How the Retiris retirement model works

Step 1 — the years before retirement

Each year before your chosen retirement age, the model grows your portfolio by the expected return and adds your annual contribution. Contributions are entered in today's money and are assumed to rise with inflation, which is what happens in practice when you save a percentage of a salary rather than a fixed cash amount.

Growth is applied to the balance you started the year with, and the contribution is added at the end of the year. That is slightly conservative compared with paying in monthly, and it avoids overstating the compounding you get in the final year before you stop work.

Step 2 — the years after retirement

From the retirement age onwards, contributions stop and withdrawals start. Spending is entered in today's money and increased each year by the inflation assumption, so the plan keeps buying power constant rather than a constant cash amount.

Any guaranteed income you enter — State Pension, Social Security, a defined-benefit pension or an annuity — is subtracted from the amount that has to come out of the portfolio, starting in the year you say it begins. This is usually the single biggest lever in a plan: a guaranteed income of £12,000 removes roughly £300,000 of pressure from a portfolio funding a 30-year retirement at a 4% withdrawal rate.

Real and nominal balances

Every year is reported twice: the nominal balance (future pounds or dollars) and the real balance (today's money). Charts on this site default to real balances, because a £1.4m nominal balance in 2056 is not a £1.4m lifestyle.

Step 3 — the 1,000 simulated runs

Real markets do not deliver 6% every year. The simulation repeats the same plan 1,000 times, drawing each year's return at random from a distribution centred on your expected return with the volatility you set. It then reports three things:

• Success rate — the share of runs where the money funded spending for the whole horizon.
• Median years funded — how long the money lasted in the middle run.
• The 10th, 50th and 90th percentile paths — an unlucky, a typical and a lucky market.

The random number generator is seeded, so the same inputs always produce the same figures. That is why a number quoted in one of our scenarios can be reproduced by anyone entering the same assumptions into the calculator.

Why volatility matters as much as average return

Two plans with the same 6% average return but 8% and 18% volatility behave very differently in drawdown. Withdrawals lock in losses when markets fall early, so a higher-volatility plan with an identical average return has a materially lower success rate. This is sequence-of-returns risk, and it is the main reason we publish a range rather than a single headline number.

What the model does not do

• Tax. No income tax, capital gains tax or tax-free lump sums are applied. Enter spending as the gross amount you need to withdraw.
• Charges. Platform and fund fees are not deducted. Subtract them from the expected return — 0.4% of charges means entering 5.6% rather than 6%.
• Changing spending. Spending is level in real terms. Most retirements are not: spending typically falls through the seventies and can rise again with care costs.
• Property, business assets or debt. Only the invested portfolio is modelled.
• Longevity. You set the planning horizon. The model does not estimate life expectancy.

How to read a success rate honestly

A 90% success rate does not mean a 10% chance of destitution. It means that in 10% of simulated markets the portfolio ran out before the horizon, at which point real people would cut spending, work part-time, or draw on other assets. Treat the success rate as a measure of how much slack the plan has, not as a probability of disaster.

Version and review

The model is reviewed whenever we change its maths or defaults, and the review date at the top of this page is updated at the same time. Our wider standards for sourcing and corrections are set out in the editorial methodology.