Finacolada
Retirement & FIRE

Sequence of Returns Risk and Monte Carlo, Explained

Sequence of returns risk and Monte Carlo in plain English: why return order matters, how to read a fan chart and what 80%, 90% or 95% success means.

Updated · 8 min read

Two people retire on the same day with the same ₹1 crore, the same spending and the same investments. Over the next ten years, their portfolios earn exactly the same set of annual returns, just in a different order. Ten years later, one has more than ₹1.3 crore and the other has about ₹80 lakh.

That gap is sequence of returns risk, and it is the reason good retirement tools now use Monte Carlo simulation instead of a single “your money lasts 34 years” answer. This guide explains both in plain English.

Key takeaways

  • Sequence risk: losses early in retirement hurt far more than the same losses later, because you are withdrawing while prices are low.
  • Monte Carlo tests your plan against thousands of random return paths and tells you how often it survives.
  • A fan chart shows the range of outcomes over time; the lower edge matters most.
  • A 90% success rate means the money lasted in 90% of simulated futures, not that you are 90% “safe” in any absolute sense.
  • The output is only as good as its assumptions. Use it to compare choices, not to predict the future.

Why averages lie once you start withdrawing

When you are saving and not withdrawing, the order of returns does not change where you end up. ₹1 crore earning +20% then −10% ends at the same place as −10% then +20%. Multiplication does not care about order.

Withdrawals change that. Each year you remove money, and the amount you remove is fixed by your needs, not by the market. In a bad year, a fixed withdrawal is a bigger share of a smaller pot. You sell more of your investments at low prices, and those units are not around when the market recovers.

A worked example: same returns, opposite order

Assumptions: both retirees start with ₹1 crore; each withdraws ₹5 lakh in year one, rising 6% a year for inflation; withdrawals are taken at the start of each year. Retiree A gets the ten returns below in order; Retiree B gets the same ten returns in reverse. From year 11 onwards, both earn a steady 9% a year.

The ten annual returns are +28%, +20%, +15%, +12%, +10%, +8%, +6%, +2%, −5% and −12%. Their simple average is 8.4% and their compound annual growth rate is about 7.8%.

End of year Retiree A (good years first) Retiree B (bad years first)
1 ₹121.6 lakh ₹83.6 lakh
2 ₹139.6 lakh ₹74.4 lakh
3 ₹154.0 lakh ₹70.1 lakh
5 ₹175.5 lakh ₹66.7 lakh
7 ₹185.7 lakh ₹65.9 lakh
10 ₹137.9 lakh ₹80.1 lakh
Money lasts about 30 years about 20 years

Without any withdrawals, both would have ended year 10 with exactly the same amount, about ₹2.12 crore. With withdrawals, the same returns in a different order cost Retiree B roughly ten years of retirement income.

Look at Retiree B’s early years. The portfolio falls from ₹1 crore to about ₹66 lakh in five years, even though returns in years 3 to 5 are positive. The withdrawals are rising and the pot is too small to recover. By the time the good years arrive, there is much less money for them to work on.

Why a single projection is not enough

A typical retirement calculator assumes a steady return, say 9% every year. That hides sequence risk completely: every year is average, so there is no bad start. The result looks precise (“money lasts 33.5 years”) but tells you nothing about how wrong it could be.

Monte Carlo fixes this by asking a different question: across many plausible futures, how often does this plan work?

How a Monte Carlo simulation works

The name comes from the casino in Monaco, because the method relies on randomness. In plain steps:

  1. Choose assumptions for the average annual return and how much it varies from year to year (volatility), plus inflation.
  2. Generate one random future: a sequence of yearly returns, each drawn at random around the average. Some years are great, some terrible, in an unpredictable order.
  3. Run your plan through it: start with your corpus, take your withdrawals, apply each year’s return. Note whether the money lasts and how much is left each year.
  4. Repeat thousands of times, each with a fresh random sequence.
  5. Summarise: the share of futures where money lasted (the success probability) and the spread of balances over time (the fan chart).

How to read a fan chart

A fan chart plots your portfolio value over time, with bands for different percentiles. Here is what one looks like as numbers.

Assumptions: ₹1 crore corpus; first-year withdrawal ₹3.5 lakh (3.5%), rising 6% a year; 30-year horizon; 20,000 simulated paths; annual returns drawn independently from a normal distribution averaging 10% with a standard deviation of 14% (an illustrative, equity-heavy mix, not a forecast); inflation fixed at 6%. Values are nominal rupees.

Year 10th percentile 25th Median 75th 90th
5 ₹84.8 lakh ₹104.5 lakh ₹129.9 lakh ₹159.8 lakh ₹190.8 lakh
10 ₹83.8 lakh ₹117.4 lakh ₹166.2 lakh ₹229.4 lakh ₹302.5 lakh
20 ₹50.9 lakh ₹128.8 lakh ₹256.1 lakh ₹456.3 lakh ₹713.4 lakh
30 ₹0 ₹59.2 lakh ₹360.8 lakh ₹894.9 lakh ₹1,660.6 lakh

In this run, the money lasted the full 30 years in about 80% of paths.

What each part tells you

  • The median (middle column) is the “typical” outcome: half the paths did better, half worse. In this run it looks comfortable at ₹3.6 crore after 30 years. But in today’s money that median is only about ₹63 lakh, because of 30 years of 6% inflation. Always check whether a chart is in nominal or real rupees.
  • The fan widens over time. After 5 years the range is fairly tight. After 30 years it runs from zero to more than ₹16 crore. Uncertainty compounds, just like returns.
  • The lower edge is the one to plan around. The 10th percentile hits zero by year 30. That is the “bad but plausible” future, and in this run it is not rare: about one path in five ran out before year 30.
  • The upper edge is not a plan. The 90th percentile is what happens if markets are kind for decades. Nice if it happens; not something to spend in advance.

What 80%, 90% and 95% success really mean

A success probability is the share of simulated futures in which money lasted the whole horizon. It is useful, but easy to misread.

Assumptions for the table below: same model as above (10% average return, 14% volatility, 6% inflation); 10,000 paths per cell; success means the corpus never runs out within the horizon.

Horizon 2.5% start 3% 3.5% 4% 4.5% 5%
25 years 98% 95% 89% 81% 71% 60%
30 years 95% 88% 80% 69% 58% 46%
40 years 87% 76% 64% 52% 39% 30%
50 years 79% 65% 52% 41% 30% 21%

How to read it:

  • 95% means about 1 in 20 simulated futures ran out of money. In this model, that needs roughly a 3% start for a 25-year retirement, or 2.5% for 30 years.
  • 90% means about 1 in 10 failed. Many planners consider 85–90% a reasonable target if they will adjust spending in bad times.
  • 80% means about 1 in 5 failed. That can be acceptable if you have real flexibility: other income, the ability to cut spending, or a willingness to work again.

Three things a success rate does not tell you

  1. How badly the failures fail. Running out at 84 is very different from running out at 62. Look at when the failing paths hit zero.
  2. What you would actually do. Real people cut spending after a crash. Most simple simulations assume you keep withdrawing on autopilot, which makes them somewhat pessimistic for flexible people.
  3. Whether the assumptions are right. Change the average return or volatility and every number moves.

Why chasing 100% is not the goal

A plan with a 100% success rate in simulation usually means you worked years longer than needed, or underspent in most futures, and died with a large unspent corpus. Many planners prefer a high but not perfect probability, combined with clear rules for adjusting if things go badly.

The limits of the model

Monte Carlo is a tool, not an oracle. Simple versions, including the one behind the tables above, make simplifying assumptions:

  • Returns each year are independent of the year before.
  • Returns follow a bell-shaped distribution, which can understate extreme crashes.
  • Inflation is fixed rather than random.
  • Spending follows a fixed rule.

Use it to compare decisions (3% vs 4%, retiring at 45 vs 48, more equity vs less) rather than to predict a precise outcome.

Practical ways to reduce sequence risk

  • Hold a buffer: one to three years of expenses in cash, FDs or short-term debt, so you are not forced to sell growth assets in a fall.
  • Start lower: the table shows how much a half-point cut in the starting rate buys.
  • Agree on guardrails in advance: skip the inflation raise, or trim spending by a set amount, if the portfolio falls by a set percentage.
  • Keep some earning ability: even small part-time income in the first few years reduces withdrawals when they hurt most.

What to do next

This is general education about how these tools work, not personalised financial advice.

Frequently asked questions

What is sequence of returns risk?

It is the risk that poor investment returns arrive early in retirement, while you are withdrawing money. Selling investments when prices are low leaves fewer units to benefit from any recovery. Two people with the same average return can end up in very different places depending on the order of good and bad years. It matters most in the first five to ten years after you start withdrawing.

What is a Monte Carlo simulation in retirement planning?

It is a way of testing a plan against thousands of possible futures. The simulator generates many random sequences of yearly returns based on assumptions about average return and volatility, runs your withdrawals through each one, and reports how often the money lasts. The result is a probability and a range of outcomes rather than a single, falsely precise number.

What does a 90% success rate mean in a Monte Carlo simulation?

It means that in 90% of the simulated return paths, your money lasted for the whole planning horizon, and in 10% it ran out before the end. It is not a certainty, and it is only as good as the assumptions behind it. It also does not show how badly the failing paths fail, or how early, so look at the fan chart too.

What success probability should I aim for?

There is no universal right answer. Many planners treat 85% to 95% as a reasonable zone, combined with a plan to adjust spending if things go badly. Aiming for 100% usually means working much longer or spending much less than necessary in most futures. A lower probability can be acceptable if you are genuinely flexible or have other income sources.

How do I read a Monte Carlo fan chart?

The middle line is the median: half the simulated paths finish above it and half below. The shaded bands show ranges, for example the 25th to 75th percentile and the 10th to 90th. A widening fan means uncertainty grows over time. Pay most attention to the lower edge, because that shows what a bad but plausible future looks like for your plan.

How can I reduce sequence of returns risk?

Common approaches are keeping one to three years of expenses in cash or short-term debt so you do not sell growth assets in a fall, starting with a lower withdrawal rate, agreeing in advance to skip inflation increases or trim spending after bad years, and keeping some flexible income. None of these removes the risk, but each reduces how much damage a bad start can do.