Monte Carlo in 30 seconds: the dice metaphor
Roll a die 10,000 times: you discover the real probability of each face. Monte Carlo applies that logic to your FIRE plan. Each roll is a full 30-year market scenario. After 10,000 rolls, you have your success rate.
Why Monte Carlo crushes the flat 7% projection
No market returns 7% every year. It returns +30% one year, β20% the next, sometimes β40% two years in a row. A smooth projection tells you 'FIRE at 52.' Monte Carlo tells you 'FIRE at 52 in 88% of cases, at 58 in the worst 10%.' Those two sentences do not lead to the same decision.
How Monte Carlo tests your plan, step by step
- Draw a sequence of annual returns, calibrated on the historical averages and volatility of your allocation.
- Compute your wealth year by year for that sequence.
- Repeat 1,000 to 10,000 times, with a different sequence each round.
- Read the results: success rate, P10 (the worst 10% of draws), P50 (median), P90 (the best 10%).
The 4 traps that distort your Monte Carlo reading
- Confusing success with a guarantee. 95% success means 1 plan in 20 ends at zero. Not a certainty.
- Reading the success rate without looking at the P10. A 92% success rate with a P10 of β¬200,000 at the end of retirement is still aggressive if your non-negotiable expenses are β¬25,000 a year.
- Plugging in overly optimistic assumptions. 9% real returns, 1.5% inflation: you manufacture a flattering success rate on numbers no one has held for 30 years.
- Not rerunning after a life change. A child, a move, a rent hike: your expenses shift by 20%, and so does your success rate.
Key Takeaways
- 1Monte Carlo is 1,000 to 10,000 simulated markets, not a prediction. You get a probability, not a future.
- 290% success = your plan holds in 9 draws out of
- 3Below 80%, rethink; between 80 and 90%, keep an eye on it.
- 4P10, P50, P90, to read together. The success rate alone hides the worst trajectories.
- 51,000 simulations to iterate, 10,000 for the final decision. Beyond that, the gain becomes marginal.
Keep going
Frequently asked questions
Monte Carlo simulates thousands of possible market futures by randomly generating returns year by year according to a statistical distribution. For FIRE, you run 1,000 to 10,000 simulations and then compute the percentage of scenarios in which the portfolio survives to the end. That figure is the 'success rate'.
That in 9 out of 10 simulations (based on the assumptions provided), your portfolio is not depleted by the end of the modeled period. It is a probability conditional on the assumptions (mean return, volatility, inflation), not a guarantee. Above 90% = robust; 80 to 90% = watch; below 80% = reconsider.
P10 = pessimistic scenario (10% of simulations are worse). P50 = median (50/50). P90 = optimistic (10% are better). Analyzing P10 is crucial: if your P10 shows depleted capital at 80 while you plan to live to 90, the plan is risky despite a flattering success rate.
100 = unreliable, noisy results. 1,000 = good speed/precision tradeoff for interactive exploration. 10,000 = maximum precision for the final FIRE decision. Beyond, the gain becomes marginal. Our simulator runs 10,000 by default for Architect accounts.