Sudden $2M: The 6 Tax Traps in the First 6 Months

Marcus Reyes, 58, single, a software engineer in San Jose, California, sold a chunk of private-company stock and netted $2,000,000 in long-term capital gains in February 2026. He assumed the tax would simply “come out at filing.” By the time we ran the numbers, the real exposure was $476,000 in federal tax (LTCG 20% + NIIT 3.8% on the gain over his $533,400 single LTCG threshold), roughly $266,000 in California tax (13.3% top rate, no preferential gains rate), a looming $8,576 Medicare premium spike for 2028, and — the part that genuinely cost him — an underpayment penalty because he made no Q1 estimated payment. Six traps, all triggered by one transaction, most of them invisible until the deadline had passed.

The mistake Marcus almost made is the universal sudden-wealth mistake: treating a windfall as a single April event instead of a six-trap obstacle course that starts the day the money lands. Here is each trap in the order it hits, the dollar math, and the escape hatch.

First, the threshold question: is your $2M even taxable?

Not every windfall is income. The single biggest variable is the source, because it determines whether you owe $0 or $700K+ before you do anything else:

Marcus’s $2M was a capital gain, so his federal exposure is the 23.8% line, not the 37% line. If yours is an inheritance, traps #1, #3, #4, and #5 below may not fire at all — but trap #2 (IRMAA) and #6 (AMT, if you exercise options) still can. Identify your source first.

Trap #1: The Q1 estimated-tax surprise (the most expensive one)

The US tax system is pay-as-you-go. Withholding from a paycheck covers wage earners, but a windfall almost never comes with adequate withholding — a brokerage does not withhold on a stock sale. If you do not send the IRS quarterly estimated payments (Form 1040-ES), you owe an underpayment penalty, which is really interest charged from each missed quarterly deadline.

For a windfall received in February (Q1), the first estimate is due April 15, 2026. Marcus’s federal liability on the gain is about $476,000. If he pays nothing until April 2027, the penalty interest accrues on that entire balance for a full year.

The escape hatch: the 110% safe harbor

You owe no underpayment penalty — no matter how enormous this year’s bill grows — if you pay, across four equal quarterly installments, the smaller of:

• 90% of this year’s actual tax, or
• 110% of last year’s total tax (the 110% figure applies because your prior-year AGI was over $150,000; it’s 100% if AGI was at or under $150,000).

This is the windfall planner’s favorite rule. Say Marcus’s 2025 total federal tax was $90,000. He pays 110% × $90,000 = $99,000 in four $24,750 installments across 2026. He will still owe roughly $377,000 more at filing on April 15, 2027 — but he owes zero penalty on that gap because he hit the safe harbor. He keeps the $377K invested for up to 14 extra months. The trap is not the tax; it’s the penalty on top of the tax, and the safe harbor erases it.

Trap #2: The IRMAA two-year lookback (the delayed-fuse trap)

If you are 63 or older, a windfall does something sneaky: it raises your Medicare premiums — but not until two years later. The Income-Related Monthly Adjustment Amount (IRMAA) sets your Part B and Part D premiums using your Modified Adjusted Gross Income from two years prior. A 2026 windfall drives your 2028 premiums.

A $2M taxable windfall lands you in the top IRMAA tier. Here is the 2026 surcharge schedule (based on 2024 MAGI; the structure is the same for the 2028 premiums your 2026 income will set):

At the top tier, Marcus pays ($628.90 + $85.80) × 12 = $8,576 for the year, versus the $2,220 a base-premium retiree pays — an extra $6,356 for one year, per person. The good news: IRMAA is a one-year hit. It resets once your MAGI drops back. The bad news: you generally cannot appeal it just because the income was one-time. Form SSA-44 only covers specific “life-changing events” (work stoppage, marriage, death of a spouse) — not “I sold stock.”

Trap #3: Crossing the NIIT threshold

The Net Investment Income Tax (IRC §1411) adds 3.8% on the lesser of your net investment income or your MAGI over $200,000 (single) / $250,000 (MFJ). A windfall of capital gains, dividends, or interest is net investment income by definition, and $2M obliterates the threshold.

This is why the headline “20% capital gains rate” is wrong for a windfall. The real top federal rate on Marcus’s gain is 23.8% (20% LTCG + 3.8% NIIT). On the portion of his $2M gain above the threshold, that extra 3.8% is roughly $68,000 of tax that does not appear on any “capital gains bracket” chart. If you model a windfall using only the LTCG table, you will understate the bill by tens of thousands.

Trap #4: The Social Security 85% taxation jump

If you are already collecting Social Security, a windfall can make up to 85% of your benefit taxable. The thresholds are brutal because they were set in 1983 and have never been inflation-indexed:

• Single: up to 50% of your benefit becomes taxable once combined income tops $25,000; up to 85% once it tops $34,000.
• MFJ: 50% above $32,000 combined income; 85% above $44,000.

“Combined income” = AGI + nontaxable interest + half your Social Security benefit. A $2M windfall — or even just the interest and dividends it throws off afterward — sails past $34,000 instantly. So the windfall is taxed, and it drags 85% of your previously-low-taxed Social Security benefit into your top marginal bracket. For a retiree drawing $40,000/yr in benefits, that means roughly $34,000 of benefit now taxed at 32%–37% instead of being mostly tax-free — an extra $11,000+ of tax that has nothing to do with the windfall’s own bill.

Trap #5: State residency and state-tax exposure

Most states tax capital gains as ordinary income — there is no federal-style preferential rate at the state level. So a windfall realized while you are a California resident is taxed at up to 13.3%; in New York, up to 10.9% (plus up to 3.876% NYC resident tax). On Marcus’s $2M, California alone takes roughly $266,000.

*Washington imposes a 7% tax on long-term gains over $250,000; the other three have no state income tax. The trap is timing: if you are about to relocate, when and where you become a resident relative to the realization event can be a six-figure decision. But do not improvise a same-week move to dodge tax — states aggressively audit “convenience” moves around large realization events, and a sham change of domicile fails. The legitimate lever is realizing income after a genuine, well-documented change of residency.

Trap #6: AMT on any ISO exercise in the windfall year

If part of your sudden wealth involves incentive stock options (ISOs), exercising them — even without selling — creates an Alternative Minimum Tax preference item equal to the bargain element (fair market value minus strike price). There is no regular-tax income at exercise, but for AMT purposes the spread counts, and AMT runs at 26% or 28%.

Stacking an ISO exercise on top of an already-massive income year is the classic AMT trap: the windfall consumes your AMT exemption phase-out, and the ISO spread gets taxed at the AMT rate this year, potentially with no offsetting cash because you held the shares. If you must exercise, the cleaner play is often to sell the same year (a disqualifying disposition) so the spread is ordinary income with cash to pay it — or to spread exercises across multiple years to stay under the AMT crossover. Model the AMT before you click “exercise.”

The full picture: Marcus’s six-trap tally

What most people miss: the recognition-timing lever defuses several traps at once

The reflex after a windfall is to think about where to invest the money. That is the wrong first question. The trap-avoidance question is when you recognize the taxable event — because a single timing decision can move you across the NIIT threshold, the IRMAA tiers, the Social Security thresholds, and the AMT crossover simultaneously.

Three timing moves people overlook:

1. Split a sale across two tax years. Selling $1M in December 2026 and $1M in January 2027 instead of $2M at once spreads the income across two IRMAA lookback years and two NIIT calculations — sometimes shaving a full IRMAA tier off each year. The total federal gains tax may be similar, but the IRMAA and threshold-cliff costs drop.
2. Harvest losses against the gain in the same year. If you hold any positions underwater, realizing those losses in the windfall year offsets the gain dollar-for-dollar, directly shrinking the NIIT base and the AGI that drives IRMAA two years out.
3. Fund a donor-advised fund with appreciated stock before selling. Donating appreciated shares (rather than cash) deducts FMV and avoids the gain entirely on the donated portion — reducing the taxable windfall, the NIIT base, and the IRMAA-driving MAGI in one move, with the deduction landing in the high-income year when it’s worth the most.

None of these change the asset allocation you eventually choose. They change the size and timing of the tax footprint — which is the part of a windfall you actually control.

The decision lever

The first move after a $2M windfall is not “invest it” and it is not “wait until April.” It is to send a Q1 estimated payment that hits the 110% safe harbor — that single action removes the most expensive trap (the penalty) and buys you up to 14 months of penalty-free time to model traps #2 through #6 deliberately. Run the recognition-timing math before you realize the next dollar of income: split the sale across tax years, harvest offsetting losses, and consider funding a donor-advised fund with appreciated shares in the same high-income year. Control the timing, and you control how many of the six traps ever fire.