72(t) vs the rule of 55: one couple at 55, three ways into the 401(k)
Leave your job at 55 with most of your savings in a 401(k), and the money is yours but not easy to reach. Withdrawals before 59½ normally carry a 10% penalty on top of income tax. Two rules in the tax code get around it: the rule of 55, which opens the 401(k) of the job you leave, and a 72(t) schedule of fixed payments, which works on an IRA. Most explanations stop at the rules. We ran one couple through both, and through the simplest alternative of paying the penalty, with all five of Plan Mirror's tests.
The answer surprised us a little. For this couple the route decides how much of their budget goes to the IRS in the first five years. It barely changes whether the money lasts. That's decided by something else.
The household
A married couple in Texas, which has no state income tax, both 55, leaving work this year and planning to 95. That's 41 years of retirement, through 2066.
- Household
- A couple, both 55
- Retires
- At 55, in 2026
- Plans through
- Age 95 (2066)
- Invested
- $1,600,000: $150,000 taxable brokerage, $1,350,000 traditional 401(k)/IRA, $100,000 Roth
- Spends
- $80,000 a year in today's dollars, taxes and healthcare included
- Social Security
- $40,000 a year combined, in today's dollars, from 67
- 401(k)/IRA before 59½
- Not touched until 59½; the Rule of 55 for the employer plan
- Assumes
- 6% average return, 2.5% inflation
- Taxes
- Married filing jointly, Texas (2026 rules)
Almost all their savings, $1,350,000, sits in the 401(k) from the job they're leaving. The brokerage account holds $150,000, enough for about two years of spending. So unlike a household with a large brokerage account, they can't simply wait for 59½. They need the 401(k) within a couple of years.
The spending figure is all-in. It covers living costs, health insurance until Medicare at 65, and the household's income tax. Plan Mirror pays the tax, and any penalty, out of the same $80,000. So $80,000 is what leaves the portfolio, and every dollar paid to the IRS is a dollar the couple doesn't get to spend.
The three routes
The rule of 55. If you leave an employer in or after the calendar year you turn 55, you can take money from that employer's 401(k) or 403(b) without the 10% penalty. It doesn't cover IRAs, or the plan of an employer you left before the year you turned 55. Roll the 401(k) into an IRA after you leave and the exception is gone. Withdrawals are flexible: take what you need, when you need it. Whether your plan allows partial withdrawals after you leave is up to the plan, so check before you rely on it. This is the couple's main plan, and the one on this page's charts.
A 72(t) schedule. Under section 72(t), substantially equal periodic payments come out penalty-free at any age. The IRS allows three ways to set them: two give a fixed annual payment from the balance, your life expectancy and an interest rate, and one recalculates the payment each year. The schedule has to run for five years from the first payment or until 59½, whichever is later, and a change the rules don't allow generally brings back the 10% on every payment so far, with interest. In this version the couple rolls the 401(k) into an IRA and takes a fixed $60,000 a year from 2026 through 2030. That amount is an illustration: whether a real schedule qualifies depends on the IRS calculation, which we didn't run. The schedule pays out in full every year whether they need it or not. They take nothing else from the IRA before 59½; the brokerage account pays the rest.
Pay the penalty. The same rollover with no schedule: draw from the IRA as needed and pay the 10% on early withdrawals.
Answer one: steady returns
Plan Mirror's first test grows the portfolio at 6% a year, raises spending with 2.5% inflation and works out each year's taxes. On that path the couple finishes at 95 with $3.2 million, about $1.2 million in today's dollars. All three routes end with exactly that. The money leaves the portfolio in the same amounts each year; only the split between the couple and the IRS changes.
The split is where the routes differ.
Rule of 55: no tax for two years, then a steady bill. Plan Mirror spends the brokerage account first. Most of each sale is the couple's own money coming back, so there's no income tax until 2028. Then withdrawals come from the 401(k), all taxable as income, and the bill is $5,505 that year and $5,784 by 59. No penalty, ever.
72(t): a small bill from the first year. The $60,000 payment is taxed as income every year, starting at $2,840 in 2026 and easing to $2,446 by 2030. That's more than the rule-of-55 plan pays in the first two years and less in each of the next three, because the schedule spreads taxable income evenly from the start instead of bunching it later. Drawing on the IRA from the first year also leaves more of the brokerage account for later, and in 2031, the year after the schedule ends, the couple pays $0 in tax, against $5,929 on the rule of 55.
Pay the penalty: the same withdrawals, plus 10%. The brokerage account carries the first year. From 2027, IRA withdrawals pay the 10% on top of income tax: $8,405 in penalty in 2028 alone, and $26,631 over the four years Plan Mirror treats as before 59½ (it counts by calendar year, so access opens in the year they turn 60). All of it comes out of the $80,000.
| Age | Year | Spending | Taxes | Social Security | From portfolio | Portfolio at year end |
|---|---|---|---|---|---|---|
| 55 | 2026 | $80,000 | $0 | $0 | $80,000 | $1,611,200 |
| 57 | 2028 | $84,050 | $5,505 | $0 | $84,050 | $1,629,116 |
| 59 | 2030 | $88,305 | $5,784 | $0 | $88,305 | $1,640,072 |
| 60 | 2031 | $90,513 | $5,929 | $0 | $90,513 | $1,642,533 |
| 67 | 2038 | $107,591 | $5,054 | $53,796 | $53,796 | $1,641,863 |
| 95 | 2066 | $214,805 | $17,372 | $107,403 | $107,403 | $3,166,778 |
Dollar figures in the table are in each year's own dollars. The $3.2 million at the end has the buying power of $1.2 million today.
Two quick stress tests
With spending 10% higher, the money still lasts, and the couple ends with about $453,000 in today's dollars. With returns 2 points lower, the portfolio runs out in 2064, at 93. The three routes give the same answers here too.
That's the first hint of what really decides this plan. $80,000 is 5% of $1.6 million, over 41 years, with Social Security 12 years away. The plan has room for extra spending but little for weaker returns.
Answer two: a century of real markets
The historical replay runs the plan through actual U.S. stock returns and inflation from every start date since 1926 Q3 with a full 41-year window behind it.
On the rule of 55, 225 of 237 start dates last to 95: 94.9%. The worst start is 1929 Q4, the Depression, and in that history the money is gone by 2038, at 67, the year Social Security was due to start.
The routes barely differ: 224 start dates last with the 72(t) schedule and 225 when paying the penalty. In the quarter of histories with the weakest first ten years, 20.3% of rule-of-55 plans run out, against 5.1% of all start dates. That gap is sequence-of-returns risk: a portfolio that shrinks in its first decade while paying the full $80,000 has too little left to recover. The route hardly moves it: 22% on the 72(t) schedule and 20.3% when paying the penalty.
The replay holds 100% U.S. stocks, and the past century was very good to them. The median history ends with $9.3 million in today's dollars. Nobody should plan on that repeating.
Answer three: 1,000 possible futures
The Monte Carlo test generates 1,000 random market paths. The typical one grows at the same 6% a year as the steady projection, with returns swinging as much as the stock market's (15% volatility a year).
On the rule of 55, 689 of 1,000 paths last to 95: 68.9%. The 72(t) schedule gives 689 and paying the penalty 691, a few paths in a thousand either way.
Almost all the failures come late. By 67, 2% of paths have run dry, against 31.1% by 95. And among the quarter of paths with the weakest first ten years, 83.2% run out. On a less generous future, this is a plan that works about two times in three, whichever door they use to reach the 401(k).
All three side by side
| Test | Rule of 55 | Roll over, 72(t) payments | Roll over, pay the 10% |
|---|---|---|---|
| Steady returns: the plan | Lasts to 95, $1.2 million left in today's dollars | Lasts to 95, $1.2 million left in today's dollars | Lasts to 95, $1.2 million left in today's dollars |
| Spending 10% higher | Lasts to 95, $453,000 left in today's dollars | Lasts to 95, $453,000 left in today's dollars | Lasts to 95, $453,000 left in today's dollars |
| Returns 2 points lower | Runs out in 2064, at age 93 | Runs out in 2064, at age 93 | Runs out in 2064, at age 93 |
| Historical replay | 225 of 237 paths last to 95 (94.9%) | 224 of 237 paths last to 95 (94.5%) | 225 of 237 paths last to 95 (94.9%) |
| Monte Carlo | 689 of 1,000 paths last to 95 (68.9%) | 689 of 1,000 paths last to 95 (68.9%) | 691 of 1,000 paths last to 95 (69.1%) |
Why doesn't the penalty hurt the odds? Because the tests measure whether the portfolio lasts, and the couple's spending is all-in. The $26,631 comes out of what they live on in their late fifties, not as an extra withdrawal from the portfolio. If they paid it and kept spending the same after tax, they would draw more each year, and these runs don't test that.
So which route?
For this couple, the rule of 55, if their 401(k) plan allows partial withdrawals after they leave. It costs no penalty and needs no commitment: in a bad year they can take less, and in a year with a big bill they can take more. Its one condition is easy to break by accident. They must leave the job in or after the year they turn 55, and they must not roll the 401(k) into an IRA before 59½.
A 72(t) schedule is the route for money the rule of 55 can't reach: an IRA, or the plan of a job left before the year they turned 55. For this couple it pays some tax in the first two years and less than the rule of 55 in the years after, because it spreads the income evenly from the start. The cost is flexibility. The $60,000 comes out every year until 2030, in good markets and bad, and a change the rules don't allow generally brings back the penalty on every payment so far.
Paying the penalty is the simplest route and the most expensive: $26,631 for this couple, in exchange for nothing the tests can measure.
But the bigger decision is elsewhere. All three routes share the same Monte Carlo odds, about 68.9%, that come with a 5% withdrawal over 41 years. A couple in this position should settle the access question in an afternoon, and spend the real effort on what they would do if the first decade is weak: spend less, or earn something for a few years.
What these runs leave out
- One plan's rules. We assumed the 401(k) allows partial withdrawals after they leave and treated the whole $1,350,000 as that one plan. If some of it sits in an IRA or the plan of a job left before 55, the rule of 55 doesn't reach that part. Plan Mirror's plan inputs treat all pre-tax savings as one account, so this page can't show that split.
- The exact 72(t) amount and end date. Plan Mirror runs the payment you give it; it doesn't apply the IRS formulas, and it ends the schedule in the year you set. The real end date depends on the date of the first payment and your birthday. Work out both with your IRA custodian or a tax professional before the first payment.
- The order of withdrawals. On the rule of 55, Plan Mirror spends the brokerage account first and the 401(k) after. Drawing a little from the 401(k) from the first year, as the 72(t) schedule is forced to, might save some tax. We didn't test that order.
- One tax profile. The household files jointly in a state with no income tax. A state income tax would come out of the same $80,000.
What we'll test next. Spending that stays level after tax, so a penalty or a heavier tax year draws more from the portfolio instead of from the budget. And the other way to time a 72(t): spending the brokerage account first and starting the schedule when it runs low.
Every figure on this page comes from 3 recorded cases run together in Plan Mirror under 2026 tax rules. Each was run through the plan with its two stress tests, a historical replay of U.S. stock returns and inflation from 1926 Q3 to 2026 Q2, and 1,000 Monte Carlo paths. It is an illustration, not financial, tax, investment or legal advice. See the methodology.
Related scenarios
- Can I retire at 47 with $2.1 million?
- Can I retire at 55 with $2 million?
- Does a cash buffer make retirement at 55 safer?
- The ACA subsidy cliff at 58: one IRA-heavy couple, three plans
Run these tests on your own plan in Plan Mirror.