Capital Gains Tax on Home Down Payments Explained

The number you pay in capital gains tax when you liquidate investments for a home down payment is not 15%. It is not 20%. For most HNW buyers liquidating from a high-cost-basis portfolio in 2026, it sits between 23.8% (Texas, Florida) and 37.1% (California) of the realized gain, with New York around 33.7% and most other high-tax states clustered between those two anchors.

This post walks the layers — federal long-term, NIIT, and state — and shows the combined number for the residency states that actually matter to HNW buyers. Where it gets interesting is the non-obvious parts: the NIIT threshold that hasn't moved since 2013, the cost-basis variable that swings the bill more than residency does, and the bracket-stacking that happens when an RSU vest and a portfolio liquidation hit the same tax year.

The four layers, in order

Federal long-term, NIIT, state, and (sometimes) local. They compose multiplicatively in some cases and additively in others, which is why the number that comes out of the wash is rarely the one buyers expect.

Layer 1: Federal long-term capital gains. 0%, 15%, or 20% based on taxable income. For 2026 (Rev. Proc. 2025-32), the 20% bracket starts at $545,500 single / $613,700 MFJ. For nearly every HNW buyer with W-2 income above $400K, the marginal long-term rate on a meaningful liquidation is 20%.

Layer 2: Net Investment Income Tax (NIIT). A flat 3.8% on top of the federal rate. Triggered when MAGI exceeds $200K single / $250K MFJ. Crucially, those thresholds are statutory — they have not been indexed since the tax took effect in 2013 — meaning more and more buyers cross them every year through inflation alone. Combined federal-plus-NIIT for the typical HNW filer: 23.8%.

Layer 3: State. Varies wildly. The states matter for HNW buyers because the dollars are large enough that the gap between low-tax and high-tax residency dwarfs the cost of moving, hypothetically, except moving has its own complications.

State	Top LTCG rate (treated as)	Combined with 23.8% federal+NIIT
California	13.3% (ordinary income)	37.1%
New York	10.9%	33.7% (state residents above $25M income)
New York	6.85%	30.65% (most NY filers)
New Jersey	10.75%	33.55%
Massachusetts	9% on > $1M gains	30.0%
Illinois	4.95%	27.85%
Texas / Florida / Tennessee	0%	23.8%
Washington	7% on > $250K LTCG	29.6% (only on the marginal portion above $250K)

Layer 4: Local. New York City adds 3.876% on top of state for residents earning above ~$50K MAGI, putting NYC's effective combined LTCG rate just under 38% for a 2026 buyer in the city — not far from California despite the much-publicized state-rate gap.

A buyer in NYC liquidating $300K of stock with a 50% basis ($150K gain) writes a tax check of roughly $57,000. A buyer in Austin doing the same thing writes a check for roughly $35,700. Same transaction, $21,300 difference, only because of geography.

The bigger lever: cost basis

Residency moves the rate. Cost basis moves the size of the gain that the rate is applied to. The latter swings the final tax bill more than the former, and most buyers don't have an accurate handle on it.

A $2M portfolio built over twenty years through patient buying and dividend reinvestment might have a cost basis of 70–80% of market value (only 20–30% unrealized gain). A $2M portfolio built mostly from RSU vests at grant-date prices that have since appreciated 4× might have a cost basis of 25%, meaning 75% of every dollar liquidated is realized gain. Same portfolio value. Wildly different tax bills.

A $300K liquidation, with combined rate of 30.65% (NY filer outside NYC):

• 80% basis: $60K gain × 30.65% = $18,400 in tax
• 50% basis: $150K gain × 30.65% = $45,975 in tax
• 25% basis: $225K gain × 30.65% = $68,963 in tax
• 10% basis (low-basis founder stock): $270K gain × 30.65% = $82,755 in tax

That's a 4.5× swing in tax bill, on the same dollar amount of liquidation, driven entirely by how the portfolio was built. The NYC-vs-Austin geographic swing was about $21K; the cost-basis swing on the same trade is over $64K.

If you don't know your cost basis cold, you don't know what your liquidation actually costs. Most brokerages report it on year-end statements. It's worth pulling.

What I didn't expect about my own bracket math

I'd been thinking about this as a static rate problem when I started researching for my own purchase: figure out the combined rate for my state, multiply by gain, get a number, decide. The version that actually showed up in my spreadsheet was much messier and much more expensive.

The thing I underestimated is bracket stacking in years when equity comp also lands. For a tech worker compensated in RSUs, the vest hits W-2 ordinary income. A meaningful vest in a calendar year — say, $200K of RSUs — pushes total taxable income up, which (a) puts more of any subsequent capital gain into the 20% federal LTCG bracket instead of the 15% bracket, (b) raises the amount of investment income exposed to NIIT, and (c) at the federal level interacts with the AMT exposure on any same-year ISO exercise. Liquidating $300K of long-held stock in that year is materially more expensive than liquidating the same $300K in a year with no vest.

The implication for my own model: don't think about "what's my LTCG rate" — think about "what's my combined effective rate in the specific tax year I'd liquidate." For me that meant the year I was eyeing for a purchase had me already at the highest combined rate I'd ever face, because of vest timing. Liquidating to fund the down payment in that year would have stacked an extra ~$20K of tax on top of the static-rate calculation. Which is, not coincidentally, roughly what the entire cost difference between SBLOC and liquidation comes out to in a single year of carry.

That's the moment the SBLOC question stopped being "is this cheaper?" and started being "is this avoidable?" The answer to "cheaper" is sometimes yes, sometimes no, depending on rates. The answer to "avoidable" is almost always yes — the gain doesn't have to be realized in 2026; it can be deferred, harvested in a deliberate low-income year, or never realized at all if the holder dies and the basis steps up.

That re-framing is what I think most HNW buyers miss when they read tax-rate posts in isolation. The question isn't "what rate do I pay." The question is "what rate will I pay in the specific year I'd otherwise force the realization."

Three planning levers most calculators ignore

If you're going to liquidate, the framework above is the static answer. But a buyer with optionality has at least three levers that change the bill.

Year selection. If you can spread a $1M liquidation across three tax years instead of one, you may keep more of it under the 20% federal LTCG bracket and below the highest NIIT exposure. This works most cleanly when the home purchase doesn't require all the cash at closing — for example, when an SBLOC fronts the down payment and the borrower paid back over a planned multi-year liquidation.

Lot selection. When you sell, you can typically specify which tax lots to sell — high-basis lots first to minimize realized gain, low-basis lots first if you're using a tax-loss-harvesting strategy that offsets them, or specific identification to optimize for the year. Default settings on most brokerages are FIFO, which is rarely tax-optimal.

Charitable offset. If you're already planning meaningful charitable giving, donating appreciated securities instead of cash means the charity gets the gross value, you avoid the gain on the donated lot, and you get a fair-market-value deduction. For HNW buyers in high-itemizing positions, this can shelter $50K–$200K of intended gain harvest. Not relevant for the down-payment dollars themselves, but relevant for the broader tax year.

When capital gains tax stops being the deciding factor

For everything I've said about how big this number is, there are profiles where it's small enough to ignore.

• High-cost-basis portfolios. If your basis is 80%+, the gain is small and the tax bill is in the noise. Liquidate.
• Recent contributions. If most of your portfolio is the last three years of paychecks, the embedded gain is modest. Liquidate.
• No-income-tax residents at the moderate end of the federal bracket. Florida filer in the 15% federal bracket with 3.8% NIIT and a 50% basis on a $300K liquidation pays about $28K — material but not huge.
• Loss harvesting available. If you have realized or unrealized losses to offset, the effective rate on the gain can be near-zero for the year.

In any of those scenarios, the tax-avoidance argument for an SBLOC weakens, and a clean liquidation becomes the simpler path. The framework is the same; only the inputs change.

Run your own combined rate

Stockstead's calculator computes the combined effective LTCG rate for your state, applies it to your cost basis, and runs the comparison against the SBLOC and traditional-mortgage paths. The static-rate answer is in the model. The bracket-stacking awareness — which year you're liquidating in and what other income hits that year — is on you, and it's worth modeling explicitly with your tax advisor before you sign.

Open the calculator →

The headline number is "23.8% to 37.1%, depending on state." The realistic number for any specific buyer in any specific year can be higher, sometimes meaningfully. Worth knowing before you write the check.

---

Sources

• IRS Rev. Proc. 2025-32 — 2026 inflation adjustments
• IRS Topic 409 — Capital Gains and Losses
• IRS — Net Investment Income Tax (NIIT)
• IRS Publication 550 — Investment Income and Expenses

Tax rates current as of 2026. Long-term capital gains brackets reflect IRS Rev. Proc. 2025-32 inflation adjustments. NIIT thresholds are statutory and unchanged since 2013. State rates (California, New York, Texas, etc.) are current as of January 2026; verify with your state's revenue department before relying on these figures. Educational, not financial or tax advice. Tyler Singletary founded Stockstead while researching his own first home purchase. He's spent 20+ years in software product roles — currently as Product Specialist for AI/ML Startups at AWS, previously CPO/COO at Tagboard, and an executive at Klout (exited to Lithium in 2014) — and has been compensated in RSUs, ISOs, and post-IPO equity at high-growth tech companies for two decades. He is not a licensed financial advisor or CPA. Tax rules are complex and your specific situation matters; consult a licensed professional before making any liquidation decisions.