You have run the numbers, the line on the chart never touches zero, and you are wondering whether that line is the whole story. It is not. The same average return can arrive in a kind order or a cruel one, and when you are already living off the portfolio, the order decides more than the average. That is sequence of returns risk.
Most explanations stop at the definition and one illustrative table. This one runs a single early-retirement plan through 500 simulated markets on the free forecast calculator, which uses the same projection engine as the Krosos app, and then changes one thing at a time to see what actually moves the odds.
The short version:
- A straight line hides the risk completely. The household below never runs out on the smooth projection and ends with about €4.8 million in today's money. Across 500 simulated markets with the same typical growth, the money runs out in 29% of them.
- In Europe, the state pension is the biggest lever. Remove it and survival falls from 71% to 44%. Sequence risk here is mostly a problem of the bridge years before the pension starts.
- Spending less and working longer both help, by similar amounts. €500 a month less takes survival to 88%. Three more years of work takes it to 83%.
- A cash buffer that you simply spend first does nothing. Three years of spending moved into a money-market fund left survival at 71%. A buffer only helps if you follow a rule about when to use it.
Why the order matters, in three years
A worked example, by hand, so you can check every step. Start with €900,000, take out €36,000 at the start of each year, and suffer one bad year (minus 20%) and two good ones (plus 10% each).
| Bad year first | Bad year last | |
|---|---|---|
| After year 1 | €691,200 | €950,400 |
| After year 2 | €720,720 | €1,005,840 |
| After year 3 | €753,192 | €775,872 |
Same three returns, same withdrawals, €22,680 apart after three years. Without the withdrawals, both orders would end at exactly €871,200, because multiplication does not care about order. The withdrawals are what break the symmetry: in a bad year you sell more units at low prices, and those units are not there for the recovery.
Three years is a toy. The question that matters is what this does over a whole retirement, and that needs many markets, not two.
The household
One worked example, labelled as such. Open this scenario, then press Range above the chart to see the 500 markets.
| Age | 45 (born 1981) |
| Stops working | November 2026 |
| Cash | €40,000 |
| Stocks and funds | €860,000 |
| Retirement spending | €3,000 a month, in today's money |
| State pension | €1,800 a month in today's money, from January 2048 (around 67) |
That is €36,000 a year from €900,000, a 4% first-year withdrawal, until the pension arrives and covers most of the bill. The assumptions worth stating: stocks grow 6% a year against 2% inflation (the calculator's default is 7%; we lowered it), with the default 15% yearly volatility. Taxes use the calculator's default settings, which follow Belgium's new capital-gains rules: 10% on realised gains above a €10,000 yearly exemption, and 30% on interest. The Belgian tax explainer has the detail.
Press Range and the calculator runs 500 markets. Each one is built so that its typical month compounds at the same 6% as the straight line, with the ups and downs drawn at random around it. The random draws are fixed, so you will see exactly the figures below, and so will anyone you send the link to.
Five runs
Each run changes one input from the first, named in the left column.
| Run | Straight line | Survives in 500 markets |
|---|---|---|
| The plan as described | Never runs out, ends at €4.8M | 71% |
| No state pension (pension income set to 0) | Never runs out, shrinks to €383K | 44% |
| Spend €2,500 a month instead of €3,000 | Never runs out, ends at €7.8M | 88% |
| Stop in November 2029 instead (three more years on the default €4,200 take-home) | Never runs out, ends at €7.8M | 83% |
| €108,000 of the stocks moved to a money-market fund (three years of spending, at the default 3% growth) | Never runs out, ends at €4.6M | 71% |
Ending values are in today's money, in October 2101, where the calculator's 75-year window ends.
What the five runs say
The straight line answers a different question. Every row in the middle column says the money lasts. That is the projection most calculators show, and it is honest about the average. It is silent about the order, and the order is what turns 29% of these markets into failures. If a single line is all you have looked at, you have looked at the best-behaved version of your plan.
The pension is doing more work than it looks. It does not start for 21 years, and it still lifts survival from 44% to 71%. The reason is the shape of the risk. Bad years hurt most while the whole bill is coming out of the portfolio. Once €1,800 of the €3,000 arrives from elsewhere, the portfolio only has to find €1,200 a month, and a bad decade at that point does far less damage. For a European early retiree, sequence risk is concentrated in the bridge between stopping work and the pension, which is the same gap retiring before pension age looks at from the other side. Before you worry about the order of returns, make sure the pension figure you entered is the one you will actually get.
Spending and working are the two levers that work, and they are close. Spending €500 a month less, so a 3.3% first-year withdrawal instead of 4%, lifts survival to 88%. Working three more years lifts it to 83%: the portfolio grows untouched for three years, takes in three years of savings, and has three fewer years to fund before the pension. Which one costs you less is a personal question. The calculator can only tell you they buy roughly the same thing.
The cash buffer did nothing, and that is worth understanding. A common defence against sequence risk is to hold a few years of spending in something stable, so you never sell stocks in a crash. We moved exactly three years of spending, €108,000, out of stocks and into a money-market fund. Survival stayed at 71%, and the straight line ended slightly lower. The reason is how the calculator spends: it draws on cash first, then the money-market fund, then stocks, every month, whatever the market is doing. So the buffer is used up in the first three years whether those years are good or bad, and it protects nothing. A buffer that helps is a buffer with a rule: spend from it only after a fall, refill it after a rise. The calculator does not model that rule, so it cannot tell you what such a buffer is worth, and we would rather say so than imply it.
Doing this with your own numbers
- Press Range. It is off by default, and the single line is the flattering view. The survival rate appears in the chart legend.
- Enter the state pension, with its real start date. As the second run shows, leaving it out understates the plan badly, and entering it too early overstates it.
- Lower the growth rate if you think 7% is generous. We used 6% and said so. Survival is sensitive to it, which is a good reason to try a few values rather than trust one.
- Zero the asset classes you do not hold. The calculator starts with sample balances in cash, money-market funds and other classes, and inheriting them quietly changes the plan.
- Change one thing at a time. That is the only way to know which change bought the improvement.
How the calculator works under the hood, including what Range computes, is in the forecast playground explainer. The withdrawal rates themselves, and what European taxes do to them, are in the 4% rule in Europe.
What this leaves out
- The window is long. The calculator runs every market to October 2101, when this household would be 120. A market where the money runs out at 105 counts as a failure, so read the survival rate as a strict test rather than a forecast of your lifetime.
- Markets here have no memory. Each month's return is drawn independently around the same average. Real markets have crashes that cluster and recoveries that follow them, and asset classes that fall together. 500 random markets bound the uncertainty; they do not describe it exactly.
- Spending never adapts. Every run spends €3,000 a month in today's money through every crash. Real people cut back after a bad year, and a flexible spending rule is one of the strongest defences there is. The calculator does not model one, so these survival rates are, if anything, on the cautious side for anyone willing to adjust.
- One household, one set of tax rules. The figures come from Belgian-style default tax settings and an all-stocks portfolio. Your own mix, country and pension will give different numbers, which is the reason every scenario above is a link you can edit.
These are estimates from stated assumptions, not financial advice. The useful output is not the 71% itself but the comparison between runs: which change moved it, and by how much.