Most retirement planning asks “how much will I have?” The question that actually determines whether a retirement works is different: will it be enough, for long enough?
The first two guides in this series covered the machinery of investing — the tax-efficient wrappers that protect your returns, and the investment construction principles that determine how your money grows. This guide asks the question that determines whether any of it actually works in practice: does the money last as long as the retirement does?
For most of a working life, the planning question is relatively straightforward: save as much as possible, invest it sensibly, and let compound growth do its work. This is the accumulation phase — putting money in, growing it, leaving it alone.
Retirement introduces a fundamentally different dynamic. Money is now flowing out of the portfolio rather than into it. The portfolio is being drawn down at the same time as markets are moving. This is the decumulation phase — and the risks it introduces are different in kind, not just in degree, from accumulation risks.
Money flows in regularly. Compound growth works in your favour. Bad markets mean you buy more units at lower prices (pound-cost averaging). Time is your ally. Volatility is uncomfortable but not damaging. The focus is on growing the pot.
The key risk: not saving enough, or investing too cautiously and missing growth.
Money flows out regularly. Growth must outpace withdrawals. Bad markets mean you sell more units at lower prices (pound-cost ravaging) — but good markets, especially early on, can build a substantial buffer that carries the plan through any difficult years that follow. Time is now a constraint. Volatility that seemed manageable in accumulation can be catastrophic in early decumulation if a poor sequence arrives — but equally rewarding if a strong one does. The focus is on sustaining the pot through whatever sequence actually arrives.
The key risk: the portfolio running out before the retirement ends.
Three things make retirement sustainability genuinely difficult to plan for — not as a reflection of poor planning, but as intrinsic features of the problem.
The formal term for the probability that a portfolio reaches zero before the end of the planning horizon is probability of ruin — or risk of ruin. It was formalised by the academic work of Moshe Milevsky, whose research in the late 1990s and 2000s provided a mathematical framework for calculating and managing this risk.
The insight that makes this framework so useful is that ruin is not binary. It is not “this plan will work” or “this plan will fail.” It is a probability — and that probability can be measured, understood, and reduced by changing the inputs. That is what the rest of this guide explores.
Ruin is not about investments performing badly. It is about a portfolio reaching zero before the retirement ends. Understanding what drives it changes how you think about the whole retirement picture.
The academic foundation for risk of ruin in retirement planning comes primarily from the work of Moshe Milevsky, a finance professor at York University in Toronto. His key insight was that the sustainability of a retirement portfolio is not just a function of investment returns — it is the interaction of return, volatility, withdrawal rate, and time horizon that determines whether a portfolio survives.
Milevsky modelled retirement portfolios as a mathematical process and derived an approximate formula for the probability that a portfolio would be depleted within a given time period. The formula involves four key variables:
Withdrawal rate (w): the annual income drawn as a percentage of the starting portfolio value. A £500,000 pot drawing £20,000 per year has a 4% withdrawal rate.
Expected return (μ): the anticipated annual growth rate of the portfolio, which depends on its asset allocation and risk level.
Volatility (σ): the standard deviation of returns — how much the annual return varies around the expected return.
Time horizon (T): how long the portfolio needs to last. This is partly a planning decision and partly a function of longevity.
The probability of ruin increases when:
• The withdrawal rate is higher relative to expected return
• Volatility is higher (more uncertainty means more paths to zero)
• The time horizon is longer
The probability of ruin decreases when:
• Expected return comfortably exceeds the withdrawal rate
• Volatility is lower (smoother journey, less chance of a catastrophic sequence)
• The time horizon is shorter
The following bands illustrate how ruin probability might be interpreted in a planning context. These are illustrative only — actual probabilities depend entirely on individual circumstances and are calculated by the cash flow model.
Milevsky's formula is useful for understanding the shape of the relationship between withdrawal rate, return, volatility, and time — but it is a simplification. In a guided session, your own ruin probability is calculated using a more thorough method: a Monte Carlo simulation.
This is a meaningfully different — and more realistic — picture than a single constant-return line. It captures the randomness that the sequencing table above illustrates by hand, but across thousands of plausible paths rather than five hand-picked examples, and tailored to your own figures rather than a generic case.
Milevsky made an important distinction between two separate but related risks. Portfolio risk is the risk that markets perform poorly. Longevity risk is the risk that you live longer than the planning horizon assumed. Both contribute to ruin — but they require different responses.
A portfolio that would survive 25 years comfortably may be exhausted by year 32 if the plan assumed a 25-year horizon. Life expectancy at age 65 in the UK is currently around 19 years for men and 22 years for women — but these are averages. Half of all 65-year-olds will live longer than average. A 65-year-old couple has approximately a 50% chance that at least one of them lives past 90.
Small differences in how much you take out each year compound into enormous differences in whether the money lasts. The withdrawal rate is the single most controllable lever in retirement planning.
If you start retirement with a portfolio of £500,000 and draw £25,000 per year, your withdrawal rate is 5%. If you draw £17,500, it is 3.5%. That 1.5 percentage point difference seems small. Over 30 years, it is the difference between a plan that likely works and one that likely does not.
This tool illustrates how withdrawal rate and portfolio size interact, and how sensitive the outcome is to small changes — and, with the good- and bad-sequence lines, how much the order of returns matters even when the long-run average is identical. All figures are illustrative only and do not account for inflation or individual circumstances.
This table shows how different withdrawal rates interact with different real return assumptions over a 30-year horizon under constant return assumptions. For illustration only — sequence of returns risk is not captured here.
| Withdrawal rate | Real return 2% | Real return 3% | Real return 4% | Real return 5% | Real return 6% |
|---|---|---|---|---|---|
| 2.5% | Survives | Survives | Survives | Survives | Survives |
| 3.0% | Survives | Survives | Survives | Survives | Survives |
| 3.5% | Survives | Survives | Survives | Survives | Survives |
| 4.0% | Survives | Survives | Survives | Survives | Survives |
| 4.5% | ~30yrs | Survives | Survives | Survives | Survives |
| 5.0% | ~26yrs | Survives | Survives | Survives | Survives |
| 5.5% | ~23yrs | ~27yrs | Survives | Survives | Survives |
| 6.0% | ~21yrs | ~24yrs | ~29yrs | Survives | Survives |
| 7.0% | ~17yrs | ~19yrs | ~22yrs | ~26yrs | Survives |
The same average return over 20 years can produce completely different outcomes depending on when the good and bad years fall. This is the most important concept in retirement planning that most people have never heard of.
Here is a puzzle. Two retirees both start with £400,000 and both draw £20,000 per year. Over 20 years, both their portfolios average exactly 5% per year. One runs out of money at year 17. The other has £340,000 left at year 20. How? The answer is sequence of returns.
During accumulation, the order of investment returns does not matter for your final pot value. If you invest a lump sum and never touch it, a 20% loss in year 1 followed by 20% gains for the next 9 years produces exactly the same outcome as 20% gains for 9 years followed by a 20% loss in year 10. The maths is commutative.
In decumulation, the maths is not commutative. When you are drawing money out every year, losses early in retirement are catastrophic — and gains early in retirement are highly valuable. Here is why.
A 30% fall in the first three years forces you to sell a large number of units at depressed prices to meet your income need. Those units are gone — they cannot participate in the recovery. When markets recover, you have a permanently smaller portfolio generating income on a much-reduced base.
The portfolio has been ravaged at exactly the point it could least afford it — when it was largest and when each withdrawal represented the largest percentage of the remaining pot.
Strong early returns grow the portfolio faster than withdrawals deplete it. By the time the inevitable down years arrive, the portfolio is much larger than it started and can absorb the same percentage loss with a much smaller absolute impact on sustainability.
The early gains create a buffer that the bad sequence never builds. Same average return. Completely different outcome.
The effect does not need 20 years to show up. Academic research into retirement ruin and the sequencing of returns has used examples as short as three years to make the point unmistakable. Take a £100,000 pension fund at the point income begins, drawing £9,000 a year, starting at age 65. Across five different cases, the average return over the first three years is identical — approximately 7% a year. Only the order in which those returns arrive is different.
| Sequence | Year 1 | Year 2 | Year 3 | Money runs out at age | vs. even return |
|---|---|---|---|---|---|
| Even return | +7% | +7% | +7% | 86.5 | — |
| Variable | +7% | −13% | +27% | 83.3 | −38 months |
| Variable | +7% | +27% | −13% | 89.5 | +36 months |
| Variable | −13% | +7% | +27% | 81.1 | −65 months |
| Variable | +27% | +7% | −13% | 94.9 | +101 months |
Source: Moshe A. Milevsky PhD, "Retirement Ruin and the Sequencing of Returns," February 2006.
Every row has the same average return. Every row started with the same fund and drew the same income. The only thing that changed is the order in which the returns arrived — and the difference between the worst and best case here is more than thirteen years of sustainability. A loss in year one, even followed by strong recovery, costs far more than the same loss arriving later — because by year one the fund is at its largest, and every pound lost there is a pound that can never participate in the recovery that follows.
This simulator shows the same £400,000 portfolio, the same £20,000 annual withdrawal, and the same 20-year average return — under three different sequences. Illustrative only. Actual returns are unpredictable.
During accumulation, investing regularly into a falling market is actually beneficial — you buy more units at lower prices. This is pound-cost averaging, and it is one of the genuine advantages of regular saving.
In decumulation, the same mechanism can work in reverse. Drawing a fixed income from a falling portfolio forces you to sell more units at lower prices — this is pound-cost ravaging, the cruel mirror image of the accumulation benefit. But the reverse is equally true in a rising market: drawing the same income from a growing portfolio means selling fewer units while the pot continues to compound underneath the withdrawals. Over a full retirement you will experience a mix of both — which is exactly why the order returns arrive in matters so much.
Your investment risk level is not just about how comfortable you are with volatility. In retirement, it directly determines how likely your money is to last. The connection between the IA sectors and ruin probability is not incidental — it is central.
Guide 2 in this series introduced the IA sector framework as the common reference point for comparing investment risk levels. In accumulation, the choice of risk level primarily affects how fast the pot grows and how bumpy the journey is. In decumulation, it affects something more fundamental: the probability of ruin.
In retirement, risk level creates a genuine dilemma that does not exist in the same way during accumulation.
Benefits: smoother journey, lower volatility, less sequence-of-returns risk. A portfolio that does not fall 30% in year one cannot be ravaged by that fall.
Problem: lower expected real return. If the real return does not comfortably exceed the withdrawal rate, the portfolio depletes slowly but certainly. Low risk does not mean low ruin probability — it means ruin by erosion rather than ruin by crash.
A 2% real return on a 4% withdrawal rate is an arithmetic certainty of eventual depletion.
Benefits: higher expected real return, more likely to grow the portfolio faster than withdrawals deplete it. In a good sequence, significantly extends sustainability.
Problem: higher volatility means higher sequence-of-returns risk. A 35% fall in year one of retirement is devastating even if followed by strong recovery. And the recovery does not restore the units already sold.
Higher risk does not mean lower ruin probability — it means trading one type of ruin risk for another.
There is an optimal risk level for any given withdrawal rate and time horizon — the point at which the expected return is high enough to sustain withdrawals without the volatility being so high that a bad sequence becomes ruinous. Finding this point is one of the core jobs of retirement planning.
It is not a fixed answer. It depends on the withdrawal rate, the time horizon, other income sources (particularly state pension and any defined benefit pension, which reduce the required drawdown from the investment portfolio), and your capacity to flex spending in bad years.
If ruin probability is too high, there are only four things you can do about it. Each has a real cost. Understanding them honestly is the foundation of a retirement conversation.
When the cash flow model shows a ruin probability that is uncomfortably high, it is tempting to look for an investment solution — a better fund, a higher return, a smarter strategy. These may help at the margin. But the four levers that genuinely change retirement sustainability are simpler and more direct — and each involves a trade-off that deserves an honest conversation.
The three guides in this series have built a complete framework. The cash flow model is where that framework becomes a plan — specific, personal, and testable.
This guide has explained what risk of ruin is, what drives it, and what can be done about it. Guide 1 explained where to put money. Guide 2 explained how to construct what goes inside. These three frameworks — tax efficiency, investment construction, and retirement sustainability — are the intellectual foundation of a financial plan. The cash flow model is where they come together as a single picture.
A cash flow model is not a spreadsheet of optimistic projections. It is a tool that takes your complete financial picture — all income sources, all assets, all expenditure and liabilities — and projects it forward year by year, testing whether the plan survives under different scenarios.
State pension (when does it start, what does it pay?). Defined benefit pensions (guaranteed income floor). Investment portfolios across all wrappers (ISAs, SIPPs, workplace pensions). Property equity. Other income. Tax position year by year.
No single variable can be understood properly without the others. The cash flow model holds them all simultaneously.
What if markets perform below expectation? What if you live to 95? What if you need care costs at 82? What if you retire two years earlier? What if inflation runs at 4% for ten years?
Stress-testing the plan against realistic adverse scenarios is what distinguishes a robust plan from an optimistic projection. The cash flow model makes this possible.
Using stochastic modelling — running thousands of simulated market scenarios — the cash flow model calculates the probability that the plan survives its full planning horizon. This is the Milevsky framework applied to your specific numbers.
The output is not a prediction. It is a probability — and that probability can be improved by adjusting the four levers.
A structured data-gathering questionnaire begins the process of building your personal cash flow model. It gathers the information needed to set the key parameters — your risk profile (which determines the return and volatility assumptions), your income needs, your assets and their wrapper structures, your planning horizon, and your other income sources.
Once the model is built, it can be run under multiple scenarios and stress tests. The output shows your probability of ruin at different spending levels, how sensitive the plan is to the four levers, and where the key vulnerabilities in the plan lie. This is not something this guide can do — it requires your specific numbers, run using the model in a guided session.
This guide has given you the framework to understand what the model is doing and why every input matters. The withdrawal rate, the risk level, the time horizon, the wrapper structure — none of these are arbitrary choices. Each is an input to a calculation that determines the probability that everything you have worked for is enough, for long enough.
This series of guides is produced by Iain Ford, Director, FEP&G Ltd for educational purposes only. Nothing in these guides constitutes financial advice or a personal recommendation. The illustrative figures, probability bands, and scenario outputs are for conceptual illustration only and do not represent predictions of future performance or probability of ruin for any individual. Investment values can fall as well as rise. You may get back less than you invest. Past performance is not a reliable indicator of future results. The risk of ruin framework draws on the academic work of Moshe Milevsky and related literature; all figures used are simplified illustrative approximations. Always seek advice from a qualified, FCA-regulated financial adviser before making financial planning decisions. © Iain Ford, FEP&G Ltd 2026.