Monte Carlo simulation replaces that single line with thousands of possible futures, so you can see how often your plan works and how badly it fails when it does not. In this guide you will learn what Monte Carlo simulation is, how to run one on your own portfolio, where it misleads people, and how to read the output like an analyst. By the end you will treat your plan the way a forecaster treats the sky: as a probability, not a promise.
A quick note on scope. This article is for any investor with a long horizon, whether you save in the US, Europe, Asia or elsewhere. All dollar figures are examples, so swap in your own local currency and the logic holds.
Monte Carlo methods use repeated random sampling to solve problems that are hard to solve exactly (Wikipedia, captured 11 October 2026).
What Is Monte Carlo Simulation?
Monte Carlo simulation is a technique that runs a model thousands of times with random inputs and then studies the spread of results. Mathematician Stanislaw Ulam conceived the method, and the name comes from the Monte Carlo Casino in Monaco, where Ulam's uncle liked to gamble. The idea is simple: if a system has too many moving parts to solve with a single formula, you can simulate it many times and count what happens.
In investing, the moving parts are annual returns, inflation, contributions and withdrawals. You feed the model an expected return and a volatility figure, which is the standard deviation of returns. For each simulated year the model draws a random return from that distribution. Chain 30 of those years together and you have one possible life for your portfolio. Repeat that 10,000 times and you have a map of outcomes.
Monte Carlo methods reached finance in 1964, when David Hertz described them in the Harvard Business Review for corporate decisions. Phelim Boyle then used simulation to value derivatives in a 1977 paper. Today the same idea sits inside most serious retirement planning tools.
86.2% is the success probability one popular free calculator shows by default for a $100,000 portfolio with $20,000 of annual contributions.
Why Monte Carlo Simulation Matters for Your Portfolio
A fixed-return projection hides the single biggest risk in investing, which is the order in which returns arrive. Two portfolios can both average 7% a year and end in completely different places, because a crash in year two hurts far more than a crash in year 28. We cover that mechanism in detail in our guide to sequence of returns risk.
Real markets are brutal about this. The S&P 500 returned roughly -37% in 2008 including dividends, and roughly -18% in 2022. Between the 2000 peak and the 2002 low it fell close to half. A straight-line forecast contains none of that. A Monte Carlo run does, because it samples bad years at the frequency your inputs imply.
The second benefit is communication. A statement such as "your plan succeeds in 82% of simulated paths" gives you a number you can act on. You can raise contributions, delay retirement or trim withdrawals and watch the probability move. That feedback loop is worth more than any single projection.
Finally, simulation sets honest expectations. Metrics like maximum drawdown describe the past you actually lived through. Monte Carlo describes the many pasts that could have happened, which is a wider and more useful view of risk.
The 1998 Trinity study defined success as a portfolio that lasted the full 30-year payout period (Wikipedia, captured 11 October 2026).
How to Run a Monte Carlo Simulation Step by Step
You can run a Monte Carlo simulation with a free web calculator, a spreadsheet or a short script. The steps are the same in every case, and the quality of your inputs matters far more than the tool.
Step 1: Define the question and the failure condition
Decide exactly what you are testing. A common question is "will my portfolio last 30 years of withdrawals?" The 1998 Trinity study used that framing and called a plan a failure if the money ran out before the period ended. Write your own definition of success before you touch a tool, such as "ending balance above zero at age 95" or "portfolio above $1 million in 20 years".
Step 2: Choose realistic return and volatility inputs
You need an expected annual return and a standard deviation. Broad global equities have historically delivered a nominal return of roughly 8% to 10% with volatility near 15% to 18%. Government bonds deliver less return and far less volatility. Blend them by your allocation. Be conservative: using a return one point lower than the historical average is a cheap way to build in humility.
Step 3: Add cash flows and inflation
Include annual contributions while you save and inflation-adjusted withdrawals while you spend. Model inflation as its own variable, around 2% to 3%, because withdrawals that rise with prices are what sink many plans. Keep everything in real, inflation-adjusted terms so the results are easy to interpret.
Step 4: Run at least 5,000 trials
With only 100 trials the result swings noticeably from run to run. Around 10,000 trials is a sensible default, since the success probability then stabilizes within a fraction of a percentage point. More trials cost almost nothing on a modern computer.
Step 5: Read the distribution, not just the headline
Do not stop at the success rate. Look at the 10th, 50th and 90th percentile outcomes. The 10th percentile path is your bad-luck scenario, and the question to ask is whether you could live with it. Our guide to value at risk covers a closely related way to think about tail outcomes.
Rule of thumb: many planners treat 75% to 90% as a healthy success range, because 100% usually means you are saving or underspending more than you need to.
Real Examples: Same Average Return, Very Different Outcomes
Consider an investor with a $1,000,000 portfolio who withdraws $40,000 a year, adjusted for inflation, over 30 years. That is the classic 4% rule that came out of Bengen's 1994 research and the Trinity study. A fixed-return projection at 6% says this plan is comfortable. A Monte Carlo run says something more nuanced, because it tests the same plan against thousands of return sequences.
The table below shows how the headline result changes as you change one lever at a time. The percentages are illustrative of the pattern you will see in most tools, not a forecast, so run your own inputs before drawing conclusions.
The lesson is that the withdrawal rate moves the result far more than small changes in expected return. A quick way to improve a shaky plan is to cut spending by a few hundred dollars a month, not to chase a higher-returning asset.
Free tools make this easy to test. The calculator shown on the cover of this article lets you change savings, contributions and horizon and watch the success probability update. Try raising the contribution by $5,000 a year and see how many percentage points it buys.
Hertz brought Monte Carlo methods to corporate finance in 1964 and Boyle applied them to derivatives in 1977 (Wikipedia, captured 11 October 2026).
Common Monte Carlo Mistakes to Avoid
A simulation is only as honest as its assumptions. These are the errors that quietly produce a comforting but wrong answer.
Mistake 1: Treating the output as a prediction
A success probability of 85% does not mean you have an 85% chance of retiring comfortably in the real world. It means that under the model's assumptions, 85% of the simulated paths worked. The model can be wrong about the world, and that risk sits outside the number.
Mistake 2: Using normal distributions that underweight crashes
Many simple tools draw returns from a bell curve. Real markets have fat tails, meaning extreme months occur more often than a bell curve allows. Look for tools that offer historical bootstrapping, which resamples actual past returns, or fat-tailed distributions. If yours does not, treat its crash risk as understated.
Mistake 3: Ignoring fees and taxes
A fund charging 1% a year takes about a fifth of a 5% real return before you see it. Enter returns net of fees, and think about the tax drag on withdrawals in your country. Leaving these out inflates success rates in a way that compounds over 30 years.
Mistake 4: Assuming you will never adjust
Real people respond to a crash. They trim spending, delay a purchase or work another year. A rigid model that withdraws the same amount through a 40% drawdown understates how resilient a flexible plan is. Test a version with a spending rule such as "cut withdrawals by 10% after a down year" and compare.
Mistake 5: Chasing 100%
A plan that succeeds in every trial usually means you are working or saving more than you need to. Underspending for decades is a real cost. Aim for a probability you can defend, then build a plan B for the remaining slice. Matching the plan to your own risk tolerance matters as much as the number.
What to Watch Next
Simulation tools improve quickly, and the inputs they depend on keep shifting. These are the checkpoints worth tracking over the next year.
- > Does your tool's expected return still match current valuations, or is it using a long-run average that may be too generous?
- > Are inflation readings in your country running above the 2% to 3% your model assumes?
- > Does your success probability fall below 75% after a market drop, which would be a signal to revisit spending?
- > Has your own situation changed, such as a new dependent, a job change or a retirement date moved earlier?
Frequently Asked Questions
How accurate is a Monte Carlo simulation?
It is accurate about the math and uncertain about the world. With enough trials the result is a precise answer to the question "given these assumptions, how often does the plan work?" Whether those assumptions match the future is something no simulation can promise.
How many trials should a Monte Carlo simulation run?
Around 10,000 is the common default. Below 1,000 trials the result can swing by several percentage points between runs. Beyond 10,000 the gain in stability is tiny.
What is a good Monte Carlo success rate for retirement?
Many planners aim for 75% to 90%. Lower than that and the plan is fragile. Much higher can mean you are being overly cautious with your spending. Choose a level that fits how flexible you can be if markets disappoint.
Can I run a Monte Carlo simulation in a spreadsheet?
Yes. Generate random returns with a normal-distribution formula, chain them across your horizon, and copy the row down several thousand times. Count the rows that end above zero. It takes about an hour to build and teaches you more than any black-box tool.
Is Monte Carlo better than a historical backtest?
They answer different questions, so use both. A backtest replays the one history that actually occurred. Monte Carlo explores many alternatives, including sequences that never happened. Pair it with risk measures such as the Sharpe ratio for a fuller picture.
Key Takeaways
- Monte Carlo simulation runs your plan through thousands of random market paths and reports how often it succeeds.
- The order of returns matters as much as the average, which is why a single straight-line projection misleads.
- Use about 10,000 trials and read the 10th, 50th and 90th percentile outcomes, not just the headline probability.
- Withdrawal rate and spending flexibility move the result more than small changes in expected return.
- Net out fees and taxes, add fat-tailed returns, and test a spending rule for bad years.
- Treat the result like a weather forecast: a probability to plan around, never a promise. Review it once a year.
This article is educational and is not personal financial advice. Consider speaking with a licensed adviser in your country before making decisions.