direct-indexing sleeve is not always a choice. For holders at the upper end of the size range described by this piece — concentrated positions in the eight-to-fifteen-million-dollar range, supported by additional liquid net worth — the cleanest planning structure often uses both, in different roles.

The exchange fund, in such a structure, addresses the immediate diversification problem on a substantial portion of the concentrated position. The direct-indexing sleeve, funded with the holder's other liquid capital, runs in parallel and produces the harvested losses against which the remaining concentrated stock can be sold over time, at a tax cost reduced by the harvest. The two strategies are not in tension. They are doing different work. The exchange fund handles the part of the concentration the holder cannot afford to wait to diversify; the direct-indexing sleeve handles the part that can be sold in pieces, more flexibly, with the tax friction managed.

The fee structure of running both is heavier than running either alone, and the operational complexity is meaningfully greater. For a holder whose concentration sits at the lower end of the range described — say, two to four million dollars in a single name — the hybrid is usually overkill, and a single direct-indexing sleeve, run alongside a deliberate quarterly sell-down of the concentrated position, is closer to the right answer. For a holder whose concentration is at the upper end and whose horizon includes mortality risk inside the lockup window, the hybrid is often, on the math, what the situation calls for.

The point is not that there is a right answer. The point is that the question, framed as a binary, tends to produce a worse answer than the question framed as an architecture.

What this piece doesn't pull

Several threads connected to this comparison have been left for separate pieces. Exchange-fund fee schedules vary substantially by provider and by minimum, and the comparison across providers is its own subject — not least because the published fees are often only part of the economic cost. The treatment of qualified-covered-call writing and prepaid variable forwards as additional paths for concentration management involves a different set of mechanics, including the constructive-sale rules of §1259, and is addressed elsewhere. State-specific overlays — California's non-conformity to certain federal partnership treatments, for instance, or New York's residency rules during the lockup — interact with both structures in ways that depend on the holder's residence trajectory and are worth their own treatment.

Each of these is substantive enough to deserve its own piece. The point of this one is not to map every consideration in the concentration-management decision, but to surface the structural differences between two products that the wider commentary tends to flatten.

Where this lands

A holder who has read this far with a meaningful concentrated position is, ideally, holding stock that has not yet been committed to a structure. The decisions described above mostly compound forward — a contribution to an exchange fund is durable for seven years, a direct-indexing sleeve takes years to mature into its tax alpha — and the room to evaluate them is widest before the deck has been chosen, narrowest after the contribution form has been signed.

The Check on this site is not a concentration audit. It is shorter, six questions, scoped to the kinds of tax drag that show up across an investment portfolio rather than to the structural decision of which product to pair with a single large position. For holders evaluating the choice described above, the questions there are upstream of this one. They are about the rest of the portfolio, the places where rate mismatches tend to live, and the long compounding question of how much of the rest of it will be lost to the tax it doesn't appear, on first glance, to owe.