Given everything you know about your savings, your spending, and how long you expect to live, what are the odds your money lasts in retirement? That is the question that a retirement Monte Carlo analysis answers. Instead of answering what is "my number"? It trades that single, tidy number, which a basic calculator gives you, for a range of outcomes with a probability attached.
This article covers what a Monte Carlo analysis of a retirement plan actually is, how it works, and why it usually gives you a different answer than the straight-line calculator on your bank's website. By the end you will know what to put into one and how to read what comes back.
Start with the alternative, because the contrast is the whole point. A conventional retirement calculator uses one fixed rate of return. You enter your balance, your yearly withdrawal, and an assumed 6% or 7%, and it grows the money in a straight line to the end of your plan. The answer is a single ending balance. It is clean, and it is almost always wrong, because no real portfolio earns exactly 7% every year for thirty years.
A Monte Carlo analysis drops the single rate. Instead it runs your plan hundreds or thousands of times, and each run draws a different randomized sequence of yearly returns and inflation from the statistical behavior of markets. One run might open with two strong years and a crash in year eight. The next might open with the crash. Every run ends by checking the same thing: did the balance reach the age you set, or did it hit zero first.
The output is the share of runs in which the money lasted. Run your plan 1,000 times, watch it survive 870 of them, and the analysis reports an 87% success rate. That figure is a probability, not a forecast of what your retirement will do.
The core of it: A Monte Carlo analysis runs your plan against many randomized return sequences and reports the percentage in which your money outlasts you. It measures how much margin your plan has against bad luck, not what will happen.
The clearest way to see what the method adds is to give the same person to both tools. Consider a retiree, call her A. She has $1 million, plans to withdraw $60,000 a year adjusted for inflation, and wants the money to last 30 years. A straight-line calculator set to 7% growth and 3% inflation shows her portfolio never running dry. It even keeps growing for the first decade. The verdict is a comfortable yes.
Feed the identical numbers into a Monte Carlo analysis, keep the 7% average but add the year-to-year swings real markets have, and the answer changes. Across 1,000 randomized sequences, A's plan survives the full 30 years in about 78 of every 100. The straight-line "yes" was really a 78% yes. The 22% it hid are the sequences where weak returns arrive early, and that is the case a single average return cannot show.
| Straight-line calculator | Monte Carlo analysis | |
|---|---|---|
| Return assumption | One fixed rate, every year | A different randomized sequence each run |
| Output | A single ending balance | A success rate across many runs |
| A bad early decade | Averaged away, never shown | Shown; some runs put the bad years first |
| Best used for | A rough first estimate | Stress-testing whether a plan holds up |
The reason for the gap is sequence risk. Once you are withdrawing, the order of returns matters and not just the average, because money you pull out of a falling market is gone before the recovery. Our piece on sequence of returns risk shows the same set of returns in two different orders producing opposite outcomes, one retiree broke and the other with more than a million left. A straight-line calculator is blind to that by construction. A Monte Carlo analysis is built to expose it.
A monte carlo simulation for retirement planning is only as trustworthy as what you feed it. Most of the inputs are the same ones any retirement plan needs: your current balances by account, how much you still contribute, your retirement age, your annual spending in retirement, any Social Security or pension income, and your assumptions for returns and inflation. Two inputs, though, move the result more than people expect.
This is the age you tell the analysis to run to. Set it too low and every plan looks safe, because you are asking the money to cover fewer years. Push it from 85 to 95 and a plan that looked solid can slip, since it now has to fund ten more years of withdrawals through whatever the market does. Longevity is a guess, so it is worth running the analysis at more than one age to see how sensitive your result is.
The average return gets all the attention, but the size of the swings around it drives the failures. A 7% average with wide year-to-year volatility fails more often than a 7% average with narrow volatility, even though the two averages are identical, because the wild version produces more of the deep early drops that sink a plan. An honest analysis asks for both numbers, not just the average.
The number of runs is a precision setting, not a quality setting. Below a few hundred runs the success figure still jumps by several points each time you press go, because the sample is too small to settle. Around 1,000 runs it lands to within a point or two. Running tens of thousands is overkill. It gets more precise, but the extra digits sit inside the margin of error, and the analysis just takes longer for no useful gain. RetirFi runs on the order of 1,000 paths for that reason. Our Monte Carlo explainer covers the mechanics of how each path is generated.
A single success rate is easy to misread. It looks like a test score, so 85 reads as good-not-great and the missing 15 reads as a warning, and neither is right. Higher is not automatically better either. A plan that scores 99% is often a sign you are underspending your own retirement, holding back travel and time you will never get back to clear a bar you would almost never hit. Most planners treat a rate in the 80s as the healthy target: strong enough to weather ordinary bad markets, honest enough that it assumes you will make small course corrections along the way. Our guide on how to interpret your Monte Carlo results works through the full range and what each band signals.
The method has real limits, and knowing them keeps you from trusting the number more than it deserves.
The first is the oldest rule in modeling: wrong inputs produce a confident wrong answer. An optimistic return assumption or a lowball spending figure can turn a shaky plan into a reassuring 90%, and the analysis will not warn you, because it has no way to know your inputs were off.
The second is that markets are not perfectly random or perfectly well-behaved. Most analyses draw returns from a statistical distribution built on history, and real crashes cluster and run deeper than that history sometimes suggests. The tails are fatter than the model assumes, which means the genuinely bad outcomes may be a little more likely than the score implies.
The third is that the analysis models the arithmetic, not your behavior. In a real downturn you would trim spending, delay a big purchase, or pick up some work, and a standard run assumes you keep withdrawing on autopilot straight into the crash. That makes the raw success rate more pessimistic than a flexible retiree's real odds, which is one more reason the number is a guide to steer by rather than a verdict.
A retirement Monte Carlo analysis will not tell you what your retirement will look like. It tells you how well your plan holds up when markets misbehave, and it does that by testing the one thing a straight-line calculator quietly ignores: the order in which returns show up. Treat the result as a range you keep an eye on, revisit it once a year and after any big change, and adjust the inputs as your life and the markets move. Run your own numbers and see where your plan lands.
Enter your numbers and run a Monte Carlo simulation to see how this plays out for your specific timeline and assets.
Open the Calculator →