By Antonio Hill
A family of three with $106,600 of income in 2026 gets $8,872 from the federal government toward health insurance. The same family at $106,601 gets zero. One dollar over the line costs almost nine grand.
Sounds terrifying. That line is the ACA subsidy cliff, and for early retirees it’s actually good news, because unlike almost everyone else in America, you get to pick which side of it you land on. This article shows you how. And it’s simpler than you think.
Your Spending and Your Income Are Two Different Numbers
Health insurance was the biggest blind spot in my entire early retirement plan. I’d put thousands of hours into accumulation, withdrawal math, and tax strategy, and the whole time I was taking my employer’s coverage completely for granted.
So when I finally dug in, I expected the worst. Everyone tells you the same thing. Retire early and you’ll pay a fortune for health insurance.
The uncomfortable part is that that belief is common because it’s true. Retire at 40 with no plan and you will get crushed.
But the fix turned out to be one number. One number, checked a couple times a year.
That number is your MAGI.
Plain English, no acronym anxiety.
MAGI means modified adjusted gross income. For most early retirees it’s basically the income that shows up on your tax return, which is a completely different thing from what you spend. The marketplace uses MAGI, and only MAGI, to set your subsidy.
The ACA doesn’t care about your net worth. It doesn’t care about your spending. A household with $2 million invested can qualify for thousands in premium tax credits if its MAGI is low enough, because MAGI is the only test written into the law.
That means every early retiree runs two budgets whether they know it or not. The first is your spending budget, the money that actually leaves your accounts. The second is what I call your MAGI Budget, the income the IRS sees. Most people assume those are the same number. They’re not even close, and the gap between them is where all the money is.
Chris Mamula, who retired from physical therapy at 41 and writes at Can I Retire Yet, nailed the reason this works. Early retirees “often have considerable control” over how they generate the income that funds their spending. An employee can’t choose what their paycheck looks like. You can.
Here’s which dollars land in your MAGI Budget and which ones don’t.
| Counts toward MAGI | Doesn’t count |
|---|---|
| Traditional 401(k) and IRA withdrawals | Cash savings you already have |
| Roth conversions, in the year you convert | The original cost of investments you sell |
| Capital gains, even ones taxed at 0% | Roth IRA contributions you pull back out |
| Dividends and interest | Qualified HSA reimbursements |
Read that third row again. A capital gain with a 0% tax rate still counts as marketplace income. Cheap for taxes, expensive for subsidies. That one catches everybody.
So the same $70,000 of spending can show up to the IRS as $110,000 of income or as $55,000 of income, depending entirely on which accounts it comes from. You spent a decade or two building those accounts. Now they’re your control panel.
The 2026 Cliff in Plain Numbers
Quick history, because the rules just changed on everyone.
From 2021 through 2025, Congress ran enhanced subsidies with no income cap. Those expired on December 31, 2025. The House passed a three year extension on January 8 and it stalled in the Senate, so as I write this in July 2026, nothing has passed. Congress could still change the rules, maybe even retroactively. Plan for the rules as they’re written today.
The rule written today is the cliff. Earn up to 400% of the federal poverty level and the government helps pay your premiums. Earn one dollar more and you get nothing. Here’s where the line sits for 2026 coverage, which runs off the 2025 poverty guidelines from HHS.
| Household size | Your 2026 cliff |
|---|---|
| 1 | $62,600 |
| 2 | $84,600 |
| 3 | $106,600 |
| 4 | $128,600 |
400% of the 2025 federal poverty guidelines, 48 contiguous states. Alaska and Hawaii run higher.
Stay under your line and the math is friendly. You pay somewhere between roughly 2% and 10% of your MAGI for the benchmark plan, which is just the second cheapest silver plan where you live, and a premium tax credit covers the rest. The exact percentage comes from an IRS table updated every year, Rev. Proc. 2025-25 for 2026. Lower income, lower percentage.
The cliff is brutal for people who can’t control their income. KFF’s May 2026 enrollment data shows households earning just over the line made up 3% of 2025 marketplace signups but accounted for 27% of the drop in signups for 2026. They got hit with full price and walked away.
You’re going to read that stat differently than they did. Their income was a fact. Yours is a decision.
The New Penalty for Guessing Wrong
One more rule change, and it’s a big one. Through 2025, if you underestimated your income and collected too much subsidy during the year, the repayment was capped for most households. That cap is gone. For tax years after 2025, the IRS says you repay every dollar of excess credit, no matter your income.
Cynthia Cox, who runs the ACA program at KFF, told CNBC in January that the repayment “could easily be $10,000” for households that guess wrong, and that an older couple could owe around $20,000.
The lesson isn’t fear. The lesson is buffer. Never plan your MAGI right up against your cliff number. I target 10% under mine, which leaves room for a surprise fund distribution or a dividend bump without blowing up the whole year.
What Staying Under the Line Is Actually Worth
Let’s use a real household shape. Two adults, one kid. Family of three, cliff at $106,600. Mine is a family of four, so my own line sits at $128,600, and I’ll come back to that.
Say they’re retired and running a MAGI Budget at $75,000. At 2026 national average premiums, the benchmark plan for a family that size costs $19,489 a year at full price. Their share would be about $7,047. The government picks up the other $12,443.
Twelve and a half grand. Every year. For managing one number.
| Your MAGI | You pay per year | Government pays |
|---|---|---|
| $70,000 | $6,177 | $13,312 |
| $80,000 | $7,968 | $11,521 |
| $90,000 | $8,964 | $10,525 |
| $106,600 | $10,617 | $8,872 |
| $106,601 | $19,489 | $0 |
That last table row is the cliff. One dollar, and $8,872 vanishes.
Now, about a friend of mine. He makes $120,000, has a family of four, and told me he could never afford to retire early because of health insurance. I told him he was flat wrong, and that with a plan he could cover his whole family for under $15,000 a year.
I ran the math for this article, and I owe him a correction. I was too pessimistic. A family of four with a MAGI anywhere under its $128,600 cliff pays between $6,718 and $12,809 a year for the benchmark plan at national average prices. Under $13,000 at any qualifying income. Under $7,000 if the MAGI Budget lands near $80,000. And the credit behind those numbers runs five figures across the entire range, from $11,170 at the very top of the income scale up to $21,116 lower down.
So what does ignoring all this cost? I modeled two identical families. Same portfolio, same $70,000 of annual spending, same everything. One runs a MAGI Budget. The other funds everything from traditional withdrawals and stacks a $35,000 Roth conversion on top, because that’s what the standard FIRE playbook says to do.
| Runs a MAGI Budget | Ignores MAGI | |
|---|---|---|
| MAGI the IRS sees | $75,000 | About $113,000 |
| Health insurance per year | $7,047 | $19,489 |
Same life. Same spending. The second family pays $12,443 more every single year, which makes their identical lifestyle about 18% more expensive. Run it as a family of four and the gap hits $17,261, almost 25% more. And it widens with age, because full price premiums climb every birthday while the subsidized share doesn’t. By the mid 50s the gap runs past 27% of that same $70,000 lifestyle.
Here’s the part that should actually change your plan. At a 4% withdrawal rate, covering an extra $12,443 of permanent annual spending requires $311,000 more in your portfolio. The family of four version needs $432,000 more. Ignoring one number on your tax return can inflate your entire FI number by three to four hundred grand.
That’s years of extra work, handed to your employer for free. And a five figure cost surprise landing in your first retirement years is exactly the kind of hit that makes sequence of returns risk dangerous.
Nobody wastes that kind of money on purpose. People waste it because they never heard of the MAGI Budget.
Healthcare blindness is a quiet mistake. It never feels like a decision, and it costs six figures. I put the ten biggest ones in a free guide. Get the 10 Quiet Mistakes That Kill Your Early Retirement and see which ones are hiding in your plan.
The Roth Conversion Problem Nobody Wants to Talk About
Now for the part that’s going to annoy some people.
The FIRE playbook says convert traditional money to Roth every single year of early retirement. Fill the low brackets. Build the ladder. I believed it too, and honestly, I need a lot of conversions in my own plan.
But every converted dollar is a MAGI dollar. And in 2026, that changes the price of a conversion completely.
Under the cliff, each extra dollar of MAGI also claws back roughly 10 to 18 cents of subsidy on top of the income tax, depending on where your income sits. So a conversion you priced at 12% might really cost you 22% or more.
And the dollar that crosses the cliff? Let’s price it. Say that family of three sits at $95,000 of MAGI and converts $15,000 because the bracket looks cheap. That pushes them to $110,000, over the line, and vaporizes the entire $8,872 credit. Even at a 12% tax rate, that conversion costs about $1,800 in tax plus $8,872 in destroyed subsidy. $10,672 to move $15,000. A 71% toll.
Would you volunteer for a 71% tax? Me neither. I already think the government manages money worse than the average American household, and that’s saying something. Ok, that part is my opinion. The 71% is just math.
Optimizing for subsidies and optimizing for Roth conversions pull against each other. They don’t go together, and pretending otherwise is how smart people quietly torch five figures a year.
Here’s what everyone writing about this conflict misses. Almost all of it is written for people retiring at 58 or 62, racing to drain their traditional accounts before required withdrawals start. You’re not them. Retire at 40 and required minimum distributions don’t touch you until 75 under current law. That’s 35 years of runway. You don’t need giant conversions in any single year. You need consistent ones, sized to fit.
The Decision Rule I Use
Three steps, once a year.
First, set your MAGI ceiling. Take your cliff number from the table above and knock 10% off for buffer. For my family of four, that’s about $115,740.
Second, add up the income you can’t avoid. Dividends, interest, fund distributions, plus the gains baked into whatever you’ll sell for spending money.
Third, the gap between those two numbers is your conversion budget for the year. Convert exactly that much. If the gap is zero, convert zero. Say it with me. Zero is a valid conversion.
Some years the ladder math will genuinely demand more. Fine. Then make it a deliberate trade. Run a heavy conversion year where you knowingly give up the subsidy, then take it back with lean MAGI years after. What you never do is stumble over the cliff by accident because a spreadsheet from 2019 told you conversions are always good.
The full mechanics of conversion ladders, and every other way to reach your money early, live in my guide to accessing retirement accounts before 59½.
Yes, This Is Gaming the System
Every time this topic comes up, somebody says it. That’s just gaming the system.
Obviously. You can turn anything in life into a game, and why would you play a game you’re not trying to win? The rules are public. The IRS prints them. The people who wrote those rules, and the wealthy people who fund their campaigns, play every single one of them. Refusing to play doesn’t make you principled. It makes you poorer.
A millionaire household qualifying for subsidies isn’t a loophole. MAGI is the test Congress chose, on purpose. Managing your MAGI is the same species of move as contributing to a 401(k) to lower your taxes, and nobody calls that cheating.
Now let me make the strongest version of the real objection, because it’s the one I actually lose sleep over. This subsidy could disappear. Congress built the enhanced version in 2021, let it die in 2025, and has been fighting about it ever since. Anyone who tells you these rules will look the same in ten years is guessing.
So here’s my answer, and it’s the same answer I gave myself. Stress test your plan at full price. A family of three at full 2026 premiums is $19,489 a year at today’s national average, and my own FI number has to carry the family of four version of that same line. If your plan only survives with subsidies, your plan isn’t finished.
Then, once the plan works without the help, take every dollar of help the current rules offer. The subsidy is upside you harvest one year at a time, for as long as the rules say you can. Flexibility beats prediction. Every time.
Your One Hour MAGI Budget
Everything above collapses into about an hour of work, once a year.
- Find your line. Grab your cliff number from the table for your household size.
- Set your ceiling. Multiply it by 0.9. That’s your MAGI Budget with the buffer built in.
- Map your money. List where next year’s spending comes from and what each source adds to MAGI. Cash and cost basis add nothing. Traditional withdrawals and conversions add every dollar.
- Size the conversion. Ceiling minus unavoidable income equals your conversion budget. Could be $30,000. Could be zero.
- Sanity check it. Plug your numbers into KFF’s marketplace calculator with your actual zip code, because your state’s premiums are what count, not national averages.
Then set a calendar reminder for early December to recheck before any year end moves. Automate the reminder so it never depends on your memory, and GGs. You’re done until next year.
One more cheat code while you’re in there. Starting in 2026, every bronze marketplace plan counts as HSA eligible under IRS Notice 2026-05, and HSA contributions lower your MAGI. So the cheapest plan tier now comes with a lever that pulls your income further under the cliff. The system handed you that one for free.
If you can build a FIRE spreadsheet, you can do this in your sleep. It’s not a maze. It’s one number with a line you stay under, and staying under it pays better per hour than almost anything else in personal finance. The weak point in every plan is the person running it, so build the system once and let it run.
Health insurance was the number one reason people told me retiring at 40 is impossible. It’s affordable and very doable. You just need a plan, like everything else.
Quick Answers
Does My Net Worth Affect ACA Subsidies?
No. The only test is MAGI. A household with $2 million invested and $70,000 of MAGI qualifies exactly like anyone else at $70,000.
Do Capital Gains Taxed at 0% Still Count Toward MAGI?
Yes. The 0% rate is a tax rule and MAGI is an income rule. Realized gains count in full even when the tax on them is zero.
Do Roth Withdrawals Count?
Pulling back your original Roth IRA contributions adds nothing to MAGI, and neither do qualified withdrawals. Conversions count in full in the year you convert.
What Happens If I Go Over the Cliff by Accident?
You repay every dollar of advance credit at tax time, with no cap for 2026 and beyond. That’s what the 10% buffer is for. If you’re close in December, delaying a sale can pull you back, and an HSA contribution helps too if you’re on an HSA eligible plan.
Is Any of This Legal?
It’s the formula printed in the tax code. Managing MAGI is the same kind of move as contributing to a 401(k) to cut your tax bill. The rules are public, and you’re allowed to read them.
Healthcare just moved from terrifying to one hour a year.
Now go find what else is quietly working against you. Get the 10 Quiet Mistakes That Kill Your Early Retirement, free, and close the other nine gaps this week. Send Me the Guide
I write about my own strategy and my own plan. The free guide linked above is genuinely free, and it puts you on my email list where I also sell The Early Retirement Blueprint. I am not compensated by any custodian, broker, or insurer mentioned here. This is educational content, not personalized tax or investment advice. Tax rules change and your situation is not mine, so confirm anything here against current IRS guidance or a professional before acting on it.
