The §1202 exclusion is granted per taxpayer, not per family. Gifting stock before a sale is how a founder turns one exclusion into several. The strategy is exactly what Congress wrote — and it comes apart on a short list of operational mistakes. This paper covers both.
Section 1202 lets a non-corporate taxpayer exclude capital gain on the sale of qualified small business stock — provided the company passes the size and active-business tests and the stock is held long enough. For QSBS issued after July 4, 2025, the per-issuer cap is the greater of $15 million or 10 times basis. For stock issued earlier, it is the greater of $10 million or 10x basis. A full exclusion requires a holding period of more than five years; under the post-OBBBA regime, partial exclusions vest earlier — 50% at three years, 75% at four.
The cap is the engine of everything that follows, because of one word: it is per taxpayer. Each separate taxpayer who holds QSBS gets a full cap of their own. A spouse has their own. An adult child has their own. A properly built non-grantor trust is its own taxpayer — and has its own.
This paper assumes the stock already qualifies as QSBS. If that is not yet settled, start with the front-door primer — Does Your Company Qualify for QSBS? — and come back. Everything below is about multiplying an exclusion the stock has already earned.
Gifting is the mechanism that adds taxpayers before a sale. Fund a non-grantor trust for your children, and the trust brings a second exclusion — a non-grantor trust is a separate taxpayer beyond dispute. Add a second trust on a separate beneficiary line, and you have a third. Nothing in the statute caps the benefit at one exclusion per family. Congress could have written that limit. It did not.
A gift to a spouse looks like the easiest add of all — one transfer, no trust. It is also the least settled question in this area, and it is worth pausing on, because the common objection that "a married couple is really one taxpayer" is not wrong so much as unresolved. Section 5 takes it head-on. The short version: build the plan on trusts, and treat a spousal second exclusion as upside, not foundation.
In late 2021, The New York Times ran a feature under the title "A Lavish Tax Dodge for the Ultrawealthy Is Easily Multiplied." Its lead example was Roblox founder David Baszucki. As reported, Baszucki, his wife Jan Ellison Baszucki, and family members held Roblox shares across a series of trusts as the company approached its 2021 public listing — each trust positioned to claim its own §1202 exclusion. The Times counted the structure as a multiplied loophole.
The framing was wrong on the law. Section 1202 grants the exclusion to each taxpayer. Transferring shares to additional taxpayers before a sale is not an exploit of the code — it is the mechanism the code leaves open. The reporting described prudent, statute-driven planning and dressed it as a dodge.
But read past the headline and the example teaches the real lesson. The Baszucki stack was not assembled the week before liquidity. The shares moved into trusts years before the listing, through real transfers, into separate trusts. That is precisely why it held. The size of the stack is not the takeaway. Timing and substance are the takeaway — and they are the two things a late, sloppy imitation gets wrong.
The transfer mechanics live in §1202(h): when QSBS is transferred by gift, the recipient is treated as having acquired the stock the same way, and at the same time, as the donor. The recipient steps into your shoes. Four elements make a gift do what it is supposed to do.
The gift has to be of the actual shares, completed before the sale becomes binding. Wait until after the deal closes and you are gifting cash. Cash carries no QSBS character and no separate exclusion — you have simply divided your after-tax proceeds. The multiplication lives in the stock, so the stock has to move first.
Under §1202(h), the recipient inherits your acquisition date and your QSBS character. If you have held three years, the recipient is treated as holding three years — the clock does not restart. The recipient also takes your carryover basis, not a stepped-up basis. This is the opposite of what happens at death, where heirs get a basis step-up to fair market value. A lifetime gift trades the step-up for the tacked holding period and the extra exclusion. For QSBS headed toward a large exclusion, that is usually the trade you want — but it is a trade, and it should be made on purpose.
An adult child or a non-grantor trust is, beyond dispute, a separate §1202 taxpayer. A spouse can hold QSBS too — but whether a married couple gets a second cap on a joint return is genuinely unsettled, and Section 5 is devoted to it. In every case, "separate" has to be real. The certificate is reissued, the cap table is updated, voting and economic rights actually pass, and the recipient behaves like an owner. A transfer that exists only on paper invites the IRS to disregard it.
The single most important fact is the gap between the gift and the sale. Months — ideally a year or more — should separate the transfer from any letter of intent. As the deal gets closer, two doctrines wake up: assignment of income and the step-transaction doctrine. Both ask whether the gain had already ripened to the founder before the stock moved. Distance is the defense.
Numbers make the lever concrete. Take a founder with near-zero basis, a company that clears every QSBS test, and a projected gain of about $60 million at sale. Assume post-7/4/2025 stock with a $15M per-taxpayer cap and a holding period past five years. The figures below build only on taxpayers whose separate status is beyond dispute — the founder and non-grantor trusts — so the result does not depend on the unsettled spousal question in Section 5. The figures are illustrative and rounded; they are not a projection of any real outcome.
| Structure | QSBS taxpayers | Gain excluded | Taxable gain | Rough federal tax @ 23.8% |
|---|---|---|---|---|
| No planning — founder holds all | 1 | $15M | $45M | ≈ $10.7M |
| Gift a third to a children's non-grantor trust | 2 | $30M | $30M | ≈ $7.1M |
| Gift to two non-grantor trusts (one per child / line) | 3 | $45M | $15M | ≈ $3.6M |
Three taxpayers shelter $45 million instead of $15 million. The delta against the no-planning baseline is roughly $7 million in federal tax on this set of facts — before any state-level effect, which can widen the gap sharply in a high-tax state and disappear entirely in a state like California that does not recognize the exclusion at all. The 23.8% rate is the long-term capital-gains rate plus the 3.8% net investment income tax; the planning operates against that rate, not the ordinary rate.
Two caveats sit on top of the table. First, gain that is not excluded under the post-OBBBA tiers at a three- or four-year hold is taxed at 28%, not 15/20% — which changes the math of when to sell. Second, the table assumes each trust is a genuine separate taxpayer. If it is not, the columns collapse back to one. A gift to a spouse may add a fourth exclusion on top of this — but that one turns on the unsettled question in the next section, so it is left out of the figures above on purpose.
One move deserves a hard look before anyone leans on it: gifting QSBS to a spouse to claim a second exclusion. It is the simplest stack to picture — one transfer, no trust — and the least settled in the law. A common, respectable view among advisors is that a married couple gets only one exclusion, full stop. If that was your instinct — that spouses filing jointly are really "one taxpayer" — you are in good company, and you are not wrong so much as ahead of an unresolved question.
A gift between spouses is a nonrecognition event,1 and the recipient spouse tacks the donor's holding period and QSBS character.2 So a spouse can hold QSBS. The contested point is narrower: on a joint return, does each spouse carry a separate per-issuer cap, or do they share one?
Section 1202 caps the exclusion for "the taxpayer."3 It says nothing about joint returns. The only provision that addresses married couples halves the cap for a spouse who files separately.4 Everything beyond that is inference.
"Taxpayer" is defined as any person subject to tax;5 a joint return is two taxpayers on one form, not two people merged into one. And the separate-return haircut is hard to explain unless joint filers get two full caps — why halve something that was never doubled? The IRS has read an analogous "per taxpayer" threshold as applying per spouse in another part of the Code.6 On this reading, each spouse who owns QSBS claims a full exclusion.
The conservative reading runs the other way. A joint return is a single taxable unit — one income, one liability.7 Married-filing-separately haircuts are a routine penalty default across the Code (the AMT exemption and several credits work the same way), so the halving need not signal doubling at all. And no regulation, ruling, or case blesses two exclusions. Advising on eight figures, many practitioners treat the cap as shared. This is a defensible view — not a mistaken one.
Here is the part that matters for a lawyer: there is no case on point. No court has decided whether spouses filing jointly get one §1202 cap or two, and the IRS has issued no regulation, revenue ruling, or private letter ruling resolving it. The case law that touches spousal gifting is the assignment-of-income line8 as applied to pre-sale transfers in Estate of Hoensheid9 — relevant to a late interspousal gift, but silent on the cap. Anyone who states this question as a flat yes or no is overstating their confidence in both directions.
Community property. In the nine community-property states — Arizona, California, Idaho, Louisiana, Nevada, New Mexico, Texas, Washington, and Wisconsin — stock bought with community funds is already owned one-half by each spouse, which independently strengthens the separate-interest reading. It is untested, and California does not recognize the §1202 exclusion at all. Penalty posture. Claiming two exclusions on a joint return is a position with at least reasonable basis; whether it rises to "substantial authority" is debatable — and that distinction governs accuracy-related penalty exposure if the IRS prevails.10
This is exactly where planning stops being math. On an unsettled question, the technical answer — the science — is that the two-exclusion position is reasonable and likely worth claiming. Whether a given client should is a different question, and it is the art: their tolerance for carrying a contestable position on an eight-figure return, and the regret they would feel either way — regret at a deficiency and interest if the IRS prevails, or regret at having left millions unexamined because the question looked hard. Advisors test a client's risk tolerance for investments as a matter of course; the tolerance for tax risk deserves the same deliberate read. The Tax Risk Profile is built for that conversation — and it loops back to the Defensibility Checklist, so the appetite for risk and the work required to support it are read together.
The strategy is settled. The failures are operational, and they repeat. Here are the six that turn a multiplied exclusion back into a single one — or into a fully taxable sale.
| # | Failure | What goes wrong |
|---|---|---|
| 1 | Gifting too late | The stock moves after the sale is "practically fixed." Under Hoensheid, the gain is assigned back to the founder and taxed to them — trust or no trust. |
| 2 | Gifting cash, not stock | Waiting until after close. You have divided after-tax proceeds; there is no second QSBS holder and no second exclusion. |
| 3 | Using a grantor trust and expecting a new exclusion | A grantor trust's income is taxed to you. It is not a separate taxpayer, so it does not add an exclusion. Stacking requires a non-grantor trust. |
| 4 | Cloned trusts | Same grantor, same beneficiaries, same terms, funded together. §643(f) lets the IRS treat them as one taxpayer — and the whole multiplication evaporates. |
| 5 | No contemporaneous record | No qualified appraisal, no gift-tax return with adequate disclosure, no issuer attestation. Ju and Leto put the burden on the taxpayer; reconstruction under audit does not work. |
| 6 | The issuer never qualified | Gross-assets test, active-business test, redemption history, or original issuance fails at the company level. Gifting non-QSBS multiplies nothing. |
This one catches careful people. Grantor trusts are common, useful estate-planning vehicles — but for income-tax purposes the grantor is the trust. Its gain is your gain. That is exactly why it cannot add a §1202 exclusion: there is no second taxpayer. The stacking vehicle is a properly structured non-grantor trust, which files and is taxed in its own right. Confusing the two is the difference between three exclusions and one.
A non-grantor trust is the workhorse of multi-exclusion planning because it is a separate taxpayer with its own cap. The risk that travels with it is I.R.C. §643(f), which lets the IRS consolidate two or more trusts into a single taxpayer when two things are both true: the trusts have substantially the same grantor and substantially the same primary beneficiaries, and a principal purpose of setting them up was avoiding federal income tax.
The defense is to make the trusts genuinely different where it counts — distinct primary beneficiaries, independent trustees, different distribution standards, and a real non-tax reason each one exists (dynasty planning, asset protection, GST allocation, blended-family equalization). One trust per child, each built and administered as its own entity, is a far stronger posture than a set of near-identical trusts funded on the same afternoon.
The §643(f) gate, the adjacent attack vectors, and the structural mitigants that clear them are the subject of a companion paper. If you are weighing a multi-trust stack rather than a single gift to a spouse or child, read it next: QSBS Multi-Trust Stacking — A Defensibility Framework.
If the exclusion is examined, the review is methodical. An examiner works through the company-level qualification, the original-issuance history, the holding period and which regime applies, the gift itself and its timing, the trust structure under §643(f), and — running through all of it — the documentation. Two recent decisions, Ju v. United States and Leto v. United States, both denied the exclusion on substantiation and original-issuance grounds and put the burden squarely on the taxpayer.
The practical implication is simple and unforgiving: the contemporaneous record is the case. A qualified appraisal at the time of the gift, a gift-tax return with adequate disclosure, an issuer attestation of QSBS status, reissued certificates, and trustee minutes are not paperwork for its own sake. They are the evidence that the gift was real, that it happened when you say it did, and that the company qualified. Built at the time, they win the audit. Reconstructed afterward, they rarely do.
That is what the QSBS Defensibility Checklist is for. It walks every element an examiner tests — issuer qualification, original issuance, holding period, the gift and its timing, trust structure, and the record — in the order the audit runs, so you can see exactly where a given situation is strong and where it is exposed before counsel is engaged.
Gifting QSBS is not aggressive and it is not a loophole. It is the plain operation of a statute that grants the exclusion per taxpayer. For a founder with an exit on the horizon and gains that outrun a single $15 million cap, it is one of the most efficient moves available — and it is fully available only before the deal is in motion.
What separates a stack that survives from one that unwinds is not the idea. It is the execution: transfer the stock and not the cash, time it well clear of the sale, use a real non-grantor trust where a trust is warranted, keep the trusts genuinely distinct, and build the record as you go. Get the sequence and the substance right, and the multiplier is exactly what the Roblox example showed it can be. Get them wrong, and the IRS has a well-worn path to collapse it.
Citation format follows The Bluebook. Statutory cites are to the Internal Revenue Code (26 U.S.C.). No regulation, ruling, or judicial decision squarely resolves the joint-return cap question; secondary sources are identified as such.
Start with the QSBS Defensibility Checklist — the audit, element by element, in the order an examiner runs it. Then run the QSBS Stacking Review, a free educational issue-spotter that reads your facts into a lane and surfaces the gimmes and gotchas specific to your situation. Both are educational tools, not advice.
Not advice. This page is published for general educational and informational purposes. It does not constitute legal, tax, accounting, financial, or investment advice for any specific person or situation. Examples and dollar figures are illustrative, not personalized recommendations. The strategies discussed involve complex rules, fact-specific application, and material residual risk.
Not a law firm. Socrates Crayon is an educational resource. It is not a law firm, registered investment adviser, broker-dealer, or accounting firm. No attorney-client, fiduciary, broker-dealer, or other professional relationship is created by use of this page, and no tax or legal opinion is rendered.
Consult qualified counsel. Implementation of any strategy referenced on this page requires written analysis by qualified tax and legal counsel applied to the specific facts of the matter. Tax and trust law are subject to change; cited authorities may be superseded. References to §1202, §1202(h), §643(f), the OBBBA (Pub. L. No. 119-21), and case law reflect publicly-available information current as of June 2026.
Third-party references. The reference to reporting by The New York Times and to the Baszucki family describes publicly-reported facts for educational illustration. It does not imply any endorsement, affiliation, or that any named person or entity used Socrates Crayon or holds any view stated here.