GRATs at scale: when the math actually works for founders.

Grantor Retained Annuity Trusts are the most discussed and least well-implemented estate technique we see in tax advisory practice. Almost every founder we work with has heard the term. Almost none of them have a GRAT positioned to actually do the work it is capable of. The gap between "we set up a GRAT" and "the GRAT moved $40M of pre-liquidity appreciation outside the estate" is enormous, and it is almost entirely a function of how the planning is modeled, by the founder's tax advisors and estate counsel together, before the trust is funded.

The mechanics in 60 seconds

A GRAT is a grantor trust funded with an asset the grantor expects to appreciate. The grantor receives back a fixed annuity stream over a specified term, calculated to return the original principal plus a statutory hurdle rate (the §7520 rate, published monthly by the IRS). Anything the asset produces above that hurdle remains in the trust at term, outside the estate, transferred without using lifetime exemption, and often payable to a continuation trust for descendants.

When the planning is sound, the gift-tax cost of funding the GRAT can be modeled to near zero. The downside, when the asset underperforms the hurdle, is also near zero. The assets simply return to the grantor and no exemption has been consumed. This asymmetry is the entire point of the technique.

The two questions every founder asks

1. "Why not just gift the shares?"

Because outright gifts permanently consume lifetime gift-and-estate exemption. Under the One Big Beautiful Bill Act, the federal estate and gift tax exemption is $15 million per individual beginning in 2026 ($30 million for married couples using portability), indexed for inflation starting in 2027, with no scheduled sunset. That is meaningfully higher and more stable than the prior framework, but it is still a finite amount, and 40% federal estate tax still applies above it. A GRAT can move appreciation off the estate balance sheet without using meaningful exemption, which preserves that exemption for the founder's other purposes. Said differently, the exemption is no longer a "use it or lose it" resource on a clock, but it is still a scarce resource worth conserving for assets where it does the most work.

2. "What if the asset goes down?"

The grantor receives the assets back through the annuity stream, no gift tax has been triggered, and no exemption has been consumed. The founder is then free to work with counsel to fund a new GRAT at a lower asset value and possibly a lower hurdle rate. This is the reason rolling short-term GRATs typically outperform a single long-dated GRAT on volatile assets, a point we will return to.

Why short-term rolling GRATs tend to beat long-dated ones

A 10-year GRAT funded with a single block of pre-IPO shares is a single coin flip. If the company underperforms the hurdle in years 1 through 4 and recovers in years 7 through 10, the GRAT can still largely fail, because the annuity payments early on strip the upside back out. A series of rolling 2-year GRATs, by contrast, captures each up-cycle independently and absorbs the down-cycles without permanent damage.

• Down year: assets return to the grantor; the founder and counsel can consider funding a new GRAT at the new lower value.
• Flat year: hurdle is met, modest excess transfers, trust continues.
• Up year: hurdle is materially beaten, larger excess transfers, founder considers funding the next layer.

Across a full liquidity cycle, the rolling approach tends to compound the wins and absorb the losses. A single long GRAT averages them together, which is generally the wrong behavior on a high-volatility asset. The right cadence for any particular founder is a planning decision made with the estate attorney.

The hurdle rate, and why timing matters more than size

The §7520 rate determines the annuity stream the GRAT must return. Funding a GRAT in a low-rate environment means a lower bar for the asset to clear. Funding the same GRAT a year later, after a meaningful move in the rate, can change the projected transferred wealth at term substantially on identical asset performance. The rate is published monthly. We monitor it closely as part of the modeling work we do for clients.

Pre-IPO vs post-IPO funding

Pre-IPO is the obvious window: illiquid shares, low 409A valuation, expected re-rating to the public market, with the GRAT positioned to capture appreciation above the hurdle. Valuation considerations on the funded shares (lack of marketability, minority interest), determined by qualified independent appraisers, can further compress the gift-tax footprint, subject to counsel's review.

Post-IPO is harder but not closed. After lockup, a founder with a view on continued appreciation can work with counsel to fund GRATs with public shares and capture upside above the hurdle. Marketability discounts are no longer available, but liquidity makes the annuity payments mechanical and reduces administrative friction. Post-IPO GRATs coordinated with a 10b5-1 plan are a common scenario, with the trading plan handled by securities counsel and the trust work by estate counsel.

What commonly undermines a GRAT

• Weak valuation work. A GRAT is only as defensible as the appraisal underneath it. Valuations done by a generalist appraiser unfamiliar with the asset class create audit exposure that compounds for the life of the trust. We help clients identify qualified appraisers; the engagement and reliance is the client's, with counsel's input.
• Grantor death during the term. If the grantor dies inside the GRAT term, assets are generally pulled back into the estate. Term length is therefore a mortality consideration as much as a return consideration, one we model, and one the attorney weighs in setting the trust terms. Shorter terms are often discussed for older grantors; longer or rolling terms for younger founders.
• Liquidity for the annuity payments. The trust must have either cash or distributable assets to make the annuity stream. For illiquid pre-IPO funding, in-kind distributions back to the grantor are common, and the mechanism must be drafted into the trust by counsel from the start.
• Lead time. For founders approaching a capital event, the productive window for GRAT modeling is roughly 18 to 24 months ahead of the event. The unproductive window is the day the term sheet arrives. By then, valuation positioning is largely fixed, §7520 timing is locked in, and the set of available moves is much narrower.

The bottom line

Used correctly, GRATs do not save current-year tax. They are designed to move the future appreciation of an asset outside the estate, with limited downside if the bet does not pay off. There are very few techniques with that asymmetry. Founders who pursue them seriously, early, and in series, with engaged tax advisors and estate counsel working in tandem, can transfer eight and nine figure sums of future appreciation to the next generation while preserving the bulk of their lifetime exemption for other uses.