← All articles
Method

Monte Carlo simulation explained simply: why a fixed return leads you astray

A projection with a 5% annual return says: your money will easily last. A simulation with the same assumptions says: in almost one case in three, it runs out early. Both calculate correctly. Only one shows you what really matters.

Christoph Urban · · 7 min read

The problem with the average return

Most calculators on the internet and many advisory meetings work with a fixed return. You enter an amount, choose 5 or 6% a year and see a smooth curve that rises steadily. It looks precise. But it is an assumption that will never come true.

Stock markets do not deliver 5% a year. They deliver +25% one year, then −18%, then +9%, then +31%. Over the long run, the average may come out at a similar figure. But what matters for your wealth is the order in which these years arrive, especially when you are withdrawing money. The keyword: sequence-of-returns risk.

A worked example

Let’s assume you retire with €500,000 and withdraw €20,000 in the first year, then 2% more every year to offset inflation. The portfolio consists of 80% broadly diversified equity ETFs and 20% money market.

The classic projection with a fixed 5% a year shows: after 30 years, around €487,000 is left. The wealth remains almost untouched. All good.

The Monte Carlo simulation with realistic fluctuations, taxes and the same withdrawals paints a different picture:

Result after 30 yearsValue
Money lasts 30 yearsin 72 of 100 scenarios
Wealth in the middle scenario (median)around €412,000
Wealth in the worst 10% of scenarios€0, the money runs out early

The middle scenario looks similar to the projection. But in a good quarter of cases, the money does not last. These are exactly the cases a fixed return hides. And they are exactly the ones you want to protect yourself against.

How a Monte Carlo simulation works

The name comes from the casino in Monaco, because the method works with chance. Instead of a single future, the simulation calculates thousands of possible futures:

  1. Set assumptions: expected return and volatility for each asset class, inflation, taxes, withdrawals, pension and other inflows.
  2. Roll the dice for a year: For each year, the simulation draws a random return that fits the assumptions. Sometimes +20%, sometimes −15%, mostly something in between.
  3. Calculate one life path: Year by year, returns, taxes, withdrawals and inflows are booked, up to the end of the planning period.
  4. Repeat thousands of times: The whole thing runs 10,000 or 20,000 times, each time with a different sequence of good and bad years.
  5. Evaluate: At the end, the simulation counts in how many runs the money lasted and shows the range of outcomes.

Reading the results correctly

Three figures are enough to understand a simulation:

  • Success rate: In how many scenarios does the money last until the end? Many planners aim for 85–95%. 100% usually means you could spend considerably more.
  • Median: the middle result. Half of the scenarios end better, half worse. It is the realistic expected value, but not a promise.
  • 10th percentile: the result that only the worst 10% of scenarios fall below. This is your stress test: could you cope with this too?

In a chart, this looks like a fan. In the middle is the median line, around it the band from the 10th to the 90th percentile. The wider the fan, the more uncertain the future.

Fan chart: the median in the middle, the band shows the 10th to 90th percentile. Illustration.

What the simulation cannot do

A simulation is only as good as its assumptions. You should know three limits:

  • Extreme events: Most models assume normally distributed returns. Real crashes are more severe and more frequent. A sound plan therefore works with cautious returns or additionally tests individual crisis scenarios.
  • Assumptions about the future: No one knows the stock market return of the next 30 years. The simulation shows what can happen under the assumptions, not what will happen.
  • Simplifications: Taxes, costs and expenses are applied as flat rates. In real life they are more irregular.

This does not speak against the method. It speaks for reading the results as a range and recalculating the plan regularly.

Conclusion: plan for the range, not for the average

A fixed return answers the question: “What happens if everything goes to plan?” A Monte Carlo simulation answers the more important question: “What happens if it doesn’t?” Anyone building wealth or wanting to live off it should know both answers. I show what this means for the well-known rule of thumb in The 4% rule recalculated for Germany.

Try it yourselfWill my money last? The free calculatorCalculate now →
Simulation assumptions
  • Starting capital €500,000, first withdrawal €20,000, then +2% per year
  • 80% global equity ETFs: expected return 7% p.a., volatility 16% p.a.; 20% money market: 2.5% p.a.
  • Tax: 26.375% flat capital gains tax incl. solidarity surcharge, 30% partial exemption, only on the gain portion of units sold
  • 20,000 simulations, annual steps, normally distributed returns
  • Comparison projection: fixed 5% p.a., without taxes
Christoph UrbanI am an independent financial planner in Berlin, focusing on retirement planning. I bought my first share at the age of twelve, paid a lot for my lessons during the dot-com era and have been studying evidence-based investing for 15 years. Today I help my clients plan how to build wealth and provide for retirement, without selling any products. Before that, I spent more than 25 years in marketing, including at MTV, Fox, Deezer and Kleinanzeigen.

General information, not investment or tax advice. Simulations are based on assumptions; past performance is not a reliable indicator of future results.