Should You Fund a GRAT, Gift Shares, or Hold for Step-Up on a $20M Estate If the AI Bubble Pops? A 5-Question Checklist
Say you're single, your estate is $20 million, and $8 million of it is one stock position that has ridden the AI wave. Your basis is $2 million. The market is at levels that make you nervous, and you've read enough headlines to know you can't tell whether it keeps running or cracks.
Should you fund a GRAT, gift the shares, or hold them and let the step-up in basis do the work at death?
The answer depends on which way the stock goes, and nobody knows that. So this post skips the prediction and runs all three strategies through a bear, base and bull case. Then it gives you a five-question checklist to run on your own numbers.
Everything below is a constructed example, not a client case or a forecast. Your numbers will differ based on your specific situation.
Why This Question Is Live Right Now
Three data points frame the decision.
The market debate. Mr. Money Mustache's piece, "Will the AI Bubble Destroy our Retirement?", starts from a familiar point. Markets keep surprising us, and we worry when they crash. A record high is no comfort either, because it can leave you holding an even bigger, more concentrated position. For estate planning, that concentration is what matters. A stock that has quadrupled is exactly the asset a GRAT was designed for. It is also the asset most likely to give back gains before you die.
The inflation and jobs picture. The Bureau of Labor Statistics' latest indicators show CPI up 0.4% in August 2026, an unemployment rate of 4.1%, payroll employment up 162,000 (preliminary), and average hourly earnings up $0.10 (preliminary). That is a solid labor market with inflation not fully tamed. A 0.4% monthly print, if it persisted, compounds to about 4.9% annualized (1.004¹² ≈ 1.049). I'm not forecasting that. But sticky inflation matters here for two reasons:
- Higher rate expectations tend to push up the IRS 7520 rate. That is the hurdle a GRAT must beat, and it works against GRATs.
- The federal exemption is indexed to inflation. That helps a little, but it doesn't change the size of the gap between your estate and the exemption.
For how a rising hurdle rate hits different structures, see how a rising IRS 7520 rate costs a $5 million GRAT vs. an IDGT.
The effort-versus-payoff lens. NerdWallet's "Should I Switch to a New Bank Just to Earn a Bonus?" makes a point that carries over here. Bank bonuses take effort, so you weigh the payoff against the work and the strings attached. Estate planning has the same shape at a much larger scale. In the example below, the difference between the best and worst strategy is over $1.8 million in one scenario. That is a decision worth an afternoon of real math, not a rule of thumb.
The Setup (Example Assumptions)
- Single filer, $20M estate: an $8M concentrated stock position (basis $2M) plus $12M of other assets, held flat for simplicity.
- Federal exemption: $15M. Estate tax rate: 40% on the excess.
- Horizon: 10 years, then death.
- Beneficiaries sell inherited or gifted shares at a 23.8% federal long-term capital gains rate (20% plus 3.8% NIIT). We'll also note what changes if they hold instead.
- GRAT: rolling short-term GRATs, with an example hurdle rate of 4.8%.
- Federal only. No state estate tax, no fees, no discounting, and no mortality modeling. Those all matter in real life.
Three growth cases for the $8M position over 10 years:
| Case | Annual return | Value at death |
|---|---|---|
| Bear (bubble pops, stays down) | −30% total | $5.6M |
| Base | 6%/yr | $14.3M |
| Bull | 12%/yr | $24.9M |
Strategy 1: Hold Everything and Take the Step-Up
At death your estate is $12M plus the position's value. Tax is 40% of anything over $15M. Basis resets to date-of-death value, so there is no capital gains bill for your heirs.
| Case | Estate at death | Federal estate tax |
|---|---|---|
| Bear | $17.6M | $1.04M |
| Base | $26.3M | $4.53M |
| Bull | $36.9M | $8.74M |
The step-up is real money here. On the base case, your heirs avoid roughly $2.9M of capital gains tax (($14.3M − $2M) × 23.8%). But holding does nothing about the estate tax. In the bull case, the very outcome you'd celebrate produces an $8.7M bill.
Strategy 2: Gift the $8M Position Now
A gift uses $8M of your $15M exemption. Because lifetime taxable gifts get added back at death, your total is still $20M against $15M. That means $2.0M of estate tax in every scenario. The growth escapes, but the exemption you used is gone for good.
The cost is carryover basis. Your heirs get your $2M basis, not a step-up. If they sell:
| Case | Estate tax | Capital gains if sold | Total |
|---|---|---|---|
| Bear | $2.00M | $0.86M | $2.86M |
| Base | $2.00M | $2.93M | $4.93M |
| Bull | $2.00M | $5.44M | $7.44M |
If they never sell, the capital gains column goes to zero. But then the step-up at their death only defers the issue, so the timing of each generation's death matters too.
Note the bear case. Gifting cost you $1.82M more than holding ($2.86M vs. $1.04M). You spent exemption on shares that then fell in value, and you gave up a step-up on a smaller gain. This is the risk that never shows up in "gift early, gift often" advice.
The gift-appreciated-stock vs. step-up break-even analysis shows the same trade-off at a smaller scale.
Strategy 3: A Rolling GRAT on the $8M Position
A GRAT pays you an annuity back and passes only the growth above the hurdle rate to the next generation. In a bear case, nothing passes, and you're essentially where you started. That is the "heads I win, tails I break even" feature.
Rough math for the excess growth over 10 years at a 4.8% hurdle:
- 4.8% compounded over 10 years ≈ 1.598×.
- Base (6%): $8M × (1.791 − 1.598) ≈ $1.54M passes out of your estate. Estate tax saved is 40% × $1.54M ≈ $0.62M.
- Bull (12%): $8M × (3.106 − 1.598) ≈ $12.06M passes out. Estate tax saved is ≈ $4.82M.
- Bear: $0 passes and $0 is saved.
The GRAT remainder also carries your low basis, so I've added a capital gains hit of up to 23.8% on the transferred growth if it's sold (about $0.37M base, $2.87M bull).
| Case | Estate tax | Capital gains on transferred growth if sold | Total |
|---|---|---|---|
| Bear | $1.04M | $0 | $1.04M |
| Base | $3.91M | $0.37M | $4.28M |
| Bull | $3.92M | $2.87M | $6.79M |
Side-by-Side: Total Federal Tax Cost (Estate Tax + Capital Gains if Sold)
| Case | Hold for step-up | Gift $8M now | Rolling GRAT |
|---|---|---|---|
| Bear (−30%) | $1.04M | $2.86M | $1.04M |
| Base (6%/yr) | $4.53M | $4.93M | $4.28M |
| Bull (12%/yr) | $8.74M | $7.44M | $6.79M |
A few honest takeaways:
- The GRAT never lost to holding in this example. That is the structural advantage of a strategy that only transfers excess growth. Real GRATs carry legal fees and mortality risk (die during the term and the assets come back into your estate), so "never lost" has a small asterisk.
- The gift only won in the bull case, and only against holding. It lost badly in the bear case. If you believe in the bubble risk, gifting the volatile asset is the riskiest way to use your exemption.
- The base case is nearly a wash across all three. Only $0.65M separates the best from the worst. If you expect middling returns, the choice matters much less than it feels like it does.
- Holding wins the bear case outright and loses the bull case badly. It is effectively a bet that your stock doesn't run.
This is the kind of scenario comparison Voritanel runs for you, so you don't have to build the spreadsheet yourself.
What Could Change This Math
A higher hurdle rate. At 4.8% the GRAT excess in the base case is $1.54M. If the 7520 rate moves up, it shrinks fast. At an example 5.5%, base-case excess growth drops to about $0.9M, and the 6% return case is barely worth the paperwork. The bull case stays strong. See how a rate jump shifted GRAT vs. IDGT results on a $10M estate.
An IDGT instead of a GRAT. An IDGT can be more efficient when growth is high and you want to keep transferring, but it uses exemption up front. The trade-offs are covered in when a GRAT beats an IDGT.
State estate tax. If you live in a state with its own estate tax and a lower threshold, your total tax could be well above the federal figure shown here. Federal portability doesn't fix that, and most states have no portability at all.
Being married. A married couple with $20M has a very different picture. With two exemptions and a portability election, this same $20M might owe nothing federal today. The planning question then becomes what happens after the first death. See the portability election break-even at 4.8%.
Charitable intent. If you'd give some of this position to charity anyway, a charitable remainder trust can defer the gain on a sale while giving you an income stream. That is a different tool for a different goal, and it's worth comparing.
The 5-Question Checklist
You don't need to guess where the market goes. Answer these five in order.
1. Is my total estate above my available exemption, today and at a realistic growth rate? If you're at $20M against a $15M exemption, you have a $5M gap that grows with the market. If you're at $9M single, the strategies above don't matter yet. Your estate is under the exemption, and holding for step-up is likely your best move. Don't add complexity you don't need.
2. How concentrated am I, and what is the basis? An $8M position with a $2M basis has $6M of embedded gain. Every strategy interacts with that gain differently. A diversified portfolio with a high basis makes gifting much cheaper than it looks in this example.
3. Can I live with the bear case of each option? Gifting the volatile asset had a bear-case cost of $1.82M more than holding. If a 30% decline would leave you wishing you'd done nothing, a GRAT's downside protection may fit better. If you'd rather not think about it at all, holding may be the calm choice, and that is a valid answer.
4. What is the hurdle rate today, and how much do I lose if it rises before I fund? A GRAT funded at a higher 7520 rate has less room to work. It is worth checking the current rate the week you act, not the week you read about it. If you want a timing framework, Should You Fund a GRAT This Week? walks through it.
5. What will my heirs actually do with the shares? Sell right away, and carryover basis costs real money. Hold for decades, and the gap between gift and step-up narrows. If they will sell, the step-up you'd give up by gifting can cost more than the estate tax you'd save.
What This Means for Your Numbers
The example above has one stock, one horizon, one state of the world, and no fees. Your real version has your actual basis, your state, your marital status, your other assets, and a 7520 rate that changes monthly. A small change to any of these can flip the ranking.
Notice too how little the market forecast mattered to the method. You don't need to know whether the AI boom fades or keeps running. You need to know which strategy holds up across the outcomes you can imagine, and what the loser's regret costs you. In this example that is $1.8M on the downside for gifting and $1.9M on the upside for holding.
If you'd like to see your own three-scenario table, you can model this for your specific situation at Voritanel. Enter your estate size, basis, state and expected growth range, and compare a GRAT, an IDGT, a gift and a hold side by side.
Nobody needs to rush. But the numbers are much easier to read before the market makes the decision for you.
This post is educational and uses a hypothetical example. It is not legal, tax or investment advice. Estate tax law changes, and a qualified estate attorney should review any plan before you act.
Sources
- Will the AI Bubble Destroy our Retirement? — Mr. Money Mustache
- Should I Switch to a New Bank Just to Earn a Bonus? — NerdWallet
- This Tahoe Hotel Got a Glow-Up, but Missed a Few Spots — NerdWallet
- National Coffee Day: Where to Find Free Coffee and Deals on Sept. 29 — NerdWallet
- Major Economic Indicators Latest Numbers — Bureau of Labor Statistics