How Monte Carlo Simulation Works (and Its Limits)
Monte Carlo tests thousands of possible return paths for your retirement plan, so you can judge probability — not prediction. Here's how to read the results.
What Monte Carlo actually does
A deterministic calculator runs one path: “if your portfolio returns X% each year, here’s what happens.” Monte Carlo runs thousands of different paths. For each path, it samples a possible return sequence (often based on historical returns, or a statistical model of those returns), then applies your spending, contributions, inflation and fees year by year. The output is the percentage of paths that don’t run out of money by the end of your time horizon — sometimes called “success rate” or “probability of success.”
This doesn’t tell you what will happen. It tells you what could happen, under the return distribution and rules you chose.
Why use it instead of a single projection?
The biggest weakness of a single line projection is that it ignores volatility and order. Two sequences can have the same average return, but very different outcomes. Monte Carlo captures:
- Volatility drag: losing 50% then gaining 50% doesn’t bring you back to par — you need ~100% to recover.
- Sequence of returns risk: poor returns in the first few withdrawal years do disproportionate damage (more on this in the related guide).
- Uncertainty: markets don’t deliver a smooth geometric mean; they swing.
For retirement planning, probability is far more useful than pretending we can see the future.
How to read success rates
Treat success rates as a sanity check, not a pass/fail exam:
| Success rate | Interpretation | Suggested action |
|---|---|---|
| 95–99% | Very conservative under the model. May be over-saving for some goals. | Check if you’re sacrificing life now for tiny probability gain. |
| 85–95% | Reasonable for many plans, especially with flexibility. | The sweet spot for many retirees who can adjust spending. |
| 75–85% | Plausible but tight. | Look at buffers, part-time work, or spending flexibility. |
| < 75% | High risk of shortfall under these assumptions. | Revisit savings rate, spending, asset allocation or delay. |
There’s no single “right” number. Flexibility (cutting discretionary spending in a bad year, working part-time) means you can tolerate a lower modeled success rate. Rigidity (fixed spending, no ability to adjust) argues for higher.
What Monte Carlo can’t do
- It doesn’t predict regime shifts. Models based on historical periods won’t capture a regime never seen before.
- It can’t know your behaviour. Panic-selling, buying at peaks, or ignoring the plan breaks any model.
- It assumes discipline. Rebalancing rules, withdrawal rules and tax assumptions must be realistic.
- Garbage in, garbage out. Small changes to expected return, inflation, fees or longevity have huge effects.
- It can overfit history. The 1926–1976 period behaved differently from 1982–2019. Your model’s choice of period matters.
A worked example
Suppose you’re 65, spending ₹1,00,000/month (₹12 lakh/year), with ₹3 crore in a 60/40 portfolio, 30-year horizon, 2.5% inflation, 0.5% fees.
- Deterministic (mean): using long-run average returns might show “works fine.”
- Monte Carlo (variable): even with sensible assumptions, ~10–15% of paths run short — usually because of a poor sequence in years 1–5.
That nuance is exactly why Monte Carlo is useful: it forces you to plan for when markets disappoint, not if returns average out.
How this site uses it
Our Monte Carlo simulator:
- Generates thousands of paths based on the assumptions you set.
- Reports success rate, median remaining, worst-case paths and the timing of shortfalls.
- Stress-tests against common failure points (poor early sequence, high inflation).
- Keeps results educational and transparent about limitations.
Limitations & guardrails
- Model choice matters. “Historical resampling” and “parametric (lognormal)” give different tails. Compare if unsure.
- Don’t optimise for 100%. Chasing 100% success at age 30 often means extreme frugality for negligible practical gain.
- Build buffers. A 2–3 year cash bucket (see Bucket Strategy) is often more effective than an extra 5% equity exposure.
- Plan for flexibility. “Success” in reality often means “we adjusted slightly and got through.”
Bottom line
Monte Carlo is a map of plausible outcomes. It won’t tell you which road you’ll take — but it will show you dead-ends to avoid. Treat the probability, not the precision, as the signal.