How long will my money last?

Here's the short version. Grow the pot a little each year, take the draw out, and count the years until it hits zero. Dan and Ruth's $740,000, paying $30,000 a year and growing 2% above prices, runs dry in year 35. That sum assumes every year is average, and they aren't, so then we play out 1,000 futures and count how many still have money: 93%.

All in today's dollars, before tax.

First, the idea: a pot, a draw and the year it hits zero

Think of a pot with a tap at the bottom. Each year it grows a bit, and then the draw comes out. Dan and Ruth Keller, a fictional couple, have $740,000 and need $30,000 a year that their checks don't cover. Say it grows 2% above prices. It goes down slowly at first, then faster, and reaches zero in year 35. One sum, one number.

Then the rule: a small change moves the end year a lot

The draw rises with prices, so what matters is growth above prices, and that's why we leave prices out and work in today's dollars. Here's the year the same $740,000 runs dry:

Growth above prices$25,000 a year$30,000 a year$40,000 a year
0%Year 30Year 25Year 19
2%Year 46Year 35Year 24
4%NeverOver 60 yearsYear 35

Read the middle row. Taking $25,000 instead of $40,000 is nearly twice as many years, and that's only one thing changed.

Then the twist: no one gets the average year

That sum assumes every year is average, and markets aren't. This is what Monte Carlo is for.

A Monte Carlo run doesn't predict one future. It plays out a thousand of them, each drawing its yearly returns from the way markets have actually behaved (ours can draw real years from 1928 to 2025), and counts how many of the thousand still have money at the end. If 850 do, the chance is 85%. Two things to take from it: the chance, not the average; and what moves the chance when you change one thing. [NYU Stern]

One thing that surprises people. A mix of 60% stocks has averaged about 6% a year above prices over those years, well above the 2% in our sum, so the odds can look kinder than the sum does. The sum is a cautious guess at one smooth future. The odds are about the bumps.

Three things to hold onto

  1. Read the chance first. If it's under about three in four, change one thing (the draw, the year you start, the age you claim Social Security) and watch which moves it most. Three in four is our rule of thumb, not an official line.
  2. Look at the worst runs, not the middle. A bad first few years hurt far more than a bad later one, because the draw is coming out of a pot that's just shrunk.
  3. This page runs a small one. Step 4 is the full tool, with your checks starting at the ages they really start.

Example 1: Dan and Ruth, year by year, then the odds

Here's the sum written out from 64, every fifth year, to 95:

End of yearAgeLeft in the pot
569$660,899
1074$573,564
1579$477,140
2084$370,680
2589$253,139
3094$123,365
3195$95,833

So they get to 95 with money left, and it runs dry in year 35, when they're 98. Now we ran the same sum changing one thing each time:

RunRuns dryLeft at 95
As above: $30,000, 2% above pricesYear 35$95,833
$5,000 a year lessYear 46$307,730
$10,000 a year moreYear 24$0
As above, no growth above pricesYear 25$0
As above, 4% above pricesOver 60 years$716,269

Rows 2 and 3 change the draw. $5,000 a year less buys 11 more years. $10,000 more costs 11. Rows 4 and 5 change the growth, the part nobody gets to pick. Now the same $740,000 and $30,000 a year against 1,000 runs of real years, to 95:

Draw a yearRuns with money leftChanceWorst one in ten, left at 95
$30,00092893%$93,848
$25,00097698%$358,941

Read the chance: 93%, and 98% with $5,000 less. This is a simple version with a level draw from today. Step 4 starts them on $78,000 a year with nothing from the checks until Dan and Ruth claim. There, drawing the same real years, the chance is 91%. Step 4 also has a smoother bell-curve model, cautious about returns, that says 68%. Different questions, different answers, and that's worth knowing before you trust any one number.

Dan and Ruth Keller are fictional; their figures are illustrative, not advice for you.

Your turn

The Kellers' numbers are in the form, so you can see it work before you touch it. Try one change: take the draw from $30,000 to $40,000 and watch the odds fall, while the growth you typed makes no difference to them. Then press "Clear the form" and put your own in. If you don't know your draw, it's what goes out in a year less what your checks pay.

The Kellers' numbers are in the form. Type over any of them. Your numbers stay in this browser; nothing is sent to us.

Empties the pot and the draw so you can type your own.

The pot and the draw

It's all in today's dollars and before tax.

The answer

On paper: $740,000 pays $30,000 a year for all 31 years, with $95,833 left, and would run dry in year 35.

The year-by-year table
YearStartGrowthDrawEnd
1$740,000$14,800$30,000$724,800
2$724,800$14,496$30,000$709,296
3$709,296$14,186$30,000$693,482
4$693,482$13,870$30,000$677,352
5$677,352$13,547$30,000$660,899
6$660,899$13,218$30,000$644,117
7$644,117$12,882$30,000$626,999
8$626,999$12,540$30,000$609,539
9$609,539$12,191$30,000$591,730
10$591,730$11,835$30,000$573,564
11$573,564$11,471$30,000$555,036
12$555,036$11,101$30,000$536,136
13$536,136$10,723$30,000$516,859
14$516,859$10,337$30,000$497,196
15$497,196$9,944$30,000$477,140
16$477,140$9,543$30,000$456,683
17$456,683$9,134$30,000$435,817
18$435,817$8,716$30,000$414,533
19$414,533$8,291$30,000$392,824
20$392,824$7,856$30,000$370,680
21$370,680$7,414$30,000$348,094
22$348,094$6,962$30,000$325,055
23$325,055$6,501$30,000$301,557
24$301,557$6,031$30,000$277,588
25$277,588$5,552$30,000$253,139
26$253,139$5,063$30,000$228,202
27$228,202$4,564$30,000$202,766
28$202,766$4,055$30,000$176,822
29$176,822$3,536$30,000$150,358
30$150,358$3,007$30,000$123,365
31$123,365$2,467$30,000$95,833

The odds

In 1,000 runs drawing real years, 928 still have money after 31 years. That's a chance of 93%. The worst one in ten still had $93,848 at the end. That's the bad-luck end of what's happened before, not a floor.

The same pot
On paper: the year it runs dryYear 35
Chance it lasts the plan93%
Worst one in ten$93,848 left

These runs are 60% stocks. Step 4 changes the mix and adds your checks at the ages they start. It opens with your balances from step 1, not this pot.

Example 2: 60 and single, $500,000, drawing $25,000

Carol is 60 and plans to 95. She needs $25,000 a year from $500,000, which is 5%.

Carol
On paper at 2% above prices: runs dryYear 26, at 85
On paper at 4% above prices: runs dryYear 42, past 95
The odds, 35 years75%
Worst one in tenDry by year 24, at 83

Read lines 2 and 3 together. Pick a friendly 4% and the plain sum says she's fine, past 95. The odds say 750 in 1,000: three in four, right on the line, and one run in four runs dry. The plain sum flatters it, because it has no bad years in it.

Carol is fictional; her figures are illustrative, not advice for you.

Example 3: 66, with a pension that covers most of it

Joan is 66. Her pension and her check bring in $54,000, and $58,000 goes out, so she only draws $4,000 a year from $250,000.

Joan
On paper at 2% above prices: runs dryNever
The odds, 29 years100%
Worst one in ten, left at 95$307,525

Read line 1: never. At 2% the pot earns more than she takes. Her question isn't whether it lasts. It's what a new roof or a bad year of health costs, and that's what the pot is for.

Joan is fictional; her figures are illustrative, not advice for you.

Example 4: 60, drawing 6%, where the number to change is the draw

Walt is 60, with $400,000, and he's drawing $24,000 a year: 6%. On paper it runs out when he's 80, and the odds are poor. Here's what each $5,000 less does:

Draw a yearOn paper, runs dryAt ageThe odds, to 95
$24,000Year 218055%
$19,000Year 288778%
$14,000Year 4310295%

Read the last column, top to bottom. 55% is close to a coin toss. $5,000 less takes it to 78%, and another $5,000 to 95%. Growth isn't something he gets to pick, and the draw is, which is why this is the number that does the work. Whether he can live on less is another question, and that's step 2.

Walt is fictional; his figures are illustrative, not advice for you.

What to make of it

  • High chance, money left in the worst runs? The sums say it holds up, before tax.
  • Under about three in four? Change one thing and see which moves it most. The draw usually does.
  • The sum and the odds disagree? That's normal. The sum shows how the pot behaves; the odds show how much luck is in it.
  • The part people miss: the first few years count the most, so two people with the same average can end up in very different places.

What this doesn't cover

Tax, which is step 5. Checks that start later: here the draw is level from year one, and step 4 lets the checks start at their real ages. Spending that changes with age, which is step 2. Fees. And history is only history. The next thirty years aren't in the table.

Next up: step 4, will it last? Change the mix and add your checks.

Questions people ask

How long will my money last in retirement?

Grow the pot by what it earns above prices, take out the year's draw, and repeat until it hits zero. $740,000 paying $30,000 a year, growing 2% above prices, runs dry in year 35. Then test it against real market years, because no one gets the average every year.

How long will $500,000 last in retirement?

It depends on the draw and the growth. At $25,000 a year it runs dry in year 26 at 2% above prices and year 42 at 4%. In 1,000 runs of real market years it lasted 35 years 75% of the time.

What is a Monte Carlo simulation?

It doesn't predict one future. It plays out a thousand of them, each drawing its yearly returns from how markets have actually behaved, and counts how many still have money at the end. If 850 do, the chance is 85%.

What is a good Monte Carlo success rate?

There's no official line. On this page we use about three in four as the point to start changing one thing and watching what moves. A high chance often means money left over; a low one means a real risk of running short.

Why does growth above prices matter, not growth?

Because the draw rises with prices. If the pot grows 5% and prices rise 3%, only the 2% above prices is helping you. So we work in today's dollars and leave prices out.

Where these numbers come from

Yearly returns on stocks and bonds, 1928 to 2025: NYU Stern, Damodaran. Inflation: Federal Reserve Bank of Minneapolis. Both checked 31 October 2026. Each run draws whole calendar years at random, 60% stocks and 40% bonds, rebalanced yearly, 1,000 runs with a fixed seed so the page always says the same thing. The rest is a simple sum, and the growth above prices is an assumption you can change.

This is general education with calculators that run on numbers you enter. It is not advice about your situation and not a recommendation to claim, invest, buy or sell anything. The author is not a financial adviser, is not licensed in any state, and has no relationship with you. Social Security's own tools and a flat-fee adviser (see step 7) can check your decision. Your numbers stay in this browser; nothing is sent to us.