The Lot You Sell First

The first sale after an initial public offering is often described in the past tense, as something that happened at the unlock, as though the unlock itself were the decision. The unlock is not the decision. The unlock is the deadline. The decision is which lots, in which order, by which identification method, against which losses, with what is left held for what reason. Almost none of this is set on the unlock date. Most of it is set, by default, weeks earlier — quietly, by the brokerage's standing rules, by counsel's draft of the 10b5-1 plan, by the cash needs the holder has not yet sized — and then executed in a window where the holder is asked to approve trades but rarely asked to choose the inventory the trades draw from.

A position that has built itself across three or four years of restricted-stock vests, each cohort acquired at a different market price during the path from late-stage private to public listing, is not a single position when the lockup expires. It is a stack of lots, each with its own basis, holding period, and embedded gain. The brokerage's default lot-identification method is first-in-first-out. For a holder whose vests began before the IPO and continued through the run-up to it, this default selects the oldest lots first — which is to say, the lowest-basis lots, which is to say, the lots with the largest embedded gains. Almost no holder would choose this method, asked on the merits. Almost every holder accepts it, because the question of how shares are identified at sale is one the brokerage answers by default rather than one the holder is asked.

This piece is about the choice that is being made by default. It is about the two structural forces that should actually be shaping the order — and that pull, importantly, in opposite directions. And it is about the three-tranche framework that organizes them into an apportionment a holder can write down.

The unlock as inventory decision

The standard treatment of an IPO lockup expiration is a calendar exercise. A date is identified in the offering documents, generally one hundred eighty days after pricing. Counsel reminds the holder of insider-window constraints, blackout periods, and disclosure obligations. A 10b5-1 plan is drafted in the weeks before the unlock to permit trading during periods when discretionary sales would be restricted. The plan specifies, in some detail, when and how much. It usually does not specify which lots.

The omission is consequential. A 10b5-1 plan that sells twenty percent of the holder's restricted-stock position over the first ninety days after unlock is a different plan, in after-tax terms, depending on whether those sales draw against the oldest lots or the newest. The market value sold is the same. The embedded gain realized is not. For a holder whose oldest pre-IPO vests have a basis of a small fraction of the post-IPO market price, and whose most recent vests have a basis close to or even equal to the price at unlock, the difference between FIFO and a deliberate specific-identification election can be the difference between a sale that realizes a sixty-percent gain and one that realizes a six-percent gain on the same number of shares.

A sale that draws against high-basis lots is not a sale that defers the gain. It is a sale that produces a different outcome from one that draws against low-basis lots. The deferred outcome, the one that involves not selling the highest-gain lots, depends on what is to be done with those lots instead. That question — what is held, by whom, and for what purpose — is the question the apportionment is built around.

Why the default is almost always wrong

There are two ways to evaluate which lots a holder should sell first, and the brokerage's default method is wrong under both.

The first frame considers the loss capacity the holder has available to offset the gain on the sale. A holder running a direct-indexing sleeve in parallel — the structure addressed at length in The Lockup You Sign For — generates realized losses through the routine harvesting of underperforming index components. Those losses constitute a budget. Against a defined loss budget, the sale that uses the budget most efficiently is the sale that realizes the largest gain per share within the budget — which is to say, the sale that draws against the lowest-basis lots. Under this frame, FIFO is doing the right thing by accident, and only for the holder whose loss budget is precisely sized to absorb the gain on the oldest lots. For everyone else, the holder either runs out of budget mid-sale, or finishes the sale with budget unused. Specific-identification is the lever that aligns the lots realized with the budget available, not a calendar.

The second frame considers what would happen if the highest-gain lots were not sold at all. Appreciated property held until the holder's death receives a step-up in basis to fair-market value at the date of death. The embedded gain accumulated over the holder's lifetime is eliminated for the heirs' federal-capital-gains purposes — a feature of the code, narrow in scope but unforgiving in its arithmetic. The implication is structural: the lots with the largest embedded gains are the lots whose sale carries the highest after-tax cost relative to holding, and the highest after-tax benefit, in expectation, from being held. Under this frame, FIFO is doing the opposite of the right thing. The holder who follows the default is realizing precisely the gains the code would later have forgiven, and is holding precisely the gains the code would have charged for.

The two frames produce different answers. Under the harvesting frame, the highest-gain lots are sold first, against the budget. Under the step-up frame, the highest-gain lots are held forever. FIFO defaults to selling them first, regardless. That this happens to coincide with the answer under the harvesting frame is incidental, and only optimal when the budget is exactly large enough to absorb the embedded gain on the lots FIFO selects. Outside that narrow case, the default is wrong; whether it is wrong by selling too much, too little, or the wrong tranche depends on the rest of the picture.

Frame one — the loss budget

The first frame treats the lockup unlock as an opportunity to realize gain against a defined loss budget that the holder has, deliberately or otherwise, accumulated.

A direct-indexing sleeve, sized appropriately to the household's broader portfolio, produces harvested losses at a rate that depends on market volatility, sector dispersion, and the sleeve's harvesting cadence. Over a typical year, a sleeve in the eight-figure range can generate harvested losses in the low single-digit percentages of its value, with meaningful year-to-year variance. The losses are a real budget. They sit on the holder's books as carryforwards if unused; they offset realized gains dollar-for-dollar against gains of the same character when applied.

The post-lockup sale, framed as a draw against this budget, becomes a question of efficiency rather than calendar. The holder asks: how much gain does this year's budget absorb, and which lots, sold in what order, fill the budget most precisely? A specific-identification election, made before the trade date, allows the holder to pair the harvested losses with the lots whose realization would produce the largest embedded gain per share. This is the harvesting frame at its sharpest: the sale is sized to the loss budget, and the lots are chosen for fit. A loss budget that absorbs the gain on the oldest pre-IPO lots — and only those — produces an after-tax outcome substantially different from one that absorbs only a fraction of that gain.

The companion piece on direct indexing — The Free Lunch Sitting in Your Losers — treats the mechanics of how the loss budget gets built. The implication for the post-lockup sale is the one this dispatch carries: the question of how much to sell is downstream of the question of what budget the holder has. The lockup expiration is not the calendar; the budget is.

There is a constraint worth naming. The character of the gain matters. Long-term capital losses offset long-term capital gains first; short-term against short-term. The pre-IPO lots, generally held longer than one year by the time of unlock, are usually long-term. The harvested losses from a sleeve held more than a year are also long-term. The pairing works. For a holder whose lots cross the long-term boundary — recent vests held under twelve months at the unlock — the character question complicates the apportionment, and the right answer is often to hold those lots into long-term status before selling, regardless of which frame governs the rest.

Frame two — the step-up at death

The second frame treats the highest-gain lots as the most valuable to keep, not the most valuable to sell.

The step-up in basis is the most consequential single feature of the federal capital-gains regime for high-basis-gap holders, and it is the one that most planning conversations underweight. Property held until the owner's death receives a basis equal to fair-market value at the date of death. The accumulated gain that produced the basis gap — sometimes the work of decades of holding through the pre-public and post-public lives of the position — is eliminated for the heirs' purposes. This is not a deferral. It is a forgiveness.

The arithmetic implication for the post-lockup holder is structural. Each highest-gain lot the holder sells is a lot whose forgiveness has been spent. The cost of selling that lot is the difference between today's after-tax proceeds and the after-tax value the lot would have produced if held to the step-up — which, for a lot with a ninety-percent embedded gain, in the limit, is the entire federal capital-gains tax on the gain. The cost of selling a low-basis lot today is, by this logic, never zero; it is the option value of the step-up the holder is forgoing.

The frame is most operative for older holders, holders in fragile health, or holders whose planning horizon legitimately includes a step-up event within a foreseeable window. For younger holders, the option value is real but heavily discounted by the long horizon to its realization, and by the regulatory risk that the step-up itself may be modified between now and then. Both populations exist in the post-IPO holder pool. The younger employee with a four-year vest schedule and a multi-decade career runway sits at one end; the founder-executive whose lockup expires after a long late-stage career sits closer to the other. The frame is not absolute. Its weight is a function of the holder's situation.

Adjacent levers operate within this frame. Lifetime-exemption gifts of appreciated stock to family members shift the basis problem without resolving it (the recipient takes carryover basis, not stepped-up basis, unless the donor dies holding the gifted property). Charitable transfers — outright gifts of appreciated stock to public charity, contributions to a donor-advised fund, contributions to a charitable remainder trust — remove the gain from the holder's recognition entirely, at the cost of removing the asset from the holder's estate. These are not step-up substitutes; they are step-up-adjacent tools that share the structural property of removing embedded gain from the holder's lifetime tax picture. They belong in the same conversation, and they sit beside the held-for-step-up position rather than instead of it.

The dispatch on inherited concentrated positions — The Step-Up That Isn't Yours Yet — treats the mechanics from the inheritor's side. The implication for the holder is the mirror image: the lots that are the largest planning opportunity for the inheritor are precisely the lots that should be considered, before sale, for whether the inheritor will be the one realizing them.

A question worth asking

Of the lots in the position today, which were chosen for sale by the holder, and which were chosen by the brokerage?

The two frames collide

A holder evaluating the post-lockup sale lives between these two frames, and the apportionment is the work of resolving them.

The holder under forty-five, with a multi-decade career runway, a meaningful continuing salary correlated to the position's outcome, and a household whose plan is not yet dominated by the position, generally sits closer to the harvesting frame. The step-up is too distant to dominate. The loss budget, by contrast, is immediate and renewable. The apportionment for this holder leans toward sale against the budget, with a smaller tranche committed to a long-horizon structure and a smaller tranche still held for the step-up.

The holder over sixty-five, with a household whose net worth is materially defined by the position, with planning that already contemplates a transfer event within the foreseeable horizon, generally sits closer to the step-up frame. The loss budget is real but not the largest lever. The apportionment for this holder leans toward less sale, more structure, and a substantial tranche held for the basis reset.

A third axis runs through both frames. A holder with sustained charitable intent has access to the gift-and-deduction structures — donor-advised funds, charitable remainder trusts, outright gifts of appreciated stock — that remove the highest-gain lots from the holder's recognition without requiring the holder to die holding them. For this holder, the second tranche of the apportionment expands. The question shifts from whether to sell or hold, to whether to sell, hold, or transfer.

The diagnostic that determines where on this spectrum a particular holder sits is the one introduced in When the Position Becomes the Plan. The post-lockup question is downstream of that diagnostic, not upstream of it. The lot-by-lot apportionment cannot be optimized in isolation; it sits inside the household's planning frame, and the frame is what determines which tranche should be largest.

The three-tranche framework

The apportionment that follows from the two frames is a three-part division of the position, not by percentage but by intent.

The first tranche is sold this year, into the holder's defined tax-friction budget. The budget is set by the harvested-loss capacity the holder has available and by direct cash needs. The lots sold are the highest-gain lots that the budget can absorb, identified by specific-ID at the trade date. This tranche is the answer to frame one. Its size is bounded by what the budget allows, not by what the calendar suggests, and not by what the brokerage's default selects.

The second tranche is committed, in the same window or in the months that follow, to a structure whose payoff arrives over a defined multi-year horizon. The candidates are the ones the publication has treated separately — the exchange fund, with its seven-year §721 lockup; the direct-indexing program, run as a deferral mechanism through accumulated carryforwards; the charitable vehicle that converts embedded gain into a deduction and a sustained philanthropic stream. Each does different work. The holder's selection among them depends on liquidity tolerance, philanthropic intent, and the holder's view of what the underlying position is likely to do over the structure's horizon. The mechanics are treated, in part, in the companion piece on exchange funds and direct indexing.

The third tranche is held, with deliberate reasons. The reasons are the step-up frame at its most concentrated form. The holder identifies the highest-gain lots, the ones whose forgiveness would be most consequential, and holds them — not as a failure of planning, but as the deliberate placement of the position's largest embedded gains inside the structure that will, in the holder's expected horizon, eliminate them. The size of this tranche is a function of the holder's age, health, charitable intent, and the household's planning frame. For some holders, it is the largest tranche of the three. For others, it is a small placeholder.

The apportionment, written this way, is the plan the unlock is asking the holder to make. It is not a single number. It is a budget — for sale, for structure, for hold — that the holder, counsel, and the broader advisory team execute over a period long enough that the unlock date itself becomes a milestone within it, not a deadline that defines it.

The 60–120 day window

Almost every consequential decision in this apportionment is made before the unlock date or in the trailing window after it.

In the sixty to one hundred twenty days before the unlock, the holder reviews and confirms the lot-identification method on the brokerage account. The default is FIFO, and the change to specific-identification, while administrative, must be on file by the trade date of the first sale. Standing instructions are reviewed and updated. Where a 10b5-1 plan is being drafted, the lot-identification method is incorporated into the plan's parameters, alongside the price triggers and volume limits that more typically receive attention. Counsel reviews insider-window calendars and blackout schedules, and the plan is structured to execute the apportionment within those constraints rather than around them.

In the unlock window itself, the apportionment is executed. The first-tranche sales are paired with the harvested-loss budget; the second-tranche structure is committed to; the third-tranche lots are identified and segregated, sometimes in a separately tracked account, so that subsequent activity does not inadvertently draw against them.

In the trailing window, the planning calendar shifts to the next tax year, and the apportionment is re-evaluated against any refresh-grant vests that have landed in the interim — a complication addressed in the companion treatment of withholding shortfalls, The 22% Problem. The position is not a static target. Each new vest extends the inventory, and the apportionment is calibrated to absorb the flow as well as the existing stock.

The 10b5-1 plan and insider-window constraints govern the timing of the trades. They do not govern the lot-identification method, the budget allocation, or the apportionment among the three tranches. Those decisions are the holder's, in principle. The brokerage default decides them, in practice, for holders who have not done the work to claim them back.

What this dispatch doesn't pull

Several threads run alongside the post-lockup apportionment and are addressed in separate pieces.

The constructive-sale rules of §1259 — which constrain how tightly a holder can collar a position with options without triggering immediate recognition — apply where the holder considers an options overlay rather than outright sale. They are their own subject. State-residence timing — particularly relevant where a holder is considering a move that would coincide with the unlock — sits with the Pillar C treatments of pre-sale planning and is its own subject. The 83(b) election, governing the treatment of restricted property at grant rather than at sale, is its own subject and applies primarily at acquisition, not at unlock. The interaction of incentive-stock-option exercise with the alternative minimum tax, of structural relevance to the holder population overlapping this one, is addressed in the Pillar D treatments. The comparative analysis of charitable structures — donor-advised funds, charitable remainder trusts, charitable lead trusts — and the holder's expected income trajectory under each, sits beside this dispatch as adjacent reading.

Each of these threads is meaningful, and each, for a particular holder, may turn out to dominate the apportionment. The dispatch you have read is the framing layer beneath them.

Where this lands

The lockup expiration produces, for most holders, a number on a brokerage statement and a calendar date. The number is not the plan. The date is not the plan. The plan is the apportionment that the holder writes between the two frames — the harvesting frame that says realize the highest gains into the budget, and the step-up frame that says hold the highest gains forever — and the three tranches into which that apportionment resolves.

The brokerage default chooses for the holder who does not choose. The default is, with high reliability, wrong; whether wrong by selling the wrong lots, or selling too many, or too few, or the wrong tranche, depends on the rest of the holder's picture. The work of the sixty to one hundred twenty days before the unlock is the work of reclaiming the choice from the default. The work of the unlock window itself is the execution of the apportionment that decision implies. The work of the trailing window is the integration of the next tranche of inventory into the same frame.

A position built across multiple years of restricted-stock vests is, at unlock, a stack of lots whose order matters more than the dollar value of the position suggests. The lot the holder sells first is the lot that defines the apportionment. The holder chooses, or the brokerage chooses. The unlock is the deadline for that choice. The apportionment is the plan it produces.