Most founders learn what a non-grantor trust is the week before a liquidity event — when it's already too late to use one.

A non-grantor trust is one of the most powerful planning tools in the U.S. tax code. It's also wrapped in jargon — grantor, non-grantor, NGT, ING, complex, simple, defective, intentionally defective — that obscures what the structure actually does. The short version: a non-grantor trust is a trust the IRS treats as a separate taxpayer from the person who created it. That single fact is what unlocks QSBS stacking, separates assets from your taxable estate, and shields what you've built from creditors.

This guide explains the structure plainly. We cover how a non-grantor trust differs from a grantor trust, why founders use them for QSBS exclusion stacking, when the structure works (and when it doesn't), and how a non-grantor trust compares to the IDGT and SLANT alternatives you've probably heard about. By the end, you'll know whether this vehicle belongs in your pre-exit planning — and what to ask the lawyer drafting it.

What is a non-grantor trust?

A non-grantor trust is a trust that the IRS treats as its own taxpayer. It files its own income tax return (Form 1041), has its own employer identification number (EIN), and pays tax on its own retained income at trust rates.

The simplest way to keep grantor and non-grantor trusts straight: a grantor trust is taxed to the person who created it. A non-grantor trust is taxed to itself. Everything else — asset protection, QSBS stacking, estate-tax separation — flows from that one distinction.

That distinction sounds technical. In practice, it's the difference between adding a new taxpayer to your tax-planning toolkit and not.

How a non-grantor trust is created

A non-grantor trust is created by drafting a trust agreement that avoids the grantor-trust trigger powers in IRC §§671–679. Get the drafting wrong and you have a grantor trust by accident.

The most common drafting traps that turn a "non-grantor" trust into a grantor trust:

  • Retained power to revoke or amend the trust (§676)
  • Power to substitute trust assets for assets of equivalent value (§675(4)(C)) — the silent killer in most contaminated trusts
  • Power to add or change beneficiaries
  • A retained income or remainder interest (§677)
  • Certain administrative powers (§675), including the power to deal with trust assets without adequate consideration
  • Naming the grantor (or grantor's spouse) as trustee with discretionary distribution power

If any of those triggers are present, the trust is a grantor trust regardless of what the document is titled. This is why having an experienced tax attorney draft the agreement matters more than the marketing language on the trust company's website.

Common types of non-grantor trusts

A few non-grantor variants come up in founder conversations:

  • Standard non-grantor irrevocable trust (NGT) — the workhorse for QSBS stacking and estate planning
  • SLANT (Spousal Lifetime Access Non-Grantor Trust) — a non-grantor trust where the founder's spouse is a beneficiary, preserving family liquidity while keeping the structure non-grantor
  • ING trust (Incomplete Non-Grantor) — used to sidestep certain state income taxes; a different toolkit
  • Charitable lead non-grantor trust — pairs philanthropy with QSBS stacking

This article is about the standard NGT. The SLANT comparison is in the trust-vehicle section below.

Grantor vs. non-grantor trust: the comparison that actually matters

Most "grantor vs. non-grantor" articles you'll find online are written for general estate planning. Here's the version that matters if you have founder stock.

  • Who pays income tax. Grantor trust: grantor pays on personal 1040. Non-grantor trust: the trust itself pays on Form 1041.
  • Tax ID. Grantor trust: uses the grantor's SSN. Non-grantor trust: has its own EIN.
  • Federal income tax bracket. Grantor trust: grantor's individual rates. Non-grantor trust: compressed trust rates — top 37% kicks in at ~$16,000.
  • Counts as a separate QSBS taxpayer. Grantor trust: no. Non-grantor trust: yes.
  • Asset protection from the grantor's creditors. Grantor trust: limited. Non-grantor trust: strong (if irrevocable and non-self-settled).
  • In the grantor's taxable estate. Grantor trust: often yes. Non-grantor trust: no (if structured properly).
  • Grantor control. Grantor trust: higher. Non-grantor trust: minimal — independent trustee.
  • Typical use. Grantor trust: living trust, or dynasty trust where the grantor wants to pay tax. Non-grantor trust: QSBS stacking, asset protection, family wealth transfer.

Three of those rows do most of the work. Walk through them in order.

Income tax: where the surprise lives

Trust tax brackets are highly compressed. In 2026, a non-grantor trust hits the top 37% federal bracket on retained income above roughly $16,000. An individual filer doesn't hit that bracket until ~$640,600. That's not a typo.

That sounds like a deal-breaker until you understand the workaround: distributable net income (DNI). When a non-grantor trust distributes income to a beneficiary, the tax goes with the distribution — at the beneficiary's individual rate, not the trust's. Most active non-grantor trusts distribute current income to avoid the compression. The compressed rates only matter for income the trust retains and reinvests.

For QSBS, this almost never matters at exit, because §1202's exclusion is of capital gain, not ordinary income. The compression is a real consideration for what the trust does with the proceeds after exit, not the QSBS event itself.

Estate tax: why this matters after the exit

Assets held in a properly structured non-grantor trust sit outside the grantor's taxable estate.

The implication: a founder can move appreciated equity into a non-grantor trust before a markup, lock in the gift-time value for transfer-tax purposes, and let all subsequent appreciation grow outside the estate. With the post-OBBBA lifetime gift exclusion permanently set at $15M per individual ($30M for married couples) starting in 2026, that's enough room to gift a meaningful slice of founder stock without triggering current gift tax. For more on that mechanic, see our lifetime gift tax exemption guide.

Asset protection: the underrated benefit

A properly structured non-grantor trust is creditor-remote. Divorce, business creditors, malpractice claims, post-exit lawsuits — none of them reach trust assets if the document and the jurisdiction are right.

For a founder holding a single concentrated position that may be worth tens of millions at exit, that's not a side benefit. It's often the second-most-important reason to set the structure up at all.

Why non-grantor trusts matter for QSBS

This is the section that drives most founders to this page.

IRC §1202 caps the QSBS gain exclusion at the greater of $15M (for stock issued after July 4, 2025) or 10× the taxpayer's adjusted basis, applied per taxpayer, per issuer. For pre-OBBBA stock, the cap is $10M with a 5-year cliff and no tiered exclusion. Two regimes can sit on the same cap table. Confirm which applies to your stock before doing anything else.

The operative phrase is per taxpayer. A single founder gets one exclusion. Each properly structured non-grantor trust is a separate taxpayer — with its own exclusion.

That's the entire mechanic of QSBS stacking. Multiply taxpayers, multiply exclusions.

Worked example — three taxpayers, $60M exit, post-OBBBA stock

A founder expects a $60M exit on QSBS issued in 2026. Acting 22 months before the exit, the founder gifts portions of the stock into two non-grantor trusts — one for each child. At exit, three taxpayers each claim a separate $15M exclusion:

  • Founder: $15M exclusion claimed.
  • Trust 1 (child A): $15M exclusion claimed.
  • Trust 2 (child B): $15M exclusion claimed.
  • Total excluded: $45M.

The remaining $15M of gain is taxed at ~23.8% federal (20% long-term capital gains plus 3.8% net investment income tax) — about $3.57M. Without the two trusts, $45M of gain would have been taxed at ~23.8% — about $10.71M in additional federal tax. The trusts save roughly $10.71M.

For the deeper mechanics, see our QSBS trust stacking guide.

What disqualifies a "non-grantor trust" for QSBS stacking

Three failure modes account for almost all of the cases where stacking falls apart at exit:

  • Grantor-trust contamination. A retained §675(4) substitution power, a power to revoke, or a retained income interest turns the trust into a grantor trust. Grantor trusts share the grantor's exclusion — they don't add a new one.
  • Self-settled status in the wrong jurisdiction. If the grantor is also a beneficiary, the trust may be treated as a grantor trust or pulled back into the estate, depending on the state's domestic asset protection trust statutes. Nevada and a handful of other states have favorable rules; most do not.
  • §643(f) trust aggregation. Two or more non-grantor trusts with substantially the same grantor and substantially the same primary beneficiaries — plus a tax-avoidance purpose — can be collapsed by the IRS into one trust for federal income tax purposes. Mitigation: distinct primary beneficiaries, staggered formation, varied terms.
"Setting up identical Nevada trusts on the same day for the same beneficiary, each funded with QSBS, is risky under §643(f). Use different beneficiaries or staggered creation times and funding levels."
— Frost Brown Todd, Transfer Planning With Qualified Small Business Stock

The reciprocal trust doctrine is a related concern when spouses set up trusts naming each other as beneficiaries. Courts have repeatedly unwound those structures. Building real differences into the trust terms is essential.

Holding-period tacking under §1223(2)

This is the single most under-explained point in QSBS-and-trust content.

When you gift QSBS to a non-grantor trust, the trust tacks your holding period under IRC §1223(2). The clock doesn't restart. The trust steps into your QSBS clock as if it had owned the stock since the day you did.

Practical implication: gifting at year four is fine. The trust crosses the 5-year line on the same date you would have. Gifting at year four and 11 months is fine. The structure works as long as the gift happens before the sale and the trust is properly non-grantor.

What does not tack: a sale to the trust. Sales reset the holding period. Stacking depends on a true gift transfer, properly documented, with the right valuation discounts and gift-tax reporting.

The 18-month planning window

The question we get most: how late is too late?

The answer Promissory uses is 18 months, and we mean it. The IRS has multiple doctrines available to challenge transfers that happen too close to a known liquidity event — assignment of income, step transaction, and the substance-over-form principle. None of them have bright-line rules. Eighteen months before a known exit gives the structure enough independent operating life to defend the substance.

Inside six months of a signed term sheet is reckless. We've seen it work and we've seen it fail; it's not a defensible plan.

When founders should (and shouldn't) use a non-grantor trust

A non-grantor trust is not a default. It's a vehicle for specific situations.

Good fits

  • Founder with $20M+ of unrealized QSBS gain expecting an exit in 2–5 years
  • Founder wanting to transfer appreciation out of the taxable estate before a markup
  • Founder concerned about asset protection — divorce, professional liability, business creditors
  • Founder in a non-conforming state for QSBS purposes (CA, PA, AL, MS, HI), where state-level planning matters as much as federal — see our state tax guide

Bad fits

  • Sub-$10M expected gain — your single $15M federal exclusion already covers it
  • Pre-Series A or pre-revenue with no near-term liquidity catalyst
  • Founder who wants to retain control of voting and dividends — non-grantor structure requires relinquishment
  • Founder unwilling to make a completed gift (gift completion is a structural requirement, not a preference)

The "almost non-grantor" trap

A surprising number of trusts marketed as "non-grantor" are accidentally grantor trusts because of one retained power, most often the §675(4) power to substitute trust assets for assets of equivalent value. Lawyers sometimes include it by default for flexibility. It is the silent killer of QSBS stacking plans.

A simple test: if you (the founder) can change beneficiaries, swap assets in or out, or revoke any provision, the trust is almost certainly a grantor trust. Have a tax attorney review the agreement against the §671–679 trigger checklist before transferring stock. The cost of a 30-minute review is rounding error against a multi-million-dollar planning failure.

Non-grantor trust vs. IDGT vs. SLANT

Three trust structures show up in almost every founder pre-exit conversation. They sound similar, and they are not interchangeable.

  • Income tax. NGT: the trust pays. IDGT: the grantor pays (intentional defect). SLANT: the trust pays.
  • Separate QSBS taxpayer. NGT: yes. IDGT: no. SLANT: yes.
  • Grantor access. NGT: none. IDGT: none (but pays the trust's tax). SLANT: the spouse can access.
  • Estate tax. All three sit outside the grantor's estate when structured properly.
  • Best for. NGT: QSBS stacking + asset protection. IDGT: estate freeze + basis-burn growth assets. SLANT: QSBS stacking + retained family liquidity.
  • Setup complexity. NGT: medium. IDGT: medium. SLANT: higher (spousal mechanics).

When a non-grantor trust beats an IDGT

For QSBS stacking, only. An IDGT is intentionally a grantor trust — that's the whole point. It shares the grantor's QSBS exclusion. It doesn't add a new one. If your goal is stacking, a non-grantor trust is the structure.

When a SLANT beats a non-grantor trust

When the founder wants the spouse to retain access to the gifted assets. A SLANT preserves spousal liquidity while keeping non-grantor (and therefore separate-taxpayer) status. For founders concerned about gifting away too much liquidity before exit, the SLANT structure is often the answer.

When an IDGT beats a non-grantor trust

When the goal is estate freezing of growth assets and the grantor wants to pay the trust's income tax. The grantor paying the tax is itself a tax-free gift to the trust each year — it's a feature, not a bug. Different objective, different toolkit.

Setting up a non-grantor trust: what founders actually need

A non-grantor trust is not a self-serve product. The components:

  • Trust agreement drafted by a tax attorney, with the §671–679 trigger checklist run before signing
  • Jurisdiction — Nevada is the common choice for QSBS-focused trusts because of directed-trustee statutes, no state income tax on trusts, robust asset protection, and friendly decanting rules
  • Independent trustee — required for non-grantor status if the grantor wants to retain any administrative role; institutional trustees are common for QSBS planning
  • EIN, separate accounting, separate tax filings — the trust files Form 1041 each year and tracks its own basis, holding period, and §1202 election at exit

Setup is a fraction of the cost of full-service Big Law trust work and meaningfully more involved than a DIY revocable living trust. Promissory builds non-grantor trusts purpose-built for QSBS stacking on a fixed-fee model. See our pricing page for the current numbers.

The most common non-grantor trust mistakes

Five failure modes account for almost every story we've seen end badly:

  • Power-to-substitute contamination. The §675(4) trap. Retained for flexibility, kills the non-grantor status, kills the stacking plan. Lead with this when reviewing a draft.
  • Self-settling in a non-DAPT jurisdiction. Naming yourself a beneficiary in a state without a domestic asset protection trust statute can re-cause grantor-trust treatment or pull the assets back into the estate.
  • Last-minute transfers. Inside six months of a known liquidity event, the substance-over-form risk is high enough that the structure becomes hard to defend.
  • §643(f) collapse. Multiple trusts with the same grantor and beneficiaries plus a tax-avoidance purpose collapse into one trust. Distinct beneficiaries and staggered formation matter.
  • Forgetting state-level rules. California's source-income rules can pull non-grantor trust income back to a California founder regardless of trust situs. Multistate trust planning is its own discipline.

The bottom line

Three things to take away from this page:

  1. Non-grantor means separate taxpayer — and separate taxpayer means separate QSBS exclusion, separation from your taxable estate, and asset protection.
  2. The structure is only as non-grantor as the trust agreement. One retained power is enough to break the treatment. Have a tax attorney review against §§671–679 before you sign.
  3. The founders who win with this strategy are the ones who set up the structure 18+ months before an exit. Inside six months is reckless. Now is almost always the right answer.

Promissory builds non-grantor trusts purpose-built for QSBS stacking, with Nevada situs and institutional trustees, on a fixed-fee engagement. If you have founder stock and a horizon to exit, schedule a free consultation to scope what stacking could look like for your situation.

‍