Guide

Monte Carlo simulation explained for retirement planning

A Monte Carlo simulation runs thousands of possible market scenarios to test whether your retirement plan is likely to succeed. Instead of one projected outcome, you get a range of outcomes and a probability.

The problem with a single forecast

A traditional retirement calculator uses one average return rate — say 6% per year — and projects your portfolio value month by month. The result is a single projected date when you'll reach your FI target.

The problem: markets don't return 6% every year. Some years return 20%; others lose 30%. The average might be 6%, but the path matters enormously — especially when you're making contributions or withdrawals along the way.

Two investors with the same average return over 20 years can have completely different outcomes depending on the order in which those returns arrived.

How Monte Carlo simulation works

Named after the Monte Carlo casino, the method uses randomness to explore possible outcomes. In retirement planning, it works like this:

1. Define the model. Start with your current portfolio, monthly contribution, return assumption, inflation assumption, withdrawal rate, and retirement horizon.
2. Generate random returns. For each month in the simulation, draw a random return from a probability distribution (typically log-normal, calibrated to your chosen mean return and volatility).
3. Simulate one path. Apply the random returns month by month, adding contributions during accumulation and subtracting withdrawals during retirement, to produce one complete portfolio path.
4. Repeat thousands of times. Yield Clarity runs 10,000 independent paths. Each uses different random returns drawn from the same distribution.
5. Aggregate the results. Count how many paths reached the FI target, when they reached it, and whether the portfolio survived through retirement. Report the probability and the range.

Interpreting the results

Probability of success

If 4,900 of 10,000 paths reached your FI target by your planned date, the probability is 49%. This means that under the model's assumptions, roughly half of possible market scenarios would result in reaching FI on time.

This is not a prediction. It's a statistical summary of simulated outcomes. A 49% probability doesn't mean you have a coin-flip chance — it means the model's assumptions produce that result.

FI window

The FI window shows the date range within which the middle 50% of successful paths reached the target. A narrow window means outcomes are clustered; a wide window means there's significant spread.

Confidence bands

The fan chart shows the spread of portfolio values at each point in time. The 50% band contains the middle half of outcomes; the 80% band contains the middle 80%. Wider bands mean more uncertainty.

Sequence-of-returns risk

The order of investment returns matters most when you're making withdrawals. Poor returns in the first few years of retirement force you to sell more shares to cover the same income, permanently reducing the portfolio's ability to recover.

Monte Carlo simulation captures this naturally. Each simulated path has a different return sequence. Some paths have poor early returns and fail; others recover. The aggregate probability accounts for sequence risk without requiring you to model it explicitly.

Limitations of Monte Carlo simulation

Monte Carlo simulation is a powerful tool, but it has real limitations:

• Distribution assumptions. The model typically uses a log-normal distribution. Real markets have fat tails — extreme events are more common than the model predicts.
• Input sensitivity. The results are highly sensitive to the mean return and volatility assumptions. Small changes in inputs can meaningfully change the probability.
• No regime modelling. The simulation assumes returns are independently drawn each month. Real markets have regimes — periods of sustained bull or bear conditions.
• Not a guarantee. The probability is only as reliable as the assumptions. If the assumed return distribution doesn't reflect future market behaviour, the probability will be wrong.

Using it wisely

Monte Carlo simulation is most useful for comparing scenarios. Asking "is my probability 49% or 62%?" is less valuable than asking "does increasing contributions by £500/month meaningfully improve my probability?" The relative comparison is more reliable than the absolute number.