The Purpose of a Monte Carlo Simulation in Retirement Planning
A Monte Carlo simulation runs your plan through thousands of possible futures instead of one average, which turns "will my money last?" into a probability you can measure. Here's how it works, how to read the result, and where it falls short.
3.9%. Morningstar uses that starting withdrawal rate as the one that survives 90% of simulated futures over a 30-year retirement. Look at how it's phrased. It doesn't say "the average return" or "a typical year." It says 90% of futures. That wording is the purpose of a Monte Carlo simulation. It swaps one guess about the future for thousands of possible futures, so you can see how often a plan works instead of only whether it works on average.
Source: Morningstar, The State of Retirement Income 2025 · Dec 2025
This article covers what the method does, where the name comes from, how to read a success rate, and where the method falls short.
The short answer: what is the purpose of a Monte Carlo simulation?
A Monte Carlo simulation is built to measure uncertainty. You run the same model over and over, and each run draws its uncertain inputs at random. You end up with a range of outcomes and the odds of each one, where a normal calculation would give you one number.
MIT News describes the method as running a model many times with random inputs to see the spread of possible results. The NordVPN glossary makes the same point: a single-number estimate hides the range of outcomes that could actually happen.
For retirement, the question shifts:
- Without Monte Carlo: "If markets return X% a year, my money lasts until 91."
- With Monte Carlo: "In 870 out of 1,000 simulated futures, my money lasts to 95."
(That second line is a made-up example, not a benchmark.)
The first statement sounds exact, but it describes only one possible future. The second tells you how often your plan holds up across many of them.
Why one average return can mislead you
Most simple retirement calculators assume a steady return, the same percentage every year. Real markets don't work like that. They rise, fall, and sometimes fall hard just after you stop working.
The order of returns matters once you're withdrawing money. Take two retirees with the same average return over 30 years. One gets the bad years early and the other gets them late, and they can end up with very different results. The first sells investments while prices are down to cover groceries, so that money isn't there for the recovery. A straight-line average treats both people the same. A Monte Carlo simulation doesn't.
So the method is more than a way to make the math look sophisticated. You use it because the average is the wrong question. What you need to know is how bad things could get and how often.
How it works, in plain English
You don't need the formulas to see what's going on. A retirement Monte Carlo run looks roughly like this:
- Start with your facts. Age, savings, monthly spending.
- Name what's uncertain. Market returns, inflation, how long you live, health care costs.
- Roll the dice. For one simulated future, the model draws a year-by-year path of returns at random and plays your plan forward: withdrawals out, growth or losses in, year after year.
- Record the result. Did the money last to the end of the plan? Yes or no.
- Repeat. Hundreds or thousands of times, each with a different random path.
- Count. The share of futures where the money lasted is your success rate.
That's all there is to it: lots of possible futures, then a count of how many worked.
Where the name comes from
The method started in physics. Researchers first developed and used it systematically during the Manhattan Project. John von Neumann and Stanislaw Ulam proposed it to study how neutrons travel through radiation shielding, a problem too messy to solve directly (Monte Carlo method, Wikipedia mirror; University of Washington, Monte Carlo history).
The idea came from a card game. In 1946, while Ulam was recovering from an illness, he wondered about his odds of winning a game of Canfield solitaire. Working them out exactly was hopeless. Playing many hands and counting the wins was easy.
Nicholas Metropolis suggested the name "Monte Carlo," after the casino in Monaco where Ulam's uncle gambled with money borrowed from relatives. The name refers to chance, which is what the method is built to measure.
What a Monte Carlo simulation can tell you about retirement
Here is where the method pays off for your own plan. Each of the risks below can't be captured by one number, and each one fits naturally into a simulation.
| Retirement risk | What the data says | Why a single estimate falls short |
|---|---|---|
| Withdrawal rate | Morningstar puts the safe starting rate at 3.9%, judged by money remaining in 90% of simulated outcomes (The State of Retirement Income 2025, Dec 2025) | An average return can make a higher rate look safe when it fails in a meaningful share of futures |
| Longevity | About 1 in 3 65-year-olds may live to at least 90 (Social Security Administration, Retirement Information for Medicare Beneficiaries, 2026) | A plan that ends at 85 is betting against a third of retirees |
| Health care | Fidelity estimates $172,500 in Medicare premiums, deductibles, and prescriptions for one 65-year-old retiring in 2025, excluding long-term care (2025 Retiree Health Care Cost Estimate, Jul 2025) | A flat guess hides costs that land year by year and change with income |
| Not knowing your number | Only 52% of workers have estimated the monthly income they'll need in retirement (EBRI/Greenwald Retirement Confidence Survey 2025, Fact Sheet 3) | You can't stress-test a figure you've never written down |
That's why the plan you build at Monte Climb runs to age 95 by default, and you can extend it. It also charges Medicare premiums and income-based surcharges year by year instead of using a single flat health care number.
How to read a success rate
A result like "85% success" is easy to misread. Here's what it means and what it doesn't.
What it means
Out of every simulated future, 85% ended with money still in the account at the end of the plan, and 15% ran out first. It answers one question: how often does this plan survive?
What it doesn't mean
- It isn't a promise. An 85% result doesn't mean you'll be fine. Your life will play out as one future, not 1,000.
- It isn't a grade. A 15% failure rate doesn't mean you'll be broke at 70. A lot of failed futures run short late in life, when smaller adjustments might have covered the gap.
- It isn't a forecast of returns. The simulation doesn't predict markets. It shows how your plan holds up across many ways markets could behave.
A probability is a measure of how much you're relying on luck. Treat it as a starting point for decisions, such as spending a little less, working a little longer, or delaying a withdrawal, and not as a verdict.
How many futures is enough?
This is where many explanations stop at "thousands of runs." There's a real rule behind the number of runs.
Monte Carlo accuracy improves slowly. The standard error of the estimate shrinks with the square root of the number of runs (σ/√N). To make the estimate 10 times more precise, you need 100 times as many runs (MIT OpenCourseWare 16.90, Error Estimates for the Monte Carlo Method). About 95% of the time, the true answer falls within two standard errors of the estimate.
Here is that formula applied to a success rate. This is our own arithmetic, not a quoted figure. At 1,000 runs, an 85% result is accurate to within roughly 2 percentage points either way, so the true rate is somewhere around 83% to 87%. Going to 100,000 runs would narrow that to about a quarter of a point.
Is that extra precision worth it? Usually not, because sampling noise isn't the main source of error. Your assumptions are.
The limits: garbage in, garbage out
Every serious explainer, including Learn Lean Sigma's guide, gives the same warning: a simulation is only as good as its inputs. More runs won't fix unrealistic assumptions about returns, inflation, or how those factors move together.
Watch for these:
- Optimistic return assumptions. If the model expects strong markets, every future looks rosier.
- Spending that's too low. If you understate your monthly spending, the success rate goes up and tells you nothing true.
- A plan horizon that's too short. Stopping at 85 ignores the roughly one in three 65-year-olds who may reach 90.
- Missing costs. Leaving out health care, or treating it as a rounding error, inflates the result.
The honest takeaway is that a Monte Carlo simulation makes your assumptions visible and testable. It doesn't make them correct. That's a reason to stress-test them on purpose. Raise spending, shorten the early good years, and stretch the horizon, then see what breaks.
When you don't need one
Monte Carlo isn't always the right tool. A simpler approach is fine when:
- The outcome is basically certain. If you have a pension and Social Security that cover your spending with room to spare, there's little uncertainty to measure.
- You're comparing two specific scenarios. "What if I retire at 62 instead of 65?" can start with a side-by-side comparison, though running each through a simulation tells you more.
- You don't have a spending number yet. Start there. The simulation needs your monthly spending to say anything useful.
For most people trying to work out when they can stop working, though, the uncertainty is the whole problem. That's the case Monte Carlo was built for.
Run your own numbers
The method was built for questions with no single answer, and "when can I retire?" is one of them.
You can try it yourself on Monte Climb's retirement calculator. Enter three numbers: your age, what you've saved for retirement, and what you spend each month. We run them through 1,000 simulated futures and show what share stay funded. It runs in your browser, and nothing is saved until you say so. If the result raises questions like Am I on track? Could I retire at 62? What spending is sustainable?, ask William, our AI planner, who shows his work.
It's planning software, not advice. The result is a probability, not a promise, and it gives you something better than a guess to work from.
Frequently asked questions
What is the purpose of a Monte Carlo simulation in simple terms?
It measures uncertainty. A Monte Carlo simulation runs a model many times with randomly drawn inputs, which gives you a range of outcomes and their odds in place of one guess. In retirement planning, that means you see the share of simulated futures in which your money lasts.
How many runs does a Monte Carlo simulation need?
Accuracy improves with the square root of the number of runs, so a 10-fold gain in precision takes 100 times as many runs. At 1,000 runs, a retirement success rate is typically accurate to within a few percentage points. Unrealistic assumptions usually cause more error than the number of runs does.
What does a Monte Carlo success rate mean for retirement?
It's the percentage of simulated futures in which your savings lasted to the end of the plan. An 85% success rate means 850 of 1,000 futures stayed funded. It's a probability, not a guarantee, and it's most useful for comparing choices such as spending, retirement age, or plan length.
Why is it called a Monte Carlo simulation?
Nicholas Metropolis suggested the name after the Monte Carlo casino in Monaco, where Stanislaw Ulam's uncle gambled. Ulam and John von Neumann developed the method during the Manhattan Project, partly inspired by Ulam wondering about his odds of winning at solitaire in 1946.
Is a Monte Carlo simulation more reliable than an average-return calculator?
It's more informative because it shows how often a plan fails, including when bad market years arrive early in retirement. An average-return projection shows only one path. Both depend on realistic inputs, though, and running more simulations won't fix bad assumptions.
When can you actually retire?
Three numbers. We run your savings through 1,000 simulated futures, the same engine behind the app, and give you one clear answer.
Show my answerGet new posts by email
We'll email you new Monte Climb posts once the newsletter starts. Nothing has been sent yet.
We keep only your email address and the post you signed up from, and delete it whenever you ask. Privacy policy
