Monte Carlo simulation for retirement — what it is and how to use it

Many simple retirement calculators assume your portfolio earns a steady 7% every year. Markets don't work that way. The S&P 500 returned +31.5% in 2024 and −18.1% in 2022. Monte Carlo simulation solves this by running thousands of randomized scenarios — each with a different sequence of market returns — to show the probability your plan survives, not just whether it works under optimistic assumptions. An 85% success rate means your plan held up in 850 of 1,000 simulations. Here's what that actually means, and how to use it.

What Monte Carlo simulation is

Monte Carlo simulation is a statistical technique that estimates the range of possible outcomes for an uncertain event by running thousands of randomized trials. The name comes from the Monte Carlo casino — named for its reliance on randomness. The technique was formalized in the 1940s by mathematicians using random sampling as a computational tool for complex probability problems.

In retirement planning, the "uncertain event" is whether your portfolio lasts as long as you need it to. The simulation replaces a single assumed return rate with thousands of randomized return sequences drawn from historical distributions — capturing the volatility, crash years, and boom years that real markets produce. After running those thousands of scenarios, it reports what percentage succeeded — i.e., your portfolio still had money at the end of your retirement horizon.

The core insight it provides that fixed-rate projections can't: average returns don't determine retirement outcomes — the sequence of those returns does. A simulation that includes crash scenarios in early retirement years reveals the true fragility of a plan in a way that a steady 7% assumption never could. This is Monte Carlo's primary value: it stress-tests a plan against the scenarios that actually hurt retirees, not just the scenarios that look good on paper.

How it works — the mechanics

The process follows a consistent pattern across tools, even if the underlying methodology varies:

1. You provide inputs. Current portfolio value, planned annual withdrawal, expected retirement length, asset allocation (stock/bond mix), and optionally guaranteed income sources like Social Security.
2. The simulator generates random return sequences. Each simulation run draws annual returns from a probability distribution based on historical market data — typically using historical mean returns and standard deviations for each asset class. "Random" here doesn't mean arbitrary: returns are sampled from a distribution calibrated to historical data or forward-looking assumptions, so the randomness mirrors how real markets behave over time. Each simulation uses a different sequence of good and bad years.
3. Each simulation runs to completion. The model tracks the portfolio year by year, subtracting withdrawals and adding or subtracting market returns. If the portfolio reaches zero before the end of the retirement horizon, that simulation is a failure. If money remains, it's a success.
4. The results are aggregated. After 1,000 or more simulations, the tool reports the success rate (percentage that survived), median ending balance, and often percentile ranges (10th, 50th, 90th percentile outcomes).

How to interpret success rates

The success rate is the percentage of simulated scenarios in which the portfolio survives the full retirement period. An 85% success rate means 850 of 1,000 simulations had money remaining at the end. This is the number most financial planners focus on — but it's frequently misunderstood.

Monte Carlo vs fixed-rate projections

Fixed-rate projections aren't useless — they're perfectly adequate for accumulation-phase planning, where sequence risk is minimal. For retirement income planning, where sequence risk is the central risk, Monte Carlo simulation provides meaningfully better insight. Most serious retirement planners use both: fixed-rate projections to understand the general picture, Monte Carlo to stress-test it.

The inputs that matter most

Monte Carlo results are only as good as the assumptions behind them. The inputs that have the largest impact on your success rate:

• Withdrawal rate. The single most important variable. The difference between 3.5% and 4.5% withdrawal rate can shift success probability by 15–20 percentage points. This is worth optimizing before any other input.
• Retirement length. A 30-year retirement is significantly harder to fund than a 25-year one. Each additional year of retirement reduces success probability. This is why retiring at 60 vs 65 is a much bigger financial decision than it appears.
• Asset allocation. A heavier equity allocation produces higher expected returns but higher volatility — which means worse outcomes in bad sequences but better outcomes in good ones. A 60/40 portfolio often shows higher Monte Carlo success rates than 100% stocks in many simulations of 30-year retirements, despite lower average returns, because volatility reduction can matter as much as average return in withdrawal scenarios — though results vary depending on return assumptions and horizon.
• Guaranteed income sources. Social Security, pension income, or annuity payments reduce the amount the portfolio must fund. Each dollar of guaranteed income significantly reduces sequence risk exposure. Including Social Security in the simulation often raises success rates by 10–20 percentage points for many retirees.
• Return distribution assumptions. Different tools use different assumptions about expected stock and bond returns. A tool using historical US equity returns (~10% nominal) produces different results than one using Morningstar's forward-looking capital market assumptions (~7–8% nominal). Recent forward-looking capital market estimates (including from Morningstar) suggest lower expected equity returns than long-run historical averages — which would reduce simulated success rates compared to historically-calibrated tools. This is worth checking when comparing results across platforms.

Limitations — what Monte Carlo can't tell you

Monte Carlo simulation is a powerful tool, but it has real limitations that are worth understanding before treating results as definitive:

• It depends entirely on its input assumptions. The return distributions used are typically based on historical data. If future market conditions differ materially from historical patterns — as they might in a period of structurally lower returns — the simulation's calibration will be off.
• It typically models constant spending. Real retirement spending isn't constant. Most retirees spend more in early retirement (travel, hobbies) and less in middle retirement, with potential increases in late retirement for healthcare. A simulation using flat inflation-adjusted withdrawals may understate early spending needs and overstate late ones.
• It doesn't model major life events. Long-term care costs, divorce, a financially distressed adult child needing support — none of these show up in a standard simulation. These can have a larger impact on retirement outcomes than market returns in some cases.
• It treats failure as binary. The "failure" flag in a simulation simply means the portfolio reached zero. It doesn't distinguish between running out of money in year 25 (with years left) vs year 29 (with months left). The severity of failure varies enormously.
• High success rates don't mean high certainty. A 90% success rate is a modeled probability based on simulated return sequences — not a guarantee. Real markets may behave in ways that no historical simulation anticipated. The 2020 pandemic, 2022 inflation surge, and 2008 financial crisis all contained elements outside normal historical distributions.

Where to run a free Monte Carlo simulation

How to use Monte Carlo simulation in practice

Running a simulation once gives you a snapshot. The real value comes from using it iteratively — adjusting inputs and watching how the success rate responds. A practical workflow:

1. Run your baseline plan. Enter your current portfolio, planned withdrawal, and expected retirement length. Record the success rate.
2. Adjust your withdrawal rate. Try reducing your annual withdrawal by 5–10% and re-run. If success rate jumps significantly, you're in a sensitive zone where small spending changes matter a lot.
3. Test retiring earlier vs later. Retiring at 62 vs 65 adds 3 years of retirement and 3 fewer years of contributions — this combination often reduces success rates by 10–15 percentage points. See the real cost of early retirement in your specific numbers.
4. Add Social Security. If you haven't included Social Security as guaranteed income, add it and re-run. For most middle-income retirees, this significantly improves success rates — often by 10–20 points.
5. Test delaying Social Security. Compare claiming at 62 vs 67 vs 70. Delaying increases the guaranteed income floor and often meaningfully improves the simulation outcome.
6. Observe the sensitivity. Which input change moves your success rate the most? That's where to focus planning attention. For most people, it's the withdrawal rate and retirement age — not investment returns.

Frequently asked questions

Is an 85% Monte Carlo success rate good enough?

For most financial planners, 85–90% is the target range. It reflects a plan with high probability of success while still allowing reasonable retirement spending rather than over-saving. The right target depends on your flexibility — if you have guaranteed income (Social Security, pension) covering basic expenses and can reduce discretionary spending when needed, 80% may be adequate. If your entire budget depends on the portfolio and you have no flexibility, 90%+ is more appropriate.

Why would a 100% success rate be a problem?

A 100% success rate typically means the plan is so conservative that in every simulated scenario, you end retirement with a large remaining portfolio — money you never got to enjoy. It suggests you're either withdrawing too little, over-saving, or both. Unless leaving a large inheritance is a specific goal, a 100% success rate may mean you sacrificed quality of life in retirement unnecessarily. T. Rowe Price and other planners explicitly note that very high success rates signal excessive conservatism.

Does Monte Carlo simulation account for Social Security?

Most tools allow you to include Social Security as a guaranteed income source that reduces the amount your portfolio must fund. Including Social Security typically increases success rates significantly — a retiree whose Social Security covers 60% of spending needs a portfolio that's much more resilient than one funding 100% from investments. Always include guaranteed income sources when running simulations, as excluding them produces an unnecessarily pessimistic result.

How does Monte Carlo relate to the 4% rule?

The 4% rule was derived from historical return sequence analysis — Bengen tested whether $40,000/year withdrawal on a $1 million portfolio survived every 30-year historical period. Monte Carlo simulation is a related but different approach: instead of testing against all historical periods, it generates synthetic return sequences from a probability distribution. The two approaches often produce similar conclusions, but Monte Carlo can be calibrated to forward-looking return assumptions and tested against non-historical scenarios. Our 4% Rule guide covers the relationship in more detail.