2026 has been the busiest year for section 1202 since Congress made the 100 percent exclusion permanent in 2015. It is the first full year under the expanded qualified small business stock (QSBS) rules enacted in the One Big Beautiful Bill Act (OBBBA), and practitioners have spent it working through what the changes actually mean for founders, employees, investors, and the trusts that hold their shares. Along the way, two academic papers put real numbers on what the incentive does, Treasury signaled discomfort with the most popular planning technique, a national accounting firm filed a 70-page request for guidance, and the financial press picked up the story.
Much of this post draws on five pieces of writing that shaped the conversation this year, and we refer back to them throughout:
With those introductions made, here are seven things that have dominated the conversation among QSBS advisors this year.
The biggest shift in day-to-day practice this year has been the realization that the acquisition date of a share now controls almost everything. Stock acquired on or before July 4, 2025 remains subject to the original rules, including the $10 million cap and the $50 million gross asset limit. Stock acquired after that date gets the $15 million cap, the $75 million gross asset limit, the new partial exclusions, and inflation indexing beginning after 2026.
As the first post-OBBBA financing rounds closed this year, advisors confronted cap tables with two or three "vintages" of QSBS outstanding at once. A founder who received shares in 2024, employees who exercised options this spring, and a Series B investor who funded in the fall cannot be analyzed as a group. The Miller article puts it plainly: practitioners should analyze each issuance independently rather than assuming all outstanding shares receive identical treatment.
The interaction rule has been a recurring surprise. A shareholder holding both pre- and post-OBBBA stock in the same issuer does not get a $10 million exclusion plus a separate $15 million exclusion. The post-OBBBA dollar limit is reduced by eligible gain already taken on the older stock. Firms that never tracked issuance dates at the share level began doing so this year, because clear data will be needed more than ever at exit.
OBBBA ended the all-or-nothing five-year cliff for new stock. Post-OBBBA QSBS held at least three years qualifies for a 50 percent exclusion, at least four years for 75 percent, and at least five years for the full 100 percent. No post-OBBBA stock has reached even the three-year mark yet, but founders, boards, and investors have already begun modeling the tiers into exit planning.
The research suggests they are right to. The Campello and Junqueira paper documented measurable "bunching" of venture exits at the five-year mark under the old rules, meaning investors were timing dispositions around the tax threshold. With three graduated thresholds instead of one, the timing calculus gets more nuanced, and an offer in year three or four now carries a real partial benefit to weigh against waiting.
A computational question also surfaced this summer. The RSM letter asks how the exclusion percentage and the dollar cap interact. Take a founder with zero basis who sells post-OBBBA stock for $18 million after three years. Apply the $15 million cap first and then the 50 percent exclusion, and $7.5 million is excluded. Apply the percentage first and then the cap, and $9 million is excluded. RSM reads the statute as requiring the cap to be applied first, but asked the Service to confirm. Until guidance arrives, advisors modeling partial exclusions have been showing clients both figures.
QSBS "stacking" is the practice of making completed gifts of QSBS to multiple nongrantor trusts before a liquidity event so each trust can claim its own shareholder-level exclusion. Because gifts preserve QSBS status and holding period, and because properly structured nongrantor trusts are separate taxpayers, a founder with four children could in theory add four additional $15 million exclusions to his or her own.
The strategy is legal under current law and became substantially more valuable after OBBBA. This year it also became visible. Bloomberg Tax reported on May 20 that Assistant Secretary for Tax Policy Kenneth Kies had told a conference audience that Treasury was working on guidance aimed at the technique. The Journal's June feature described Silicon Valley's enthusiasm for it and the IRS's discomfort. The RSM letter notes that Treasury and the Service are contemplating guidance on stacking, most likely under the anti-avoidance authority in section 1202(k), and asks that any rule be drawn narrowly enough to avoid sweeping in unrelated structures.
Practitioner advice in response has centered on substance. Section 643(f) already lets the IRS treat multiple trusts as one where the grantors and primary beneficiaries are substantially the same and a principal purpose is tax avoidance. Trusts should have different beneficiaries, different distribution standards, different trustees where possible, and documented non-tax reasons for existing. They should be funded well before a deal is on the table, since the anticipatory assignment of income doctrine applies once a sale becomes a near certainty. And every trust instrument needs a careful read for provisions under sections 672 through 677 that could inadvertently create grantor trust status, which Miller memorably calls the "unintentionally defective grantor trust."
Section 1202(b)(3) says a married individual filing a separate return gets half of the dollar cap. The statute says nothing about married couples filing jointly. Do they share one $15 million cap, or does each spouse have a separate cap for a combined $30 million?
With the cap now $15 million (up from $10 million), the stakes of this old ambiguity grew, and both the Miller article and the RSM letter treat it as an open interpretive issue. The argument for two caps rests on the long-standing principle that spouses are separate taxpayers even when they file jointly, on legislative history describing the limitation as applying shareholder by shareholder, and on the fact that partners in a partnership each get their own cap. The argument for one cap is structural: it would be odd for Congress to give a couple a smaller benefit for filing separately than for filing jointly, so the omission of joint filers from the half-cap rule looks like a drafting oversight.
RSM has asked Treasury to resolve the question. In the meantime, most advisors have been planning conservatively while telling clients with substantial gain that reliance on the two-cap reading calls for a formal tax opinion and a clear understanding of the audit exposure.
Section 1202(e)(3) excludes a list of service fields including health, law, consulting, financial services, and brokerage, along with any business whose principal asset is the reputation or skill of its employees. The statute does not define those terms, and the IRS has not issued private letter rulings on the active business requirement since 2024. That leaves practitioners with a handful of non-precedential rulings, one Chief Counsel Advice memo, and the statutory text.
The AI boom made this the year the problem moved from theoretical to unavoidable. Consider a platform that connects patients with physicians, bills the patients, and pays the doctors a fixed fee. Is that a software business or a health business? Consider an online marketplace that facilitates short-term real estate rentals. A 2021 private letter ruling found that an insurance agency performing substantial back-office work was not a "mere intermediary" and therefore not a brokerage. A 2022 Chief Counsel Advice memo reached the opposite conclusion for a real estate leasing website, treating it as a brokerage even though the section 199A regulations define brokerage narrowly as securities brokerage.
RSM's recommendations here cut in both directions. The firm asks Treasury to adopt the section 199A regulatory definitions of the excluded fields for section 1202 purposes, which would give taxpayers a clearer roadmap. But it also asks Treasury to look to the source of revenue rather than the form of the service relationship, so that a company whose value is tied to excluded services does not qualify simply because it delivers those services through contractors or a software layer. RSM further proposes an 80 percent bright line for the undefined "substantially all" holding period requirement. For now, Miller's guidance has been the working standard: labels rarely determine the outcome, and companies with novel models should document how they actually generate revenue well before a liquidity event.
The jump from $50 million to $75 million drew a lot of companies back into eligibility this year, and with them came a wave of misunderstandings. Clients hear "$75 million" and assume the test is a snapshot on the issuance date. It is not. The corporation's aggregate gross assets must not have exceeded the threshold at any time before the issuance, and must not exceed it immediately after the issuance counting the new money. Once a company crosses the line, later issuances cannot be cured by shrinking back below it, although earlier qualifying stock is unaffected.
Two wrinkles got particular attention. First, gross assets are measured by cash plus adjusted tax basis, not fair market value, which is favorable for software companies with little basis and large enterprise value. But when an LLC taxed as a partnership converts to a C corporation, the contributed assets count at fair market value. That rule can disqualify a conversion founders assumed would work, or, if value is still under the threshold, it can produce a very large ten-times-basis exclusion on post-conversion appreciation. Miller notes that a conversion at $74.9 million of value after July 4, 2025 could in principle support up to $749 million of excluded gain. Many LLC-to-C-corp conversions were run this year with exactly that arithmetic in mind.
Second, RSM flagged the phrase "immediately after" as undefined for staged financings. If three investors sign one purchase agreement and close on different dates, and the third closing pushes the company over $75 million, does the first investor's stock still qualify? RSM recommends applying ordinary step transaction principles, but acknowledges that many practitioners currently do not. Until Treasury weighs in, advisors have been urging companies to structure and document rounds with this question in mind.
OBBBA changed the numbers but left in place the mechanical rules that most often disqualify otherwise valuable stock. Four came up repeatedly this year.
Redemptions. Stock repurchases from the shareholder or related persons within two years before or after an issuance, or significant redemptions exceeding 5 percent of outstanding stock, can taint new issuances. RSM has asked whether dividend-equivalent redemptions, recapitalizations, and founder redemptions that merely return excess startup capital should count, and the answer today is unclear.
Contributions to partnerships. A partnership can distribute QSBS to a partner and preserve its status, but a partner who contributes QSBS to a partnership, including a family limited partnership formed for estate planning, generally destroys it. Miller describes this as a pitfall that continues to trap otherwise sophisticated taxpayers.
Tiered structures. Venture funds routinely hold QSBS through multiple partnership layers. Most advisors believe the exclusion flows through to the ultimate non-corporate partner, and Treasury has adopted a look-through rule for section 1045 rollovers, but no regulation confirms that result under section 1202(g).
Rollovers and reorganizations. RSM has asked whether a taxpayer who excludes $10 million of a $22 million gain must reinvest all $22 million or only $12 million to defer the remainder under section 1045, and whether stock received in a pro rata section 355 spin-off inherits QSBS status when no actual exchange occurs. Both were live questions for portfolio companies that restructured this year.
One reason these technical debates carried more weight this year is that section 1202 has become a much larger fiscal item. The Campello and Junqueira paper cites Treasury estimates that the program costs about $3 billion annually in forgone revenue, projected to reach $7 billion by 2035. Provisions of that size attract legislative attention, as the 2021 Build Back Better proposals demonstrated when they sought to cut the exclusion in half for high earners and nearly all trusts, retroactively.
The research published this year offers a counterweight. The Chen and Farre-Mensa paper found that the 2010 expansion increased firm births in eligible industries by roughly 10 percent and startup patenting by more than 20 percent, and that eligible startups expanded equity compensation and hired more high-skill talent away from incumbents. The Campello and Junqueira paper showed that venture investors became far more willing to fund pre-commercial companies after the 2009 and 2010 changes, and that the resulting portfolio companies failed more often but also reached higher valuations and were roughly twice as likely to become unicorns. In its September editorial, the Journal's editorial board cited that work as evidence that when it comes to capital gains taxation, the "something" being taxed is innovation itself.
None of this guarantees the current rules (as they are) will last. It does mean the policy conversation heading into 2027 is slightly better informed than it was a year ago.
