A JC Wealth resource

The Drawdown
Atlas.

Mapping the decisions that shape the first decade of retirement — and every decade that follows.

Jessica Cook · LLB (Hons) Chartered MCSI

JC Wealth The Drawdown Atlas Mapping the decisions that shape the first decade of retirement Jessica CookChartered MCSI

A working reference, not a book.

The Atlas is the resource I would want my own family to have in front of them in the first decade of retirement. Ten sections, working through the structural decisions that retirement actually asks of a household - withdrawal sequencing, sequence risk, the four percent rule, annuities, inheritance tax, the international dimension, care planning, and when an outside voice in the room genuinely helps.

You can read it cover to cover, or you can let me ask you three short questions and point you to the sections most relevant to where you are. The sections sit alongside my weekly column in The Times of London and the broader work I do with private clients.

The ten sections

Browse all →

Show me where to start

Three short questions.

I'll point you at the sections of the Atlas most relevant to where you actually are.

Skip — let me browse instead

Your starting point

Start with the frame, then the sections built for your situation.

However you answered, the Atlas reads best against the picture set up at the very start. Begin with my note and the two foundation sections — a few minutes each, and they make everything after them land — then move on to the sections matched to what you told me.

Start here first

And for your situation specifically

And the rest of the Atlas if you want to go further:

A note from Jessica

Welcome.

If you have arrived here, you are at one of the more difficult moments in a financial life. After a working career spent saving, accumulating, paying down a mortgage, and adding to pensions and ISAs, you are now being asked to do something quite different. You are being asked to spend what you have built, in a way that lasts. Much of the work that determines how the rest of retirement feels gets done in its first decade.

This is harder than it sounds. The savings discipline that brought you here will, if you let it, follow you into retirement and turn into an under-spending discipline. The investment habits that worked for accumulation will, if you do not reorganise them, work against you in decumulation. Market falls that were inconvenient when you were earning become much more consequential when you are drawing. And nearly every received-wisdom rule — the four percent rule, draw pensions last, take the maximum tax-free cash up front — needs more thought than the shorthand suggests.

This Atlas sets out the framework I use with my own clients when we are building a retirement income arrangement from the actual numbers. It is a working tool in long form, written for readers who would rather think structurally than collect tips. The middle of the Atlas does the structural work: mapping the full picture, deciding the order of withdrawal, handling sequence risk, putting the four percent rule in its proper place, and thinking honestly about whether annuitisation has a place in your particular household. The sections that follow take on estate planning, the international dimension, and the later years. The closing section is about when to ask for personalised help.

You can read it cover to cover, or you can dip into the sections that meet you where you are. Sections two and four are the ones many households come back to. The reader I had in mind is somewhere in the first decade of retirement, with a household picture in the high six- or low-seven-figure range, possibly with a cross-border element, and quite reasonably reluctant to accept that the next twenty or thirty years of their financial life can be reduced to a single rule of thumb.

I hope it helps.

Jessica Cook Summer 2026

Before you begin

A note on what this Atlas is.

This Atlas is educational. Nothing in it constitutes personal financial advice, a recommendation, or a solicitation. The framework set out across the ten sections is structural; the specific decisions any household needs to make depend on its actual numbers, family pattern, longevity assumptions, residency position, and tax position.

The figures, allowances, and tax rules described are accurate as at Summer 2026 and may change. Any reader considering action on the substance covered here should commission personalised advice from a qualified financial planner regulated in the relevant jurisdiction.

The fuller note

What this Atlas is and is not.

This Atlas is educational. Nothing in it constitutes personal financial advice, a recommendation to take or refrain from any specific course of action, or a solicitation to engage any regulated firm. The framework set out across these ten sections is a structural one; the specific decisions any household needs to make depend on its actual numbers, family pattern, longevity assumptions, residency position, and tax position, and those decisions should be taken with personalised advice from a qualified planner.

Tax rules, allowances, thresholds, and rates change. The figures, regimes, and policy positions described in this Atlas are accurate as at Summer 2026 and may have changed by the time you read them. The 2027 pension-into-estate change, the post-April 2025 long-term resident regime, the four-year Foreign Income and Gains regime, and the country-specific tax positions in §8 are all areas where the operational detail is still bedding in and where current advice will be more reliable than current writing.

Worked examples — Susan and Peter at £3.08 million combined wealth, David and Helen in Spain, the annuitisation routes at £200,000 and £100,000, the bucketing illustrations at various capital sizes — are illustrative of the mechanics rather than recommendations. Any reader's own numbers will sit at a different place inside the same framework.

Jessica Cook is a Chartered MCSI & Financial Planner, regulated by the Financial Conduct Authority. The Atlas is published independently of any product provider, platform, or investment scheme.

All ten sections

The full Atlas. Filter by the topic that matches where you are, or browse all.

Peter is eighteen months into retirement. He retired from a senior marketing role at a FTSE 250 business, took the package, told the colleagues he would be in touch, and walked into the rest of his life on a Friday afternoon. The first thing he did the following Monday was check the Standard Life balance, and he has been checking it more often than he means to ever since.

The habit has the texture of a low-level anxiety rather than a deliberate exercise. He logs in, sees a number, translates a market move into months of essential spending without quite meaning to, and catches himself doing it. Peter has stopped thinking of his money as something that grows and started thinking of it as something finite, but he has not yet built the structured framework that replaces how much am I earning with how much can I safely spend. The instinct is the natural starting point. The framework is what the rest of this guide gives him.

Decumulation is a different discipline

The mindset matters before the framework. Accumulation is a discipline of regularity. The household earns, saves a portion, invests the savings into a long-horizon portfolio, holds through the noise, and repeats the cycle for thirty years. The arithmetic forgives a great deal. A bad year early on barely shows by the time the working years are finished. Time is the load-bearing variable, and time is on the saver's side.

Decumulation works differently. The household is no longer adding to the pot; it is taking from it. Time is no longer accumulating; it is being spent. A bad year early on now matters disproportionately, in the precise sense §4 will set out under sequence risk. The household's income is no longer arriving from outside the pot; it is being generated by drawing from inside it. Each of these shifts is small in isolation. Together they make decumulation a different game from the one the household has just spent thirty years playing well.

The first twelve to twenty-four months

The discipline does not arrive automatically the day the salary stops. It is built deliberately, over the first twelve to twenty-four months, through a small set of decisions taken with intention rather than by default. The household that enters retirement without naming the shift continues to operate on accumulation instincts — saving harder than the situation requires, sitting in cash that erodes against inflation, panicking at the first market fall, deferring spending the plan can comfortably support. Much of the avoidable damage in early retirement comes from this gap between instinct and discipline, and the section's job is to help close it.

Shift one: from income thinking to sustainability thinking

The first shift is from income thinking to sustainability thinking. Working years organise around the question how much am I earning? Pay rises, bonuses, promotions all answer that question, and the household calibrates its spending against the answer. The mental model is straightforward: income arrives, is partly spent and partly saved, and the saving has its own purpose somewhere in the future.

Retirement asks a different question: how much can I safely spend across the rest of two lives? The answer is not a number arrived at by adding the pots together and dividing by the years. It is a sustainable draw rate, modulated by the structure of the assets, by the household's longevity assumptions, and by the income floor already in place from state pensions and any defined benefit entitlement. Section 5 develops the technical machinery; the shift this section asks for is mental. The household stops asking "what is the value of the pot today" and starts asking "what is the annual spend the pot can support across the horizon I plan to live across". The two questions are different and they yield different answers.

Shift two: from saving discipline to spending permission

The second shift is harder. A household that has spent thirty years saving has built a deep, often unconscious habit of treating spending as the thing to be minimised. The habit served the household well. It is also the habit that, left unchecked, becomes an under-spending discipline in retirement. The household draws materially less than the plan could safely support — the holiday is deferred, the kitchen refit is deferred, the help to adult children with a deposit is deferred, the trip to see grandchildren overseas is deferred and then deferred again. The plan made room for those things. The instinct will not let the household take them.

Giving oneself permission to spend what the plan supports is a deliberate act. It usually requires a structured income arrangement that makes the safe-spend number visible and trustworthy. Without that structure the household reverts to the saving instinct and treats every withdrawal as a small transgression, even when the plan was built to support exactly that withdrawal. With the structure in place, the household can spend what it has earned the right to spend without the anxiety of wondering whether each pound out of the pot is the pound that breaks the arithmetic. The arithmetic has been done in advance, and the answer is in front of them.

The word "permission" matters. The shift moves from one form of care to another: the careful saving discipline that served the household well during the working years gives way to the careful spending discipline that serves the household well now. Both are forms of care. They look different in practice.

Shift three: from single-horizon investing to multi-horizon investing

The third shift is structural rather than emotional. Working years organise the household's invested capital around a single horizon — the date of retirement — and a single risk-and-return profile sized to that horizon. Decumulation needs more than one. The money required for the next year or two has a different job from the money required for years two to five, which has a different job from the money required for everything from year five onward. Each layer holds the assets best suited to its horizon: cash and short-dated instruments for the near layer, fixed income and lower-volatility growth for the middle layer, equities and return-seeking growth for the long layer. Section 4 develops this structure substantively. The shift this section asks for is recognition that decumulation requires a different portfolio architecture from the one accumulation called for.

A pattern from practice

Consider the typical household in the first year of retirement. The income floor is already in place from state pension and a defined benefit entitlement, the invested capital is sized to support the spending the household has been planning for, and the arithmetic comfortably accommodates the holiday they have been promising themselves for a decade. The household defers the holiday, and then defers it again. The plan supports it; the saving instinct will not give them permission to take it. By the time a structured arrangement is in place that lets them see the safe-spend number on a single sheet of paper, two years have passed. The trip happens in year three rather than year one. The arithmetic does not change in the meantime. What changes is whether the household feels permitted to act on what the arithmetic has been telling them.

A note on patience

Not all three shifts land at once, and the section does not pretend they will. The household that recognises the income-to-sustainability shift first usually finds the multi-horizon shift second; the spending-permission shift often takes the longest. The first twelve to twenty-four months are when the discipline gets built, not when it is finished. Naming the three shifts matters because the household can then see what is changing as it changes, and recognise the discomfort of the transition for what it is rather than for something more alarming.

Before the next section

The action this section asks for, before §2 begins, is small and concrete. Gather the most recent statement for every pot the household holds. State pensions, defined benefit entitlements (request a fresh statement if the most recent is more than a year old), defined contribution pensions, ISAs, the general investment account, and any foreign-held holdings. The next section asks the household to put all of it on a single page. Without the inventory in front of you, the structured thinking the rest of this guide depends on cannot begin.

Most households arrive at retirement with pots in several places. A workplace pension or two, a self-invested personal pension holding the transfers from earlier employers, one or two ISAs, and often a general investment account for bonus years when the ISA allowance was already full. Frequently a defined benefit entitlement from a role held long enough ago that the statements are filed in a drawer rather than an inbox. Two state pensions due at different dates if the partners are different ages, and sometimes a foreign pension left behind after a stretch of overseas work.

Very few households have seen all of it on a single page. Susan and Peter had not, until they sat down one Saturday with the most recent statements from six providers, a projection from the pension service at gov.uk, and rather more coffee than the job strictly required.

Before any decision about how to draw can be made, the household needs to see what it holds. Every later decision in this guide rests on having that picture in front of you. A household drawing for a decade from an incomplete map has been making good decisions against bad information.

Why the picture comes first

The logic is simple. Withdrawal ordering, sequence risk, annuitisation, estate planning, the international question. Each of these rests on knowing exactly what the household holds, in what wrappers, producing what income, taxed how. The errors that creep in are rarely dramatic. A defined benefit entitlement whose survivor provision was misremembered when the household last looked. A general investment account whose value sits at last quarter's number rather than this week's. A forgotten pot from a former employer that nobody has thought to add to the list. Each of these on its own is small; together they shape what the household believes its position to be, and the household that is drawing for a decade against a partial map has been making good decisions on imperfect information.

The pot map is also the input to the organising structure §4 will introduce. Where each pot goes, and how the household thinks about the time horizon of its assets, will be reorganised there. You do not need to anticipate that yet. The job of this section is inventory.

The Different Categories

For most people, retirement resources fall into the five main categories described below, although some may also hold other assets such as EIS investments, AIM portfolios or investment bonds. This list is not exhaustive. Each is doing a different job, and each is taxed differently. Knowing which category each of your pots belongs to is the first layer of the mapping exercise.

State pension. Indexed income, paid for life from state pension age, uprated each April under the triple lock. The full new state pension currently pays around £11,500 a year, though anyone with gaps in their National Insurance record receives less. Taxable at your marginal rate. A state pension forecast from gov.uk is the starting point; it tells you how much you are on track to receive and from what age. State pension ends with the person who earned it; there is no transfer of the whole entitlement to a surviving spouse, though some uprating rules apply.

Defined benefit pensions. Scheme pensions paying an indexed income for life, usually from a selected retirement age, with survivor benefits that vary by scheme. The capital behind the income is held by the scheme, not the member. Ongoing income is taxable at marginal rates. At retirement most schemes offer the option to commute part of the annual pension into a tax-free lump sum, at a scheme-specific conversion rate; the choice is between cash at the start and a higher income for life. Survivor provision is typically half or two-thirds of the member's pension, though the specifics live in the scheme booklet. A defined benefit entitlement is often the most valuable income stream in a retired household, and also the one most likely to have been forgotten about if the member left the scheme many years ago.

Defined contribution pensions. Capital sitting inside a pension wrapper, drawn flexibly from age 55 (rising to 57 in 2028). A quarter can be taken tax-free across the lifetime of the pot; the remaining three-quarters are taxable at marginal rates on withdrawal. Defined contribution pensions currently sit outside the estate for inheritance tax purposes. From April 2027, unused pensions come into the estate, which reshapes the withdrawal sequencing of §3 and the estate planning of §7. The change is named here for completeness. Its substantive treatment is in §7.

ISAs and general investment accounts. Two different wrappers holding invested capital. They can look similar on a platform statement and behave quite differently in tax terms. ISAs grow and withdraw free of income tax and capital gains tax, and sit inside the estate for inheritance tax purposes. General investment accounts attract capital gains tax on realised gains and income tax on dividends and interest as they arise; on death, assets receive an uplift to market value that resets the unrealised capital gains position for the heirs.

Primary residence. Not income-producing, but the household's largest single capital asset for many readers of this guide. Its role in retirement is shelter rather than income. It earns its place in the pot map because it is the biggest single line in any estate calculation, and because the residence nil-rate band, an additional IHT allowance of up to £175,000 per person tapered away above £2 million of total estate, applies where a direct descendant inherits the main home.

The five categories below, set out in one place, are the taxonomic spine of the rest of the guide.

Category	How it pays out	Tax in life	Tax at death
State pension	Indexed income, lifetime	Income tax at marginal rate	Ends with the person; limited spousal uprating only
Defined benefit	Indexed income, lifetime; tax-free cash via commutation; survivor provision varies	Income tax at marginal rate on pension income	Scheme-defined survivor; otherwise ends
Defined contribution	Capital drawn flexibly; 25% tax-free	75% of withdrawals at marginal rate	Outside estate pre-April 2027; inside from April 2027
ISA	Capital withdrawn at will	No income tax or CGT	Inside estate for IHT
GIA	Capital withdrawn at will	CGT on realised gains; income tax on dividends and interest	Inside estate for IHT; CGT uplift at death
Primary residence	No income; shelter	None while occupied	Inside estate; RNRB available where a direct descendant inherits

The table is reference. It does not say which pot to draw from first, or in what order, or how to sequence across a retirement. That is §3. The job of this section is to see which of these categories the household has, and in what amounts.

Susan and Peter's picture

Susan and Peter's map, once assembled, sits on a single sheet of paper. They are the running example across this guide, and the numbers below are the ones every later section builds on.

Their state pensions are split by age. Peter, now 67 and eighteen months into retirement, is drawing his at the full rate, around £11,500 a year. Susan, 64, reaches state pension age in two years and her projection from gov.uk shows the full new rate when it begins. Once both are in payment the household has around £23,000 a year of indexed state pension income.

Their defined benefit position is asymmetric. Peter's FTSE 250 scheme pays £38,000 a year, indexed, with a 50 per cent survivor pension for Susan. Susan has a smaller entitlement of £7,000 a year, indexed, paid from her current age, from an early-career role she sometimes forgets to count because it was accrued so long ago. Combined, their defined benefit income runs at £45,000 a year. Once both state pensions are fully in payment, the guaranteed indexed floor of the household comes to around £68,000.

Their defined contribution position, by contrast, sits in capital rather than income. Peter's Standard Life arrangement holds £920,000, consolidated two years ago from three separate ex-employer pots into a single drawdown contract. Susan's Aviva pot holds £230,000. Between them the defined contribution capital totals £1.15 million.

Their joint ISAs at Hargreaves Lansdown hold £380,000, held across a mix of equity income funds and sensible trackers. A general investment account of £150,000 sits beside them, the residue of the Surrey house sale that put them into the Cotswolds property. The GIA has not yet been properly invested; Peter has left it sitting in a cash-like fund while he works out what to do with it. He will revisit that in §3.

Their primary residence, the Cotswolds house, is worth around £1.4 million and is mortgage-free.

Their two adult children, Ben and Anna, are both based in the UK. Ben and his partner have two young children of their own — the household's grandchildren show up in planning conversations more often than the numbers strictly require, and the family is closer-knit than the pot map alone would suggest.

The completed map looks like this.

Category	Provider / scheme	Current value or accrued entitlement	What it produces
State pension (Peter)	DWP	In payment	~£11,500 / year, indexed
State pension (Susan)	DWP	Starts in 24 months	~£11,500 / year from age 66, indexed
Defined benefit (Peter)	FTSE 250 scheme	In payment	£38,000 / year, indexed; 50% survivor
Defined benefit (Susan)	Early-career scheme	In payment	£7,000 / year, indexed
Defined contribution (Peter)	Standard Life drawdown	£920,000	(drawn flexibly)
Defined contribution (Susan)	Aviva	£230,000	(drawn flexibly)
ISAs (joint)	Hargreaves Lansdown	£380,000	(drawn flexibly)
General investment account	(house-sale residue)	£150,000	(drawn flexibly)
Primary residence	Cotswolds home	£1,400,000	Shelter

Two numbers matter most. Guaranteed indexed income of around £68,000 a year once both state pensions are fully in payment. Investable capital outside the house totalling £1.68 million. Every later section of the guide treats those two numbers as its starting point.

What the exercise produced for Susan and Peter was less a new set of facts than a reorganisation of what they already knew. Peter had been calculating withdrawals against the Standard Life pot as though it were the only pot, mentally excluding Susan's Aviva holding because it was in her name. Susan had been treating the Aviva pot as a back-up resource rather than as part of the household picture. Neither had internalised the scale of the guaranteed income floor, because the defined benefit statements had been filed, paid quarterly, and never added up. Each piece of the picture they had been planning against was correctly known in isolation; what they had been missing was the shape the pieces formed once set out together.

For readers with foreign-held resources

The mapping discipline is the same for households with assets or entitlements outside the UK; the categories widen. A Spanish-held investment account, a French pension entitlement earned during a working decade there, a US Roth IRA from time in New York, or a Hong Kong MPF pot carried since a posting in the late nineties each sits outside the UK pot map until it is deliberately included. Each foreign holding also brings its own tax treatment under the destination country's rules, which may or may not align with the UK treatment in the table above. Section 8 covers the international layer in detail. For the purpose of this exercise, every holding belongs on the same sheet of paper, in whichever category it sits, with a note of the country of origin against its line.

The forgotten pot

The exercise above is the one discipline that forces a complete inventory, and completing it tends to surface at least one pot that had not been properly counted. Consider the typical household in the first three weeks of retirement planning, both partners gathering their statements, scrolling through old email folders, and sifting drawers in the study. One partner's list of former employers includes a role from the 1990s where the pension scheme was left behind at the point of leaving. A letter to the old employer's HR department, or a search on the Pension Tracing Service, eventually surfaces an entitlement that has been quietly accruing for a quarter of a century. Sometimes the entitlement is a small defined benefit of three thousand pounds a year, indexed and payable from a retirement age already in the past. Sometimes it is a defined contribution pot of sixty to a hundred and twenty thousand pounds that received four years of contributions and has sat invested ever since.

The size of the rediscovery varies; the effect on the map is the same regardless. A forgotten pot is the most common reason a household's completed picture looks materially different from the one they started the exercise believing in. Without the inventory discipline, the forgotten pot stays forgotten, and every subsequent decision rests on a picture with a hole in it.

Your own pot map

The exercise in this section is the same for every household; only the numbers change. Complete the table below before moving on to §3. A blank template follows, with space for many households.

Category	Provider / scheme	Current value or accrued entitlement	What it produces
State pension (yours)	—	—	—
State pension (partner's)	—	—	—
Defined benefit (yours)	—	—	—
Defined benefit (partner's)	—	—	—
Defined contribution (yours)	—	—	—
Defined contribution (partner's)	—	—	—
ISAs	—	—	—
General investment account	—	—	—
Primary residence	—	—	—
Other (overseas pensions, inherited assets, business interests)	—	—	—

↓ Download printable worksheet PDF · A4 · one page · Summer 2026

State pension projections are available through the gov.uk pension forecast service. Defined benefit entitlements are on the most recent scheme statement, which arrives annually; if the last statement is more than a year old, request a fresh one. Defined contribution valuations come from each provider's platform; ISAs and the general investment account from the platform statement. The value of the primary residence can reasonably be estimated from comparable sales in the neighbourhood within the last six months.

A downloadable spreadsheet version of this template is planned for a future release of the guide. For the purposes of the work this guide is asking of you, a clean printed copy of the table, filled in by hand, does exactly the job it is meant to do.

Before the next section

Complete the table before moving on. A household that has not done the inventory work has a picture with gaps in it, and every later decision in this guide — the order of withdrawal in §3, the bucketed architecture in §4, the choice of whether to extend the income floor in §6, the estate question in §7 — rests on the picture being complete. Susan and Peter's numbers are in front of you; your own need to be on the same sheet.

The next section takes the completed map and asks the first question a retired household should be asking of it. Given what we hold, in what order should we draw? What goes first, what stays invested, and what we reasonably leave alone.

The pot map is the input. The order of withdrawal is the first thing it lets a household decide.

Many retirees make this decision once, in the first few weeks of retirement, by instinct. The defined contribution pension gets drawn first because it is the easiest to start. The ISAs get deferred because they feel like savings rather than income. The general investment account gets ignored because no-one in the household has thought of it as a retirement resource. Each of these choices is reasonable in isolation. Together, they are rarely the choices that hold up across the full retirement.

This section asks the question deliberately. Given what the household holds — and now sees in front of it — in what order should it draw?

The two-stage framework

The order-of-withdrawal decision is two questions, asked in sequence.

The first question is which time-horizon layer of the household's wealth funds this year's spending. The household's resources have different jobs across the retirement horizon. Some are sized for the next two or three years. Some for the next ten. Some to grow across the rest of the retirement and to provide for the years that begin twenty years out. The first question is which of those layers takes the draw. That question sits at the architectural level, and §4 covers it substantively under the bucketing structure. Here, it is enough to name it.

The second question is which specific pot inside the chosen layer to liquidate. Once the layer is chosen, the household still has to decide which wrapper provides the cash. This question sits at the tax level, and it carries more weight than the first. Many retirees have never thought about it systematically.

Visual A · Two-step decision diagram

Step One

Which time-horizon layer funds this year's draw?

→ §4

›

Step Two

Which wrapper inside that layer do we liquidate?

this section

The framework that follows develops the second question. The bucketing structure of §4 stands behind it.

The three layers, in brief

The bucketing structure has three layers, sized by horizon. The near-term layer covers years one to two and holds the cash, money market funds, and short-dated gilts that the annual draw is taken from. The mid-term layer covers years two to five and holds the fixed income and lower-volatility growth that refills the near-term layer on a sensible cadence. The long-term layer covers everything from year five onward and holds the equity weighting a long retirement requires.

§4 develops the architecture in full and explains why this structure is the right answer to the sequence-of-returns question, with the visual layout that makes each layer's job concrete. For this section, the relevant point is simpler: the draw comes from the near-term layer, and the wrapper question — which specific pot to liquidate inside it — is what carries the rest of §3.

The three lenses on the wrapper question

Once the layer is chosen, the wrapper question has three lenses. All three apply at the same time, and they pull in slightly different directions. The reader's job is to hold all three in view and to accept that no single answer wins on every lens.

The first lens is the tax this year. Which pot is cheapest to draw from in the current tax year, given the household's other income and the personal allowances and basic-rate bands in play? The answer is often the pot whose withdrawal lands in unused tax space — the lower-earning partner's defined contribution pension, the general investment account where the gain sits inside the annual capital gains tax allowance, the ISA where the withdrawal carries no tax at all.

The second lens is the lifetime tax. Which order minimises tax across the full retirement, not just this year? The answer often differs from the first lens. Taking ISA money this year because it is tax-free now may be the wrong choice if it leaves a larger pension pot to draw down later in higher-rate territory.

The third lens is the estate. Which pot is best left in place because of how it will be treated at death? The answer here has been reshaped by the pension change due in April 2027, and the section returns to that in a moment.

Where the lenses agree, the wrapper that wins all three is the right pot to draw from this year. Where they disagree, the framework asks the household to weigh them. For many households at this wealth level, the lifetime-tax lens becomes the deciding consideration when it pulls clearly against the tax-this-year answer. The estate lens carries more weight at older ages and at higher wealth.

Two short illustrations show how the lenses argue with each other before the framework gets put to work on Susan and Peter.

Visual F · Two scenarios where the lenses disagree

Scenario one — ISA versus pension

Drawing £15,000 from the ISA this year is free of tax; the tax-this-year lens favours the ISA. Drawing the same £15,000 from the lower-earning partner's pension uses unused basic-rate band and costs around £2,250 in income tax this year. But if the pension is left untouched and grows to £900,000 over the next decade, drawing it then in higher-rate territory costs more than the £2,250 saved. The lifetime-tax lens points the other way.

Scenario two — GIA versus pension

Drawing £12,000 from the general investment account this year, where the gain sits inside the annual capital gains tax allowance, costs nothing; the tax-this-year lens favours it. But the GIA carries the capital-gains-uplift advantage at death; drawing it down now sacrifices that uplift, while the pension has lost its inheritance-tax exemption from April 2027. The estate lens points toward draining the pension first and leaving the GIA in place.

What each wrapper costs in tax

The table that follows is the wrapper-by-wrapper picture of what each pot costs in tax when drawn — both in life and at death. It is the reference behind the three lenses above. The 2027 change sits in the defined contribution row; the substantive treatment is two sections on.

Wrapper	Tax on withdrawal in life	Tax position at death
State pension	Income tax at marginal rate on the full amount	Ends with the person; no transferable balance
Defined benefit pension	Income tax at marginal rate on the pension income; tax-free cash via commutation if elected at retirement	Scheme-defined survivor pension; otherwise ends
Defined contribution pension	25 per cent tax-free up to the Lump Sum Allowance cap; remaining 75 per cent at marginal rate	Outside estate pre-April 2027; inside estate from April 2027
ISA	None	Inside estate for IHT
General investment account	Capital gains tax on realised gains above the annual allowance; income tax on dividends and interest	Inside estate for IHT; capital gains uplift to market value at death

The defined benefit commutation cell is included for completeness. The parallel decision — whether to take the tax-free cash from a defined benefit pension as a lump or to take the larger income — sits in §6, where the annuity-versus-drawdown question carries the same shape and is treated together.

Susan and Peter, two phases

The household. Susan, 64, and Peter, 67. He retired eighteen months ago from a senior marketing role; Susan is winding down her freelance HR consulting. Their pot map shows a guaranteed indexed floor today of £56,500 — Peter's state pension at £11,500, his defined benefit pension at £38,000, and Susan's earlier-career defined benefit at £7,000 — reaching £68,000 in twenty-four months when Susan's state pension begins. Their investable capital is £1.68 million across Peter's Standard Life drawdown pot at £920,000, Susan's Aviva pot at £230,000, joint Hargreaves Lansdown ISAs at £380,000, and a general investment account at £150,000. Their target net spending is £72,000 a year.

Visual E · The bridge years

The next twenty-four months are the bridge years. Peter's combined state-and-defined-benefit gross of £49,500 sits just below the higher-rate threshold. Susan's defined benefit pension of £7,000 plus her £8,000 of freelance leaves her with material unused basic-rate band that the household has not been making use of. Combined net from non-investment sources falls short of the £72,000 target by something in the region of £12,000 to £15,000.

The question §3 asks of the bridge years: which pot funds the gap? Three sensible candidates present themselves. Susan's Aviva pot — her unused basic-rate space is the cheapest income tax anywhere in the household this year, and the pension can be drawn either as an Uncrystallised Funds Pension Lump Sum or through flexi-access drawdown (the next subsection covers the choice). The general investment account — the gain sits inside the annual capital gains tax allowance and the ISA wrapper is preserved. The joint ISAs — zero tax this year, with the wrapper available to be taken last.

The framework asks the household to weigh the three lenses against the three candidates.

Visual B · The three lenses applied to Susan and Peter's three candidate pots

	Susan's Aviva pot (£230k)	Joint ISAs (£380k)	General investment account (£150k)
Tax this year	FavourableUses Susan's unused basic-rate space.	Most favourableZero tax on the withdrawal.	FavourableGain sits inside the annual CGT allowance.
Lifetime tax	Most favourableBasic-rate now is cheaper than higher-rate later.	UnfavourableLeaves pension growing into higher-rate territory.	NeutralDepends on the gain trajectory.
Estate (post-2027)	Most favourablePension is now in the estate; drawing reduces IHT exposure.	NeutralAlready in the estate; drawing changes nothing.	UnfavourableSacrifices the capital-gains uplift at death.

No single column wins on every row. Where the lenses pull in different directions, the framework asks the household to weigh them.

The grid does not produce a single winner. Susan's Aviva pot wins on lifetime tax and on estate. The ISAs win on tax this year. The general investment account wins on tax this year and looks neutral on the others. The framework points clearly toward a combination — drawing some from Susan's Aviva pot and some from the general investment account is more tax-efficient than the household's current pattern of drawing predominantly from Peter's Standard Life pot. Peter and Susan had not modelled it that way. Many households have not.

Twenty-four months from now, the picture shifts. Both state pensions are flowing; the guaranteed floor reaches £68,000. Susan's freelance income has stopped. The combined net from non-investment sources still falls short of the £72,000 target by something close to £15,000, but the composition is different and the tax bands have moved. Peter is now further from the higher-rate threshold than he was, because Susan's state pension has reduced the share of household income flowing through his name. Susan has a full basic-rate band against which her Aviva pot can be drawn — her own state pension uses some of her personal allowance, and the pension drawing fills the rest of the basic-rate space at the lowest income tax rate available.

The framework now favours Susan's Aviva pot more clearly than it did in the bridge years. The basic-rate-band use is unambiguous. The general investment account's capital-gains-uplift advantage is stronger as the household ages. The post-2027 logic on the defined contribution pension reinforces the case for drawing it down rather than leaving it as a deferred estate asset.

The teaching point is the contrast between the two phases. The right order is configuration-dependent — it is the answer to the lens question at the household's current pattern of income, allowances, and wrappers, and that configuration changes when income flows shift. The bridge years are the window in which the most consequential sequencing decisions get made, because they are when the lens-vs-lens trade-offs are sharpest and when changes to the household's pattern can be made without disturbing income that is already in flow.

UFPLS versus flexi-access drawdown

When Peter draws from his Standard Life pot, the mechanical question that lands first is whether the right method is an Uncrystallised Funds Pension Lump Sum (UFPLS) or flexi-access drawdown (FAD).

The two methods deliver the same gross result and treat the 25 per cent tax-free element differently. UFPLS pays 25 per cent tax-free and 75 per cent taxable on every withdrawal. FAD designates a portion of the pot, takes 25 per cent of the designated portion as the Pension Commencement Lump Sum (PCLS) tax-free, and leaves the rest in a drawdown wrapper to be taken as taxable income on whatever schedule the member chooses.

Visual D · The same £20,000 draw, two different shapes

	UFPLS — £20,000 withdrawn	FAD — £80,000 designated, full PCLS taken
Tax-free element this year	£5,00025 per cent of £20,000	£20,00025 per cent of £80,000 designated
Taxable element this year	£15,000added to income	£0this year; £60,000 stays in drawdown
Tax-free cash drawn so far	£5,000 of LSA used	£20,000 of LSA used
Position next year	Repeat the same shape	£60,000 available as taxable drawdown on any schedule

The choice matters because it shapes the timing of the tax-free element. Spreading it through UFPLS or staged FAD designations holds back some of the tax-free room for later years. For many households at Susan and Peter's wealth level, staging is more tax-efficient than a single front-loaded lump, because the tax-free slice on each withdrawal pulls a portion of taxable income out of higher-rate territory along the way. The right answer depends on what the household plans to do with the cash and on whether other tax thresholds — the personal allowance taper above £100,000, the higher-rate threshold itself — are in play.

The 2024 Budget introduced the Lump Sum Allowance, which caps the cumulative tax-free cash that can be drawn from registered pensions across a lifetime at £268,275. Many households with a pension pot above £900,000 will run into this cap if they take all available tax-free cash, which is itself an argument for staging.

The parallel question — whether Peter should commute defined benefit pension income into a tax-free lump at retirement — sits in §6, where it is treated alongside the annuity-versus-drawdown decision that has the same shape.

The 2027 change to pensions in the estate

Until April 2027, unused defined contribution pensions sit outside the estate for inheritance tax. The received wisdom — draw pensions last, draw ISAs and general investment accounts first — followed directly from that fact. Draining the inheritance-tax-free wrapper before the inheritance-tax-included wrappers was the structurally correct sequencing for households whose estate would face an inheritance tax charge.

From April 2027, unused defined contribution pensions enter the estate. The wisdom inverts.

Visual C · Sequencing logic before and after April 2027

	Pre-April 2027	Post-April 2027
DC pension at death	Outside estate; passes to beneficiaries free of IHT	Inside estate; subject to IHT at the marginal estate rate
Sequencing implication	Draw pension last; preserve the IHT-free wrapper	Draw pension earlier; pay income tax on the way out rather than IHT at the end
ISA position	Draw earlier; ISA sits in the estate, so use it before pension	Draw later or leave; the ISA's tax-free growth and tax-free withdrawals retain their own argument
GIA position	Draw before pension; the capital-gains uplift at death is a useful side-effect	The capital-gains uplift at death becomes more valuable in relative terms

For households at Susan and Peter's wealth level, the new logic favours drawing the defined contribution pension down faster, paying the income tax on the way out, and leaving the ISA wrapper in place. The reasoning is symmetrical: the ISA inside the estate now costs no more than the pension inside the estate, but the ISA's tax-free growth and tax-free withdrawals retain their own argument.

This section names the change. The substantive estate-side mechanics — the order in which assets fall into the calculation, the residence nil-rate band, the gifting decisions that sit alongside — are the work of §7.

A practitioner observation

Consider the typical household at this wealth level whose retirement plan was written before April 2024. Guaranteed income floor close to or above essential spending. A defined contribution pension of seven figures or close to it, sitting outside the estate under the rules of the time. Joint ISAs in the mid-hundreds of thousands. A sequencing plan inherited from an earlier adviser that placed the pension last on the grounds it sat outside the estate.

The Budget reshaped that logic. The household that revisited its sequencing — bringing pension withdrawals forward, leaving the ISAs in place, and accepting an income tax bill in the early years that the old plan had deferred indefinitely — undertook the kind of restructuring that, at this wealth level, can protect a six-figure share of the eventual inheritance from a tax change that the old plan would have walked the household straight into.

What the Budget changed was which sequencing was the good plan. The households that looked at the question early are in a better position now than those who left the inherited plan in place. The difference is not insight; it is simply having looked again when the rules changed.

A second household type sits at the other end of the same teaching point. The pension is small relative to total wealth — a SIPP at one or two hundred thousand against several million in non-pension wrappers — and the household does not need it to fund retirement. Here the framework's verdict on the pension changes. The pension is not an income vehicle to be sequenced against the other pots; it is a tax-free cash entitlement whose value grows with the pot, up to the £268,275 LSA cap.

The income-tax-efficiency case for letting the pot grow before crystallising is sharper for this household than for households whose pension is doing structural retirement-income work, because no income is foregone by waiting. The April 2027 change reshapes what happens to the residual after PCLS is taken — for a household with no income need from the pension, the structurally consistent answer is often to draw the residual down faster than instinct suggests, accepting the income tax along the way rather than leaving the wrapper to absorb the new IHT exposure.

A note for international readers

The framework above sits on UK tax assumptions: UK personal allowances, UK basic-rate bands, UK capital gains tax allowances, the UK 25 per cent tax-free pension cash, and the UK 2027 estate change. For readers already retired abroad, considering a return, or with a destination country in view, the lens framework still applies, but the inputs change. Treaty residence determines which country has the right to tax what; many destination countries (Spain among them) treat the UK 25 per cent tax-free pension cash as ordinary income rather than as tax-free; the order of withdrawal that is efficient in the UK can be inefficient or worse abroad.

The most consequential point for households in transition is the timing of major draws relative to a residence change. A tax-free pension lump sum taken while UK-resident may be taxed as ordinary income if drawn the year after the household becomes Spanish-resident; the same draw made twelve months earlier sits inside one regime instead of two.

The shape the household is arriving with carries its own version of the question. A returning expat who built household wealth overseas across a long international career — in non-UK portfolio bonds, in a GIA, in property — and who holds a comparatively small UK SIPP from earlier UK working years, faces a sequencing question that runs differently from the framework above. The pension is not funding retirement; the bond and GIA stack does that. The 25 per cent PCLS calculation applied to whatever value the pot reaches at the point of crystallisation becomes the live planning question, and the answer pulls toward deferral rather than toward early draw. The April 2027 estate change cuts the other way on the residual.

§8 covers the country-specific treatment in detail.

Closing the section

For each pot in the §2 template, write down which of the three layers it sits in — that is the input §4 picks up. Then, holding the lens framework in mind, identify one change the household could make to its current withdrawal pattern that better aligns the order of draw with the three lenses. The change does not have to be implemented this week. Naming it is the work this section asks for.

The next section takes the bucketing structure and develops it in full.

The wrapper question is exactly the territory where personalised modelling earns its fee. The lens framework points at the right shape of answer; the right number — for this household, in this tax year, against these specific bands and allowances — needs the actual figures on a spreadsheet. If the framework above has surfaced a question worth thinking through, this is the kind of decision that benefits from sitting down with a qualified financial planner who can see the household's actual numbers.

A particular worry sits in the background of every retirement in its early years: that a bad stretch of markets in the first few years could reshape the rest of it. Readers who have read deeply about pensions half-know this already. Readers who have not read deeply tend to sense the same thing in a less articulated form. This section works for both, and lands on the same structure.

The worry has a name — sequence of returns risk, usually shortened to sequence risk. The mechanics are mathematical rather than rhetorical: when an income is being drawn from an invested pot, the first five to seven years of returns carry more weight than any five to seven years that follow. The rest of the section is about what to do with that fact.

A note on the word "risk"

It is worth pausing on a distinction the industry has done a poor job of making. Risk and volatility are often used interchangeably, and the conflation does real damage. Volatility is the movement of a portfolio's value along the way — the up and down of any given quarter. Risk, in the precise sense that matters to a retired household, is the chance of failing to meet the objective the portfolio is there to support: paying a sustainable income for the rest of two lives. A portfolio that feels comfortable through a calm year may be poorly placed to meet that objective. A portfolio that moves around more in any given month may be far better placed.

The biggest risk a retiree faces is rarely the next bad quarter. It is the risk of running out of money before they run out of life — of being forced to cut spending later, or of taking losses with no time left to recover. The bucketing structure this section sets out is designed to manage volatility precisely so that the real risk does not materialise. The two concepts are not the same, and a household that treats them as the same has been measuring the wrong thing.

Why the first years weigh more

A pot that falls thirty percent in the year after the first withdrawal is in a materially worse position than a pot that falls thirty percent in year thirty, because a pound drawn from a depleted pot is a pound that cannot grow back. The effect compounds across the remaining horizon.

An example makes this plain. Two retirees each start with a pot of £1 million and draw £40,000 a year in real terms. Retiree A has a bad first year — the portfolio falls thirty percent — and then posts ordinary returns of around four percent real for the next twenty-nine years. Retiree B has twenty-nine years of those ordinary returns first, and the same thirty percent fall in year thirty. The market delivers the same thirty annual returns to both retirees, in a different sequence; the geometric mean return on the investments themselves is identical.

What the retirees live through is not identical at all. Retiree A runs out of money a year short of the thirty-year horizon. Retiree B takes the full thirty years of withdrawals and ends with a substantial pot intact. Same portfolio, same withdrawal plan, same thirty returns in a different sequence.

This pattern is not exotic; it is the ordinary arithmetic of drawing an income from an asset whose value fluctuates. The reason the early years weigh more is straightforward. That is when the pot is both largest in absolute terms and most exposed to being drawn down below the level from which it can recover within a retirement horizon. Well before the horizon closes, the pot has either proved itself capable of supporting the plan or has not; the outcome is, by then, already largely written.

The question is what to do about that fact. The answer is structural.

Bucketing: the structure many of my clients use

The structure I use with many clients at this stage of life organises retirement assets by when the money is needed, rather than by what type of asset it is. Three layers, doing three different jobs, sized to horizons that respect how equity markets actually behave.

The near-term layer holds one to two years of essential spending, in cash, money market funds, and short-dated gilts. Its job is to keep the household from being forced to sell growth assets at a bad moment. In a year where the invested portfolio has fallen, the household draws its income from the near-term layer and leaves the rest of the portfolio alone.

The mid-term layer covers the next two to five years of expected withdrawal, in a mix of fixed income and lower-volatility growth assets chosen to absorb ordinary year-on-year moves and broadly outlast a prolonged fall. Its job is to refill the near-term layer on a sensible cadence. In years when markets perform normally, the near-term layer is topped up from the mid-term layer so that it stays at one to two years' depth.

The long-term layer holds everything from year five onward: equities and other return-seeking assets. The five-year cut is not arbitrary. Equity markets tend to move in roughly five-year cycles — long enough to ride out a fall and a recovery, short enough that a household drawing for thirty years can plan around it. The long-term layer is the layer that carries the equity weighting a long retirement requires, and it is also the layer most exposed to volatility. Volatility here only bites the household if the household is forced to sell from it during a fall, and the other two layers exist to prevent that.

The segmentation is about function as much as allocation. A pot of six hundred thousand pounds can hold exactly the same assets under one mental model, "a diversified portfolio drawn down at a fixed rate", and work quite differently under another: "three layers, each sized to a horizon, refilled on a cadence." The difference shows up when markets fall. Under the first model the reader's lived experience is that everything is at risk. Under the second, only the long-term layer is down, and that layer is not being drawn from this year anyway. Retirement money has a timetable — a departure board, not a departure lounge.

Layer 1

Years 1–2

Near-term

What it holds

Cash, money market funds, short-dated gilts. One to two years of essential spending.

The job

Insulates the household from forced selling in a bad year. The annual draw comes from here.

Layer 2

Years 2–5

Mid-term

What it holds

Fixed income and lower-volatility growth, sized to two to five years of expected spending.

The job

Refills the near-term layer on a sensible cadence — tops up in normal years, pauses in falling ones.

Layer 3

Years 5+

Long-term

What it holds

Globally diversified equities and other return-seeking growth. The bulk of the capital.

The job

Grows across the long horizon. Volatility here only bites if the household is forced to sell during a fall — which the other two layers exist to prevent.

Dynamic withdrawal: the discipline that goes with the structure

Bucketing is the architecture. A small discipline sits alongside it.

In years where the long-term layer has fallen meaningfully, the refill from it into the mid-term layer is paused. The household continues drawing from the near-term layer, which is doing what it was put in place to do. Discretionary spending is trimmed modestly for the year (a holiday deferred, a kitchen refit pushed out), and the long-term layer is left untouched until its recovery begins.

The discipline is tolerable because the bucketing structure has already done most of the work. The decision to hold equities through a fall was made in advance, when the structure was put in place. The adjustment now is small. Without the structure, the same adjustment feels catastrophic, because the household is being asked to improvise a response under stress.

In the literature this kind of rule is sometimes called a Guyton-Klinger guardrail. The naming is fine; the principle is older and simpler than the name suggests. It is the ordinary discipline of a household that is paying attention to its own circumstances.

What this looks like at £1.68 million

Consider Susan and Peter, eighteen months into retirement, with drawdown assets of around £1.68 million, spread across a £920,000 DC pot in Peter's name, a £230,000 DC pot in Susan's, £380,000 of joint ISAs, and a £150,000 GIA left over from a house sale. Their defined benefit entitlement of £45,000 a year combined, plus one state pension now and the second in two years' time, provides an inflation-protected floor that already covers most of their essential spending. The household is drawing around £72,000 net, the larger part of which comes from the invested pots.

A bucketed arrangement looks like this. The near-term layer holds £80,000, held in a money market fund inside one of the ISAs and a short-dated gilt ladder in the other, just under two years of essential spending. The mid-term layer holds around £200,000 in a mix of fixed income and low-volatility growth, spread across the two ISAs and the GIA — roughly four years of essential spending at the household's current rate. The long-term layer holds the balance, around £1.4 million, in a diversified equity allocation across both DC pots.

Assume the twelve months after retirement deliver a thirty percent fall in global equities. Under a rigid four-percent draw from one undivided pot, the household's lived experience is that everything is at risk. They are drawing seventy-two thousand out of a portfolio that has lost three hundred thousand. And because most retirement money sits in unitised funds of the 60/40 shape, every pound withdrawn sells sixty pence of equity and forty pence of fixed income pro rata — so the £72,000 draw has liquidated roughly £43,000 of equity units at the low, and those units do not come back when markets recover. The question at the lunch table is whether they should have been more defensive, and whether they should become more defensive now.

Under the bucketed structure, the same household's lived experience is different. The long-term layer is down around £300,000; the mid-term layer, less volatile, is down a modest amount; the near-term layer is untouched. The household is drawing from the near-term layer this year, as planned, and refilling from the mid-term layer is paused for the year because the near-term layer does not need topping up. No equities are being sold. The question at the lunch table is whether to trim the summer holiday budget. Both households hold the same portfolio, take the same withdrawal, and sit under the same arithmetic. Their lived experience of the year is very different.

How the structure played out in March 2020

Consider two households at materially similar financial positions when the markets fell in March 2020. The first had a near-term layer already in cash, holding two years of essential spending outside the invested pots. When the fall came, the household drew from the cash that had been set aside. The invested portfolio was not touched. The fall in the long-term layer was real, but it registered as a number on a screen rather than a decision being forced under stress.

The second household, without that structure, de-risked across the whole portfolio at the wrong moment. The sale took a week. The repurchase never quite happened at the right time. By the time the markets had recovered, that household had locked in a permanent loss of around fifteen percent of its capital, an avoidable loss, entirely a product of having no structure in place to frame the bad news against.

Neither household was unusually foolish or unusually wise. One had a structure in place before it was needed and the other did not. That was the difference.

A note on my qualifications

The substance of this section sits on the Advanced (Ofqual Level 6) CII qualifications I hold in retirement income — AF8 (Retirement Income Options) and AF7 (Pension Transfer Specialist) — supported by my Diploma-level units J05 (Pension Income Options) and R04 (Pensions & Retirement Planning). Bucketing is the structure I use with many clients at this stage of life, and it is the structure I would want in place for my own family.

Before the next section

A decision made in advance is substantially easier to implement than one made in the middle of bad news. Before you move on, name the three layers for your own household. Write down what sits where, sized to roughly how many years of which kind of spending. Identify the single biggest gap between the current arrangement and a properly bucketed structure.

That piece of paper is the difference between a household that has a plan for a bad year and a household that has opinions about a bad year.

The next section takes the four percent rule, the single piece of received wisdom many retirees have absorbed and very few have read the research on, and shows how it becomes useful once it is understood as an input into the cadence of bucket refill, rather than a draw against an undivided pot.

A particular number has lodged itself in retirement conversations over the past three decades: the idea that a retiree can safely withdraw four percent of their portfolio in the first year, index that amount each year for inflation, and expect the money to last a thirty-year retirement. Many readers who have given serious thought to drawdown have encountered the number; few have read the research that produced it.

This section gives the reader that understanding. The four percent rule is useful in a particular and limited way: it is a first approximation, an anchor against which to judge whether a proposed draw is broadly sensible or broadly implausible. It works as a historical finding about a specific market carrying specific assumptions, rather than as a guarantee or a universal law. What matters more than the specific number is what the reader does with it, and the rule works best when it is understood as one input into the structure §4 introduced rather than as a rule governing an undivided pot.

Where the rule came from

The rule is a finding from two pieces of work published roughly a decade apart. William Bengen, an American financial planner, published a paper in 1994 asking a specific question: across all rolling thirty-year periods in the US stock and bond markets since 1926, what was the highest fixed real withdrawal rate that had not run the portfolio dry in any of those windows? His answer, after modelling portfolios weighted between US equities and intermediate-term government bonds, was four percent. A rigid four percent real draw, inflation-adjusted each year regardless of how the portfolio had performed, would have survived every thirty-year window in his data set, including the retirees who started drawing in 1929 and in 1973.

The mechanic is worth stating plainly, because it is more specific than the shorthand implies. In the first year of retirement the household draws four percent of the portfolio's value: at £1.68 million, that is £67,200. In the second year the household draws the same amount in real terms, which at two percent inflation is £68,544 in nominal pounds; in the third year, with further inflation, a little higher again. The draw does not respond to the portfolio's performance. It is set in year one and indexed for inflation thereafter, and the question the research asked was whether that draw had run the portfolio dry by year thirty.

The Trinity Study, published by three finance professors in 1998 and updated since, asked the same question from a slightly different angle and reached broadly the same landing: a rigid four percent real draw from a portfolio with a material equity weighting had a very high survival rate across historical US data. Both studies were careful and specific. Both reached a number that became, in the telling, more definite than the studies themselves claimed.

The assumptions, and how they land for a UK retiree

The rule carries five assumptions that matter.

The first is the data. US equities and US intermediate government bonds, from 1926 onwards. The intuitive objection is that a UK retiree's portfolio must be materially different, and therefore that the finding must be less portable. On the equity side the objection is shakier than it looks: a globally diversified equity allocation today is typically sixty to seventy percent US by market weight, which leaves the equity distribution closer to Bengen's than the headline geography suggests. The bond side is usually different in shape from intermediate US Treasuries, and that matters at the margin. The shift widens the distribution around the four percent figure rather than moving the centre of it, and portfolios drifting further from a diversified shape, whether through heavier UK home bias, narrower concentrations, or smaller fixed-income sleeves, widen the distribution further.

The second is the horizon. Thirty years. For a household at Susan and Peter's age, with family longevity patterns suggesting a reasonable chance that one of them lives past ninety, the honest planning horizon is closer to thirty-five. A four percent rate is a little more conservative against a longer horizon, not catastrophically so but not neutral to the assumption.

The third is the portfolio mix. Bengen modelled specific equity/bond splits; Trinity explored a range. The rule's survival rate is meaningfully better at fifty percent equities than at twenty-five, and falls off again at very high equity weights for reasons sequence risk makes plain.

The fourth is the withdrawal discipline. The rule models a rigid real withdrawal, indexed for inflation each year regardless of what markets did the previous year. That is not a rule a sensible household would actually follow through a material fall, and households that have come to grief against the rule have usually been following it too literally.

The fifth is the pot structure. The rule is a rule about what a household takes out of one undivided portfolio each year. It does not contemplate the time-segmented structure §4 introduced, and under that structure the rule's work looks different.

For a UK retiree in 2026, the rule lands as a useful starting anchor: a little conservative for a diversified portfolio and a thirty-five-year horizon in competent hands, a little generous for the same horizon if the withdrawal discipline is rigid and the structure is an undivided pot. The question is how to use the number well.

Inside the bucketing structure

The reframe is the reason the rule becomes useful rather than dangerous. Under the original framing, four percent is a rule about what the household takes out of one undivided pot each year. Under the structure §4 put in place, four percent is not doing that work. It is the cadence at which the near-term layer is drawn down each year and refilled from the mid-term layer when markets are behaving normally. The household's draw for the year comes from the near-term layer, the refill runs on a cadence the rate helps to govern, and the long-term layer is not being drawn on at all in a given year.

Sequence risk is absorbed by the structure, not by the rate. A bad year in the long-term layer does not threaten this year's draw, because the draw is not coming from the long-term layer; it threatens the cadence at which the mid-term layer refills the near-term layer, which is a quite different problem and a substantially more manageable one. The rate becomes an input into refill cadence rather than a statement about what the household can afford to take out of one pot.

The practical consequence is that the same four percent figure can be held steady through a bad year without imperilling the household's lived experience, because the household is not drawing on the falling part of the portfolio.

Guardrails: the discipline that replaces rigidity

The rule as originally written asks the household to index the draw every year regardless of what the portfolio has done. A small set of rules, written down before the bad news arrives, replaces that rigidity with a discipline that is substantially easier to honour.

Two rules carry the work. The first says that in a year following a material fall in the portfolio, conventionally understood as a fall of more than around ten percent, the household holds the annual draw flat in cash terms rather than applying the usual inflation uplift. Essential spending is protected, because the guaranteed income floor and the near-term layer are doing their jobs; discretionary spending absorbs the real-terms erosion for a year or two. The second says that in a year following an exceptional rise, conventionally around twenty-five percent above the pre-committed portfolio trajectory, the household can take the inflation uplift plus a modest step up, giving the good years their due and preventing the gradual under-consumption that can otherwise set in. Between those thresholds, ordinary inflation indexing applies.

The specific numbers are illustrative; different advisers use narrower or wider bands. What matters is that the rule lives on a piece of paper drafted when the household is calm, and therefore survives the market stress that would otherwise rewrite the rule in real time. In the literature this kind of structure is sometimes called a Guyton-Klinger guardrail, after the two American advisers whose 2006 paper named it. The principle is older and simpler than the name suggests.

Three approaches compared

The rule is one of three approaches worth holding in mind. The other two are variations on it: different ways of using the same broad arithmetic with a different treatment of the response to markets. The chart below shows the three approaches running through one shared market scenario; the table that follows captures the trade-offs as the household would weigh them.

Visual G · Three approaches through one market scenario

Rigid 4% plus inflation Guardrails (Guyton-Klinger style) Percentage of portfolio

Three retirees, three rules, one shared market. The rigid 4% household keeps indexing its draw for inflation regardless of what the portfolio is doing — by year ten its annual withdrawal has risen towards £90,000, even though the portfolio fell hard in year two. The guardrail household holds its draw flat through years three and four, resumes indexing in year five, and ends the decade drawing meaningfully less. The percentage-of-portfolio household cuts its draw the moment the market falls and steps it back up as the market recovers: the steadiest portfolio outcome but the bumpiest annual income.

Approach	How the draw responds to markets	Behaviour in a bad year	Behaviour in a good year	Pre-commitment	Long-horizon resilience
Rigid 4% plus inflation	Does not respond; the draw is set in year one and indexed thereafter	Continues to rise with inflation regardless of portfolio performance	Indexes with inflation only	Low; real adherence through a fall is psychologically hard	Depends on the sequence of returns the household encounters
Guardrails (Guyton-Klinger style)	Responds to two pre-committed thresholds above and below	Held flat for the year; discretionary spending absorbs the erosion	Indexes and may step up after an exceptional rise	High; the rule is written down in advance	Absorbs sequence risk through the rules; tends to outlast the rigid approach
Percentage of portfolio	Set each year as a percentage of that year's portfolio value	Falls with the portfolio	Rises with the portfolio	Mechanical; no judgment required	Absorbs sequence risk through the volatility; exposes the household to consumption volatility

The three approaches are three different answers to the same question: what should the relationship be between the draw and the portfolio's actual behaviour?

What this looks like at £1.68 million

Consider Susan and Peter, introduced in §2 and seen again in §4. Invested pots of around £1.68 million, a guaranteed income floor of roughly £68,000 a year once both state pensions are in payment, essential spending of around £44,000, a total current draw of around £72,000 net of tax. The question is what draw rate and what rule govern the invested pots' contribution.

Under a rigid four percent approach, the household draws £67,200 a year in real terms, indexed each year for inflation regardless of how the invested pots are performing. In ordinary years this works; in a year following a thirty percent fall the draw is still £67,200, indexed again the year after. Whether the arithmetic holds over the full horizon depends on where the fall lands in the sequence.

Under a guardrail approach, the household starts at £67,200 and writes two rules down. In a year after a material fall the draw is held flat in cash terms, discretionary spending is trimmed, and essential spending is undisturbed because the guaranteed floor is covering it. In a year after an exceptional rise the draw is indexed and may be stepped up modestly. In ordinary years the draw indexes normally. Five years in, the cumulative draw sits within ordinary indexing distance of the rigid approach; the portfolio's trajectory through a bad patch is markedly healthier.

Under a percentage-of-portfolio approach, the household draws four percent of each year's portfolio value. In year one that is £67,200; in year two, after a thirty percent fall, it would be closer to £47,000; in year six, after a strong recovery, it could be £78,000. The mechanical responsiveness is the point; the consumption volatility is the cost.

For the Susan-and-Peter shape, the guardrail approach fits best. The guaranteed floor means a flat-held year on the invested pots does not threaten essential spending, and Peter's temperament responds more easily to a rule he has written down than to a formula that produces a lower number in a bad year. Each of the three is a way of using the four percent anchor; each produces a different cadence of refill on the bucketing layers the household already has in place. The draw rate and the bucketing cadence are the same decision viewed through two lenses.

A pattern from 2020 to 2022

Consider two households at materially similar financial positions heading into 2020. Both with retirement pots around the £1.5 million mark, both with broadly similar essential and discretionary spending, both drawing in the neighbourhood of the four percent anchor.

The first household held to the rule rigidly. It drew its inflation-indexed amount through the fall in early 2020 and applied the full inflation uplift in 2021 and 2022 as the rule requires, at a time when inflation was running hard. By the end of 2022 the portfolio was visibly down from where a disciplined approach would have left it: not ruinously, but enough to weaken the arithmetic over the full retirement horizon. The rigidity was the cost.

The second household had a guardrail rule drafted when the retirement was first structured. It drew the same amount in 2019, held the draw flat through 2020 and 2021 while the portfolio recovered, and resumed indexing in 2022 once the pre-fall trajectory was within reach. Five years in, the portfolio was close to where it would have been without the fall at all. Discretionary spending had absorbed a couple of flat years; essential spending had never been threatened, because the guaranteed floor was doing the work.

Both households held the same portfolio. Both received the same sequence of returns. The difference was whether a piece of paper with a rule on it had been drafted before the news arrived or was being improvised afterwards.

Before the next section

Pick a withdrawal rate and a rule for adjusting it, not a rate alone. The rate is the anchor; the rule is the discipline that lets the anchor hold when markets do the unusual thing markets reliably do every few years. Write the rule down, and put it with the household's other retirement paperwork so it can be found in the year it is needed.

The next section takes up the question of whether any part of the invested pots' draw should come from guaranteed income bought from an insurer: the annuity question that many retirees hold as a binary and that the rest of the guide handles as a component decision inside the bucketing architecture already in place.

Most retirees never explicitly decide whether annuitisation has a place in their retirement arrangement. The state pension arrives. Any defined benefit entitlement arrives. The DC pot sits in drawdown, or is moved into drawdown from a workplace scheme without much conversation. The guaranteed-income floor ends up being whatever state pension and DB happen to combine to, and the question of whether to extend that floor by buying guaranteed income in the market goes unasked. Readers who have not made the decision explicitly have made it by default. The default is not always the right answer for the household in question, and it is not always the wrong one. This section is about making the decision deliberately rather than drifting into it.

The question inside the §4 structure

§4 introduced bucketing: retirement assets organised by when the money is needed, in three layers. The near-term layer holds one to two years of essential spending, in cash, money market funds, and short-dated gilts, and its job is to insulate the household from forced selling in a bad year. The mid-term layer covers the next two to five years and refills the near-term on a sensible cadence. The long-term layer holds everything from year five onward, carrying the equity weighting a long retirement requires across a horizon long enough to absorb volatility.

Bucketing does much of the income-floor work that annuitisation is sometimes proposed to do. The near-term and mid-term buckets, properly sized, provide the household with predictable income across the two-to-ten-year horizon where sequence risk actually bites. An annuity extends the floor further — guarantees it for life — in exchange for surrendering the underlying capital.

The practical question for this section is whether extending the floor earns its place in the specific shape of the reader's household.

For many households, and Susan and Peter's shape is the common case at this wealth level, the bucketing structure performs the floor function without requiring annuitisation at all. Combined state pension and DB entitlement already exceed essential spending. The invested pots fund discretionary spending at a draw rate the bucketing cadence can hold. The near-term layer provides the insulation against bad years that a structural architecture is designed to provide. Adding guaranteed income on top surrenders capital that the household is already using productively.

For many other households, DB entitlement is modest or absent and state pension is the only guaranteed stream. Bucketing carries these arrangements too, in my practice and in many others, provided the household's essential spending sits within a sustainable draw against the invested pots once state pension is added in. The invested pots have to do more of the lifting in this shape, and a well-sized near-term layer is what makes the arrangement safe to live in across a bad year. The absence of a DB pension is not, on its own, an argument for annuitisation. The §4 structure does the floor work a household needs in plenty of arrangements where state pension is the only guaranteed line on the income statement.

For a narrower set of households, the question becomes more live. It is rarely any single factor that moves it. A combination tends to be at work: essential spending sitting close to what the invested pots can safely sustain as a draw on top of state pension; a family pattern or medical history suggesting a thirty-year or thirty-five-year horizon; a household temperament for which market volatility is genuinely corrosive to daily life rather than passingly uncomfortable. When several of these factors coincide, the annuitisation question moves from "not for this household" to "worth thinking about seriously".

The question running through the section is whether annuitisation adds something the bucketing structure is not already providing for the specific shape of this household, and if so, how much. Assess against household shape rather than default in either direction.

How annuities work

An annuity is a contract. The household surrenders a lump sum of capital to an insurer and in return receives a guaranteed income stream, at a rate the insurer sets at purchase and holds for the life of the contract. The income is guaranteed. The capital is no longer the household's.

The cost of that guarantee is threefold. The capital itself is gone, no longer available to the household for emergencies, large purchases, gifts to Ben and Anna, or estate value. The growth the capital would have earned over the retirement horizon is gone, no longer compounding inside an ISA, a GIA, or a DC pot. The estate value at death, for a conventional lifetime annuity with no guarantee period, drops substantially, because the insurer's income obligation ends with the annuitant.

The rate environment matters. Annuity rates are driven largely by gilt yields at the moment of purchase. Through most of 2010 to 2022, gilt yields were unusually low and annuity rates looked correspondingly poor. A household considering annuitisation today is being offered a different deal to the one their parents or older siblings turned down a decade ago. Rates move, and today's rate is not a permanent fixture, which is an argument for care in timing rather than for rushing.

Three structural options are worth understanding before any decision. A conventional lifetime annuity surrenders capital permanently in return for income guaranteed for life. A fixed-term annuity surrenders capital for a set number of years, typically five to ten, with guaranteed income through the term and the contract reverting to drawdown or a new annuity at the end. A deferred annuity surrenders capital now in return for guaranteed income that begins at a later date, often timed to cover later-life costs when longevity risk has become more material and the household's other income sources may have thinned.

The guide describes annuitisation as a category without advocating for or against. The restraint reflects practitioner caution: annuities are a useful tool in the right shape of household and the wrong tool in the wrong shape, and the section's job is to help the reader see which shape theirs is.

A parallel decision sits in front of members of defined benefit schemes at retirement: whether to commute part of the annual pension into a tax-free lump sum, a trade that runs in the opposite direction to annuitisation (guaranteed income surrendered for capital rather than the other way round) but sits under the same shape-of-household framework this section lays out.

The three options side by side

Option	Income guarantee	Inflation protection	Capital flexibility	Death benefits	Typical income rate per £100,000	Complexity
Conventional lifetime	For life	Available at additional cost	None; capital surrendered	Limited; optional guarantee period or joint-life	Roughly £7,000–£7,800 per year at age 65 in current rates (single life, level, healthy applicant)	Low
Fixed-term (5–10 years)	For the term	Available at additional cost	At term end, capital returns to a choice of drawdown or new annuity	Unused capital returns to estate	Lower than a lifetime annuity when a meaningful maturity value is retained; the larger the capital-back-at-end, the lower the during-term income	Medium
Deferred (income begins later)	For life, from the deferred start	Available at additional cost	Very limited	Typically limited	Highest rate per £ of capital once income begins; nothing in the deferral years	High

Rates indicative as at May 2026; annuity rates move with gilt yields and the figures above are guidance for proportion, not for shopping.

Read across the table carefully. The fixed-term row can look like the best of both worlds — guaranteed income for the term and capital returning at the end — but the pricing reflects the trade. Insurers charge for the maturity value by paying a lower income rate during the term. The larger the amount of capital scheduled to return at the end, the lower the during-term income. That is why many households who choose to annuitise at all settle on a conventional lifetime contract: it gives the highest income rate per £ of capital surrendered, in exchange for surrendering the capital itself permanently. Fixed-term and deferred contracts are useful in specific shapes of arrangement, not as defaults. Rates are illustrative of current market conditions and move with gilt yields; the figures here are guidance for proportion, not for shopping.

What the decision looks like at £1.68 million

Take Susan and Peter, introduced in §2 and seen again in §4. Drawdown assets of around £1.68 million, spread across the two DC pots, the joint ISAs, and a GIA left over from a house sale. The guaranteed income floor is in the process of assembling. Combined state pensions will reach around £23,000 once both are in payment. DB entitlement runs at £45,000 combined. The floor settles at £68,000 once the second state pension starts in two years' time.

Essential spending runs at around £44,000 a year, covering utilities, council tax, food, insurance, motoring, basic maintenance, and the modest non-negotiable travel two retired adults treat as part of ordinary life. Discretionary spending adds another £28,000: holidays, the house projects deferred through the working years, regular gifts to Ben and Anna, the wider social life two retired professionals keep up. Total household draw: around £72,000 net.

Read the arithmetic carefully. The household's guaranteed floor (£68,000) covers essential spending (£44,000) with roughly £24,000 of headroom. That £24,000 of headroom is available to deploy into discretionary spending before the invested pots are touched at all. Discretionary spending runs at £28,000, which means the guaranteed stream funds £24,000 of it and the invested pots fund the remaining £4,000, plus the small tax cost of drawing from a pension to meet that gap. The effective draw on the £1.68 million portfolio sits at around one percent a year, well inside a sustainable range, and the bucketing structure of §4 carries that comfortably.

Put the two halves together and the picture is straightforward. Guaranteed income almost covers total planned spending on its own. The invested pots are funding a thin residual slice of discretionary and providing the flexibility, capital, growth, and estate value that a long retirement needs. There is no floor gap to close.

With the picture clear, three routes can be compared.

Route A: bucketing alone. The current arrangement. Guaranteed floor covers essential with headroom and most of discretionary; invested pots fund the thin residual through the bucketed structure; near-term layer insulates against bad years. Capital, growth, estate value, and discretionary flexibility are all retained.

Route B: partial annuitisation of £200,000 of the larger DC pot. At current rates, that roughly adds £12,000 of additional guaranteed annual income. The guaranteed floor rises from £68,000 to around £80,000, which now sits above total planned spending of £72,000 by around £8,000 a year. The invested pots are no longer drawn against for ordinary outgoings. £200,000 of capital is surrendered to generate guaranteed income beyond what the household is actually spending, and £200,000 of future growth and estate value goes with it.

Route C: a smaller annuitisation combined with a bucketed arrangement. A £100,000 annuitisation, adding around £6,000 of guaranteed income and removing £100,000 of capital. The guaranteed floor rises to around £74,000, marginally above total planned spending, and the pot draw falls to nothing for ordinary outgoings. Half the trade-offs of Route B, the same structural question at smaller scale.

No recommendation lands in the worked example; the reader is helped to see what each route actually does. For a household whose guaranteed income already covers essential spending with £24,000 of headroom and funds most of discretionary as well, the real question is whether to convert flexibility the household already has into guarantee the household does not need. Read from the numbers, Route A has the strongest case for this household.

Typical shapes of household

Annuitisation earns its place in some shapes of household and does not earn it in others. Three patterns make the distinction clear.

The DB-supported shape where bucketing renders annuitisation unnecessary. The Susan-and-Peter shape. Combined state pensions and DB already comfortably exceed essential spending. Invested pots fund discretionary rather than essential. Draw rate on invested pots is well below four percent. The bucketing structure of §4 carries the arrangement without strain. In this shape, annuitising surrenders capital for guaranteed income the household's floor does not need.

The no-DB shape where bucketing also renders annuitisation unnecessary. A household where state pension is the only guaranteed stream and DB entitlement is absent or negligible. The invested pots are doing more work, funding both essential and discretionary above the state pension line. Provided the total draw rate on the invested pots is sustainable and the near-term layer is sized to one or two years of drawdown, the bucketing structure carries the arrangement in the same way it carries the DB-supported shape. In my practice, a great many households of this kind run retirement comfortably on a well-structured bucketed drawdown with no annuity in the picture. The absence of DB on its own is not a reason to annuitise.

The shape where partial annuitisation genuinely earns its place. A narrower case, and it tends to arrive as a combination rather than a single factor. A single-person household, often widowed. A smaller DB entitlement of perhaps £12,000 to £18,000 a year, or none at all. The bulk of the retirement capital sits in a DC pot of several hundred thousand, and essential spending runs close to what that pot can sustain as a draw on top of state pension. The family pattern suggests longevity into the nineties. The household's temperament is anxious about markets in a way that corrodes the quality of retirement on its own. For this shape, a partial annuitisation of a portion of the DC pot closes the floor gap and removes the weekly FTSE-checking that would otherwise wear the household down. The remainder of the DC runs as a bucketed drawdown. The structural argument for annuitisation is genuine, and the section's framework supports it. The same framework would decline annuitisation for any of these factors in isolation.

In most arrangements at this wealth level, whether or not a DB pension is part of the picture, the bucketing structure performs the floor function and annuitisation does not earn its place. That is said here once, without hedging, and without closing the door on annuitisation as a category. The third shape above shows exactly where it does earn its place.

There is also a wider set of households whose decision sits outside the structural argument entirely. The comfort of a guaranteed income line that the household neither has to manage nor has to watch the market move against is, for some retirees, worth more than the capital and growth surrendered to obtain it. The §4 bucketing structure can carry the arrangement on the numbers; the household may nevertheless prefer not to live with the management responsibility, even at a partial annuitisation. The framework above does not argue them out of that choice. It asks only that the choice be made with the trade-offs named clearly rather than absorbed by default.

Before the next section

The practical action for the reader, before moving on, is to draw the household's own version of the floor-versus-essential picture.

On a piece of paper, write down the household's essential spending for a typical year. Below it, write the combined guaranteed income already in place once state pension and any DB are fully in payment. Below that, write the income the near-term bucket could sustainably provide through the §4 refill cadence. Add the second and third lines together, and compare against the first.

If the combined floor comfortably exceeds essential spending, the annuity question is usually not live for the household, and Route A is the starting point. If state pension is the only guaranteed stream and the invested pots are doing most of the work, the annuity question is still not automatically live; the bucketing structure of §4 carries many such arrangements perfectly well. It is when essential spending runs close to the sustainable draw rate, and a longevity factor or a market-anxious temperament coincides with it, that the question moves to the foreground. In any shape where the picture is unclear, the decision warrants proper practitioner input — annuitisation is one tool among several, and the conversation about when it earns its place is exactly the conversation a practitioner is there to have.

This is a decision readers rarely make well alone. The choice is permanent, the capital surrender is irreversible, and the shape-of-household question is exactly what a qualified financial planner can help a reader assess honestly. For any reader finishing this section unsure which shape their household is, a personalised conversation with a planner who can see the actual numbers is the natural next step.

The most expensive mistake in estate planning is not bad planning; it is no planning at all. Many households arrive at retirement having read about inheritance tax for years and never having done anything about it. The topic is uncomfortable, the arithmetic is unforgiving, and the postponement is where the avoidable damage gets done. Susan and Peter are typical. Their wills were last updated when the children were small. They have never sat down with the household's combined estate position on a single sheet of paper, looked at the bill that would be due at current numbers, and worked through the structural responses available to them. The job of this section is to make that conversation easier to have by giving the household the structural picture and the names of the planning patterns.

The IHT architecture

The mechanics, named once and clearly. The Nil-Rate Band is £325,000 per person, frozen since 2009 and currently scheduled to remain frozen to April 2030. The Residence Nil-Rate Band is up to £175,000 per person where the main residence passes to direct descendants, with a taper of £1 of allowance lost for every £2 of estate above £2 million. Spousal transfer is unrestricted between two long-term residents — the surviving spouse inherits free of IHT and inherits the unused allowances of the first to die, so a couple with a clean transfer can shelter up to £1 million between them. The marginal estate rate above the allowances is 40 per cent. The frozen thresholds are a real problem for households at the wealth level this guide is written for; what was a comfortable allowance a decade ago is now exposing many households to charges that earlier generations would not have seen at the same real wealth level. The honest description belongs here; the campaigning belongs in the column.

The 2027 pension change

The substance §3 forward-referenced. From April 2027, unused defined contribution pensions enter the estate for inheritance tax purposes. The change inverts the old wisdom that placed pensions last in the withdrawal sequence on the grounds that they sat outside the estate. From April 2027, drawing pensions last leaves the unused balance inside the estate at death and exposes it to IHT at the marginal estate rate.

The interaction with income tax also matters. Under current rules, beneficiaries who inherit a defined contribution pension take it free of income tax if death occurs before the holder's 75th birthday and at the beneficiary's marginal rate thereafter. The post-April 2027 IHT layer sits on top of whatever income tax position applies, which means a death after age 75 can trigger both an estate IHT charge and an income tax charge on the beneficiary's eventual draws. The combined effect at the wealth level this guide addresses can be material.

The structural response, for many households, is to restructure the withdrawal sequencing — drawing the defined contribution pension faster than instinct suggests, paying the income tax on the way out across a number of years rather than leaving the unused balance to take the new IHT charge at death. The detail of the post-April 2027 mechanics is still being finalised in secondary legislation as this guide is written, and the treatment described here should be taken as the change as legislated rather than as a settled set of operational rules.

What this looks like at Susan and Peter's wealth level

Their combined estate at death assembles as follows. Investable capital £1.68 million, primary residence £1.4 million, total £3.08 million before any pension treatment. Combined allowances start at £1 million on a clean spousal transfer — two NRBs of £325,000 each plus two RNRBs of £175,000 each — but the RNRB taper bites above the £2 million threshold. With an estate of £3.08 million, the excess over £2 million is £1.08 million, which removes £540,000 of RNRB. The household's gross RNRB is only £350,000 between them, so all of it tapers away. The effective allowance is £650,000 of NRB only. The taxable estate is £2.43 million, the IHT bill is approximately £972,000, payable on the second death.

That is the picture under the pre-2027 regime. Under the post-April 2027 regime, the £1.15 million of combined defined contribution pension capital sits inside the estate rather than outside it. The taxable estate moves from £2.43 million to £3.58 million. The IHT bill moves from approximately £972,000 to approximately £1.43 million, a difference of around £460,000 between doing nothing and restructuring the household's withdrawal pattern in the years before. The numbers are illustrative of the mechanics rather than a recommendation for this household; the actual planning sits with a qualified adviser working from the household's specific position.

Lifetime gifting strategies

Gifting is the most flexible lever a household has during life. Several allowances run in parallel each tax year. The annual exemption is £3,000 per person, with one year of unused allowance available to carry forward. The small gifts allowance is £250 per recipient per year, with no cap on the number of recipients. Wedding gifts are exempt up to £5,000 from each parent and £2,500 from each grandparent. Gifts above these allowances are Potentially Exempt Transfers, falling out of the estate after seven years, with taper relief on the IHT charge if death occurs in years three to seven of the seven-year window.

The most under-used lever in this set is the gifts-out-of-normal-income exemption. Gifts made from the household's regular surplus income are immediately exempt from IHT, with no seven-year wait, provided the income is genuinely regular, the gifts are habitual, and the household's standard of living is not affected by making them. The exemption is precise about evidence: the household needs to document the income, the gifts, and the standard-of-living point in a way that survives scrutiny on the second death. Households that use the exemption properly often build a five-year programme that combines annual exemptions, regular income gifts to children and grandchildren, and one-off PETs above those amounts, taking material value out of the estate while the household is still alive to see what it does.

Trusts

Two patterns come up most often in estate planning at this wealth level. A loan trust takes a sum of capital from the donor and lends it to the trust; the loan remains repayable to the donor at any time, but the growth on the lent capital sits outside the estate from day one. A discounted gift trust takes a portion of a gifted sum as immediately outside the estate (the discount, calculated by reference to the donor's age and life expectancy at the time of the gift), with the remainder treated as a Potentially Exempt Transfer falling out of the estate after seven years; in return, the donor receives a fixed regular payment from the trust for life. Bare trusts hold assets for a specific named beneficiary, transparent to that beneficiary's tax position. Discretionary trusts hold assets for a class of beneficiaries with the trustees deciding how and when to distribute. The patterns are named here; the implementation of any of them sits squarely in regulated-advice territory because trusts of this kind are typically established using onshore or offshore investment bonds, and the choice of structure, the choice of wrapper, and the trustee arrangements all carry their own technical and tax consequences. Trusts of this kind are properly set up under professional advice.

Insuring the IHT bill, written in trust

A related pattern that some households execute alongside the gifting and trust work is a life policy written specifically to cover the anticipated IHT liability on the second death. The policy must be placed in trust at outset. Without the trust, the policy proceeds fall into the estate on death and are themselves charged to IHT at 40 per cent, defeating the policy's purpose. Written in trust, the proceeds pay out outside the estate, providing the executors with the liquidity to settle the IHT bill at the moment it falls due rather than forcing the sale of property or other estate assets under time pressure. The pattern is straightforward technically and surprisingly often missed in households where the rest of the estate planning has been done properly.

Estate planning sits on top of the bucketing structure

Worth holding the estate plan and the bucketing structure of §4 alongside each other rather than as separate exercises. The long-term layer (years five and beyond) is the layer most likely to remain at the household's death. It is also therefore the layer that estate planning most directly affects. A pension restructuring driven by the 2027 change is, viewed through the bucketing lens, a question about how fast the long-term layer is drawn down across the household's horizon. The two frames describe the same set of decisions from different angles — same household, same pots, same set of structural choices.

Wills and powers of attorney

The two pieces of paperwork that many households put off, and that many households later wish they had not. Wills should be revisited at every material life event — marriage, divorce, the death of a beneficiary, a significant change in the asset position, a change of preferred executor. Beneficiary designations on pensions and life policies sit separately from the will and need their own update discipline; an out-of-date beneficiary designation on a pension can override what the will says about who inherits what.

Lasting Powers of Attorney run on a separate track. There are two types, both worth registering: one for property and financial affairs, one for health and welfare. Registration costs around £82 per LPA in England and Wales as of 2026 and takes approximately 8 to 10 weeks once submitted. The critical point is timing. The LPA must be registered while the donor still has capacity to grant it; once incapacity is established, the LPA route closes and the family has to apply through the Court of Protection at substantially greater cost and delay. Many households leave this for too long and discover the option has gone.

Consider two households at materially similar positions, both in their late seventies, one of whom suffers a stroke. The household whose LPAs were registered ten years earlier moves seamlessly into a position where the spouse and adult children can manage finances and make health decisions on the donor's behalf within days. The household without the LPAs spends the next six months and several thousand pounds going through the Court of Protection, while the day-to-day arrangements are made awkwardly through holding accounts and family workarounds. The structural difference is not which household had better luck; it is which household had registered the paperwork before the news arrived.

Residency, the spousal exemption, and what changed in 2025

The 2025 reforms reshaped the IHT picture for households with cross-border histories. From 6 April 2025 the UK abolished domicile as the test for IHT scope and replaced it with a long-term resident test. An individual is a long-term resident for IHT purposes once they have been UK resident in at least 10 of the last 20 tax years immediately preceding the chargeable event. A long-term resident is within UK IHT scope on their worldwide assets. A non-long-term-resident is within scope only on UK-situated assets — UK property, UK-listed shares, UK bank accounts.

The tail period catches departures. A long-term resident who leaves the UK remains within scope for between three and ten years depending on prior length of residence. The minimum tail of three years applies at 10 to 13 years of prior residence; the tail then increases by one year for each additional year of residence, to a maximum of ten years at 20 or more years of prior residence. After 10 consecutive tax years of non-residence the long-term resident clock resets entirely, and a household that returns to the UK starts the 10-of-20 count again from scratch.

The spousal exemption mechanics under the long-term resident regime mirror the shape of the old domicile rules. Transfers between two long-term residents are unlimited under the standard spousal exemption. Transfers from a long-term resident to a non-long-term-resident spouse are capped at £325,000 cumulatively across lifetime gifts and death transfers. Transfers in the reverse direction — from a non-long-term-resident to a long-term resident — are unlimited. The non-long-term-resident spouse can elect to be treated as a long-term resident to access the full unlimited exemption. The trade is that the electing spouse's own worldwide assets come into UK IHT scope from the date of election, and the election runs until she has been non-UK resident for 10 consecutive tax years.

Consider the household where the position bites hardest. A British man, UK resident throughout his career and therefore a long-term resident, worth several million pounds, married to a spouse who has been in the UK for fewer than 10 of the last 20 tax years and is therefore not a long-term resident. He dies first. The spousal exemption to the wife is capped at £325,000. The bulk of the estate passes through IHT on the first death rather than rolling over to the second. The wife's election to be treated as a long-term resident is a substantial structural decision — it removes the cap and restores the unlimited exemption, but at the cost of pulling her own worldwide assets into the UK IHT net. Whether the election is the right call depends entirely on the size and location of her own pre-existing wealth, and on her intentions about staying in or leaving the UK across the decade ahead.

The mirror case is the long-term-abroad British household. A British national who has been non-UK resident for more than 10 consecutive tax years has reset the long-term resident clock. They are a non-long-term-resident for IHT purposes. Their non-UK assets are outside UK IHT scope. Their UK-situated assets — UK property, UK-listed investments, UK bank accounts — remain within scope on the same basis as before. If the household returns to the UK, the 10-of-20 count starts again, and there is a window of up to 10 years before the worldwide net closes around them again. §8 covers the country-side counterpart of these arrangements.

A note for international readers

The country-side counterpart of the residency mechanics. Whether the destination country has its own inheritance tax regime varies considerably — Spain levies Impuesto sobre Sucesiones with regional variation that makes the position genuinely complex; France levies droits de succession at progressive rates with allowances varying by relationship to the deceased; Portugal and Cyprus have no inheritance tax in the UK sense at all. The UK has very few estate-tax treaties (a small handful, most signed before 1980), which means UK-situated assets held by a household resident abroad can face IHT in both the UK and the destination country at the same death. §8 covers the country-specific treatment in detail and walks through the practical positions for the most common destination countries.

Before the next section

The first action this section asks for, before any structured estate conversation, is to book the will update appointment. Without an updated will, the rest of the planning runs against documents that may no longer reflect the household's intent. The second is to register the Lasting Powers of Attorney if they are not already in place. Both are floor items; everything else this section has covered builds on top of them. Estate planning of any meaningful structural kind sits with qualified professional input — the IHT architecture, the 2027 pension change, the residency reforms, and the trust patterns are all areas where the actual decisions need a planner, a tax adviser, and often a solicitor working together from the household's specific position. The job of this section has been to give the household the picture it needs to start that conversation well-prepared.

The international material is core to this guide rather than an appendix to it. Three shapes recur most often in the practice and they cover most of what arrives on the desk. The UK retiree who has already moved abroad — David and Helen, five years into Spanish residency, asking whether their current arrangement is still the right one. The UK retiree considering a move — Susan and Peter, who have looked at Portugal on and off for three years and never quite settled the question. The retiree with family abroad who is being drawn toward joining them — a daughter in France, a son in Lisbon, a small grandchild changing the calculation. The shapes are different from one another in the lived sense; the underlying decision architecture is the same.

The three layers

International retirement decisions sit on three layers and are easiest to think about when the layers are kept distinct. The first layer is residency. UK tax residence is determined by the Statutory Residence Test, a set of day-counting and tie-breaking rules that decides whether the retiree is UK-resident for tax purposes in any given tax year. Residence governs whether UK tax applies to worldwide income or only to UK-sourced income, and getting the residency answer wrong is the single most common cause of a cross-border arrangement that does not work the way the household believed it did.

The second layer is treaty. The UK has double taxation agreements with most of the countries a UK retiree might plausibly move to; the treaty specifies which country gets the taxing rights on each kind of income. Pensions are typically treated under the treaty's pension articles, with the general rule that private pensions are taxable in the country of residence and that government service pensions remain taxable in the country that paid them.

The third layer is destination country. Once the treaty has assigned the taxing rights, the destination country's own rules govern the rate and the timing. A retiree who has internalised the three layers — residency first, treaty second, destination country third — has the structure to read any retail commentary on international retirement without confusion.

Portugal

The Non-Habitual Resident regime, which for a decade made Portugal the most popular destination for UK retirees by some distance, closed to new applicants at the end of 2023. Existing NHRs continue to run their ten-year window on the original terms — foreign pension income taxed at a flat 10 per cent for the duration. New arrivals from 2024 onward fall under the standard Portuguese tax regime, which taxes foreign pension income at progressive rates running up to 48 per cent. The replacement regime announced in 2024, sometimes referred to as NHR 2.0 or IFICI, is targeted at scientific and technical roles rather than at retirees and does not extend the favourable pension treatment that drew so many UK households across the decade just past.

Portugal continues to offer a strong UK retiree story for households that secured NHR status before the window closed. For UK households starting from scratch in 2026, the headline tax case for Portugal is materially weaker than it was three years ago, and any move now needs to be evaluated against the standard regime rather than against the closed NHR window that informed everyone's prior expectations.

Spain

Spanish tax residence applies to anyone present in Spain for more than 183 days in a calendar year, or whose centre of economic interests is Spain. Spanish residents are taxable on their worldwide income at progressive rates, with 2026 marginal rates running up to around 47 per cent depending on the autonomous community. Foreign private pensions are taxed in Spain under the UK-Spain double taxation agreement, with UK government service pensions remaining UK-taxable.

Spain levies a separate wealth tax in many regions, with substantial regional variation in thresholds and rates, and the temporary solidarity tax on large net wealth has applied since 2022 alongside the regional regime. The Modelo 720 reporting requirement obliges Spanish tax residents to declare foreign assets above €50,000 on a per-category basis; failure to file carried penalties that the European Court of Justice ruled excessive in 2022, and the regime has been reformed since, but the filing obligation itself remains.

The combination of progressive income tax on the pension, regional wealth tax, and ongoing reporting compliance gives Spain a higher operating-tax burden than NHR-era Portugal, and any household considering Spain needs to model the regional variation rather than working from a national average.

France

France offers a tax election that comes up repeatedly in UK retiree planning. Lump sums drawn from a UK pension can be taxed at a flat 7.5 per cent under Article 163 bis of the Code Général des Impôts, after a 10 per cent deduction, provided the contributions to the underlying pension were tax-deductible in the UK. The election is irrevocable once made, and is the reason France has become attractive for UK retirees taking large lump sums in retirement.

The implication catches many UK retirees off-guard: the 25 per cent tax-free cash that UK pension rules permit is not tax-free in France. France taxes it as foreign pension income. The Article 163 bis election is the mechanism that holds the rate at the flat 7.5 per cent rather than at progressive French marginal rates. The election needs to be made before the lump sum is drawn — once taken at the higher rate, the position cannot be retroactively unwound. Households planning a move to France in advance of taking the PCLS have a meaningfully different decision in front of them from households that have already drawn it.

Regular pension income drawn after retirement is taxed at French marginal rates, with a 10 per cent deduction up to a cap of around €4,300 per year. French marginal rates run to 45 per cent at the top, with social charges adding a further layer that varies by income type and by residency status. Inheritance is governed by droits de succession at progressive rates, with allowances varying by relationship to the deceased; spouse-to-spouse transfers are exempt, and child transfers carry a personal allowance of €100,000 per child before the progressive rates begin.

France is a strong jurisdiction for households drawing large lump sums; the regular-income picture is more ordinary, and the inheritance regime needs separate planning attention for households with substantial estates.

Cyprus

Cyprus offers the most distinctive flat-rate option for UK pension income at the wealth level this guide addresses. Cypriot tax residents can elect to have foreign pension income taxed at a flat 5 per cent above an annual personal allowance of €3,420; alternatively they can elect to be taxed under the standard progressive regime, which has a tax-free band up to €19,500 and a top rate of 35 per cent above €60,000. The election is made annually, which gives the household some flexibility year on year.

Cyprus has no inheritance tax in the UK sense and no wealth tax. The 60-day residency route, available to retirees who do not spend more than 183 days in any single other country and who meet the other statutory conditions, lets a UK retiree establish Cypriot tax residence with comparatively limited physical presence. The combination has made Cyprus increasingly attractive across the past five years, particularly for households leaving Portugal as the NHR window closes.

The four destinations side by side

The comparison below summarises the four destinations across the dimensions that matter most to a UK retiree pricing the move: tax treatment of regular pension income, treatment of UK pension lump sums, position on UK government service pensions, inheritance tax in the destination country, and any reporting compliance the destination country imposes. The table does the side-by-side; the prose around it does not need to repeat what the cells say.

Country	Regular pension income	UK lump sums	Government service pensions	Inheritance tax	Reporting
Portugal	Standard progressive (post-NHR) up to 48%; 10% under existing NHR	Standard progressive; partial exemption rules apply	UK-taxed	None on residents in the UK sense	Standard
Spain	Progressive up to ~47% (varies by region)	Marginal rate at draw	UK-taxed	Regional Impuesto sobre Sucesiones (varies)	Modelo 720 above €50,000
France	Marginal rate (10% deduction to ~€4,300 cap)	7.5% election (Art. 163 bis CGI)	UK-taxed	Droits de succession (progressive)	Standard
Cyprus	5% flat above €3,420 (annual election) or progressive	Treated under chosen regime	UK-taxed	None	Standard

David and Helen, in numbers

The Spanish position, in numbers. David, 71, retired from a UK private-sector career, drawing a £42,000 a year flexi-access pension from his SIPP plus a £14,000 a year DB pension from his old employer. UK state pension £11,500 a year. Helen, 68, drawing a £9,000 a year DB and her UK state pension of £11,500, with a smaller SIPP not yet in payment. They moved to a coastal village near Valencia five years ago, became Spanish tax residents in their first full year, and have been filing in Spain since.

Their combined gross UK pension income is £88,000; under the UK-Spain double taxation agreement, Spain has the taxing rights on the private pensions and on the DB amounts, which fall under Spanish progressive rates after the relevant Spanish allowances. Their effective Spanish tax on the pension income works out at roughly 23 per cent in the Valencian region, against the 0 per cent they would have paid under Portuguese NHR if they had moved there at the same time.

The choice they face now is whether the lifestyle case for Spain still earns its higher operating-tax bill, or whether a move to Cyprus or — at the cost of dropping the NHR-era saving entirely — a return to the UK is now the structurally cleaner answer. The decision sits with a qualified planner working both the UK position and the Spanish position on the same piece of paper.

Returning to the UK

This case appears in the practice repeatedly and is almost never covered well in retail commentary. A British retiree who has been non-UK-resident for many years — not unusual after a Middle East career or a long stint in Singapore — eventually decides to return. The headline tax point is that returning UK residents become subject to UK tax on their worldwide income and gains from the date of return.

The less-obvious point, the one §7 covered on the inheritance tax side, is that the long-term resident clock starts again from scratch if the household has been non-UK resident for ten consecutive tax years before the return. There is a window of up to ten years before the worldwide-IHT net closes around the household's non-UK assets again, and the window is structurally significant for households with substantial offshore wealth — non-UK property, offshore investment accounts, foreign-domiciled trusts — because it gives time to restructure under the new regime before the long-term resident test bites.

The mirror point, for households that have not yet reset the clock, is that returning before ten consecutive non-resident years are up means the long-term resident status carries forward, with no fresh window. The decision of when to return is structurally as important as the decision of whether to return.

A second regime, distinct from the long-term resident reset but triggered by the same ten-years-non-resident eligibility test, runs alongside on the income tax and capital gains tax side. The Foreign Income and Gains regime, introduced from 6 April 2025 to replace the remittance basis, allows a returning household that has been non-UK resident for at least ten consecutive tax years to claim relief from UK tax on its foreign income and foreign gains during the first four tax years of UK residence. Claims are made annually; the household can claim for foreign income, foreign gains, or both. The trade-off is that any tax year in which the claim is made forfeits the household's UK personal allowance and capital gains tax annual exempt amount for that year, so the regime earns its place where the foreign income or gains during the four-year window are substantial relative to UK-source income. Households returning with material offshore pension drawdowns, foreign rental income, or foreign-held investment portfolios crystallising gains during the window are usually inside the case for claiming; households whose offshore position is modest are usually outside it. The decision sits with a tax adviser working from the household's actual numbers. The regime pairs with the ten-year residency reset on the IHT side covered in §7, and should be on the table in any return-to-UK planning conversation.

Two shapes worth naming

Two shapes recur on the international side of the book and they are worth naming because the structural decision in each case is cleaner than its emotional shape suggests. The first is the household that moved to Portugal in 2017 or 2018, secured NHR status before the regime closed, has now seen out half of its ten-year window, and is approaching the question of what happens when the window closes. For this household the decision in front of them is whether to move on to a successor jurisdiction — Cyprus most often, less commonly Greece or Italy under the latter's flat-tax regime — or accept Portugal's standard regime for the remainder of retirement. The structural question is essentially when the next move is timed; the answer turns on the household's settling-in tolerance and on lifestyle preferences as much as on the marginal tax saved.

The second is the household that returned to the UK at the right moment. Twelve years non-resident in Singapore, full long-term resident clock reset achieved, returned to the UK three years before the worldwide net would have caught their non-UK pension wrappers, and used the open window to restructure offshore holdings under the new regime before the long-term resident test bit. Both households moved on structural grounds rather than emotional ones; the decisions held up through the years that followed.

The 2025 reforms, as they apply on the international side

§7's residency reforms beat covers the UK-side mechanics of the 2025 long-term resident regime — the ten-of-twenty test, the £325,000 spousal-exemption cap when one spouse is non-LTR, the tail period on departure, the ten-year reset — and the cross-border worked scenario that goes with them. The honest description for international readers is that the abolition of domicile and its replacement by long-term resident status materially reshapes the position for British retirees abroad and for non-British spouses living in the UK.

The reforms are described in §7 because the mechanics are UK-side. The consequence on the international side of the line is that the rules many households have planned around for years no longer apply in the form they did, and any household with a cross-border footprint should treat the post-2025 regime as a structural change requiring fresh advice rather than as a tweak to the prior position. The campaigning belongs in the column; the description belongs here.

A forward note on care

One consideration this guide picks up properly in §9 deserves a brief signpost here. UK retirees moving abroad commonly assume that destination-country provision for care in later life will resemble the UK's local-authority safety net. In several of the destinations this section covers, that assumption is wrong, and any plan that runs into the eighth or ninth decade of life needs to test it explicitly rather than defer it.

Before the next section

Before any international move, before any return decision, before any substantive change to a cross-border arrangement that has been running unreviewed for years, the practical action this section asks for is to commission a piece of advice that covers the UK and the destination country on the same piece of paper. The two tax systems, the relevant treaty, the residency mechanics, and the destination country's local rules cannot be reconciled by reading either side in isolation. International retirement is the area of the guide where the cost of advice is most reliably justified by the cost of getting the structure wrong without it. A qualified planner who works cross-border, ideally with a tax adviser in the destination country alongside, is the standard route for a household that wants the position settled rather than carried forward in hope.

Susan's mother has dementia. Susan has watched the cost of her mother's care accumulate in a way that frightens her, and she has not yet mapped that fear onto her own and Peter's planning. The conversation many couples do not have, despite seeing it coming for a decade, is the one about what happens if one or both of them needs care in their eighties or nineties — and the separate question of who pays and how. The conversation is uncomfortable because it forces a household to think honestly about decline, and uncomfortable conversations get postponed. The job of this section is to make the conversation easier to start by giving the household the framework, the rough arithmetic, and the one piece of paperwork that matters more than the rest.

The funding system

The English self-funding system, named once and clearly. A person in residential care with assets above £23,250 — the upper capital threshold in 2026 — is responsible for the full cost of their care. Below that threshold and above the lower threshold of £14,250, the local authority contributes on a tariffed basis with a partial contribution from the resident. Below £14,250, the local authority bears the cost in full. The household home is included in the assets test for residential care unless a spouse, dependent relative, or qualifying carer continues to live in it.

The forthcoming funding cap, originally legislated for introduction in 2023, deferred to October 2025, and now deferred again, is best treated as honest description rather than as a current piece of architecture; the reader should plan as if the cap is not yet in force. Scotland operates a separate regime in which personal and nursing care payments are made at fixed weekly rates to self-funders, with free personal care available from age 65 and extended to under-65s with assessed care needs since 2019. The Scottish position is genuinely different from the English; this section does not cover it in detail, but the reader resident in Scotland should not assume the English treatment applies. Wales and Northern Ireland operate further variations.

Care cost arithmetic, at 2026 rates, runs from around £1,200 a week for a basic residential placement, to over £2,000 a week for nursing care, and meaningfully higher again — £2,500 to £3,500 a week is not unusual — for specialist dementia care in southern England, with the very top of the market running higher still. A household model that does not stress-test against four to six years of care for one or both partners, longer where the family pattern includes dementia, is missing the largest single financial risk of the later years.

Three practical responses

Three responses to the care risk are available and are worth distinguishing because the household tends to drift toward the third by default rather than choosing among them.

The first is to earmark a specific portion of the estate as the care reserve. The household decides, in advance, that some defined sum — often a six-figure amount sized to the realistic worst case for one partner needing four or five years of care — is held back from the rest of the financial plan. The earmark is usually held inside an investment wrapper rather than in cash, so it does not erode against inflation across the years before it is needed.

The second response is insurance-backed. Pre-funded long-term care insurance, common in some other markets, is rare in the UK and many of the schemes that existed twenty years ago are no longer open to new entrants. Immediate Needs Annuities — bought at the point of entering care, paying a guaranteed income to the care provider, taxable only in part — remain available and earn their place in some specific situations, particularly where a household wants to fix the cost of care once it has begun rather than continue running market exposure on the care reserve.

The third response, common at the wealth level this guide addresses, is the simpler one of building enough total resource that the care question is absorbed by the plan rather than specifically provisioned for. A household whose invested wealth is materially in excess of what its lifetime spending requires can treat care as a draw against the long-term part of the portfolio if it materialises, without needing a separate reserve. None of the three responses is universally right; the framework depends on the household's wealth level, family pattern, and longevity assumptions.

Powers of attorney

This is the substantive practical message of the section. A Lasting Power of Attorney is a legal document, executed while the donor still has capacity, that nominates one or more attorneys to make decisions on the donor's behalf if the donor later loses capacity. There are two LPAs and they cover different territory. The Property and Financial Affairs LPA gives the attorney authority over the donor's bank accounts, investments, property, and other financial matters. The Health and Welfare LPA gives the attorney authority over decisions about medical treatment, care arrangements, and where the donor lives. Both can be drafted to take effect only on loss of capacity, or, in the case of the Property and Financial Affairs LPA, drafted to take effect immediately and used in tandem with the donor while they retain capacity.

Cost runs at around £82 per LPA in the registration fee paid to the Office of the Public Guardian, with solicitor fees additional where a solicitor draws the document; many households execute LPAs straightforwardly without a solicitor, though a solicitor's hand is useful in more complex family or asset situations. Registration takes around 8 to 10 weeks, and sometimes longer when the OPG is under load.

The single timing constraint is the load-bearing point. An LPA must be executed while the donor still has capacity. Once capacity is lost, an LPA cannot be put in place at all, and the family is forced into a Court of Protection deputyship application — slower, considerably more expensive, and with the deputy's authority running on a much tighter leash than an attorney's. Many households postpone the LPA conversation because it feels distant; many of those households then face the Court of Protection process at the worst possible moment, with a parent or partner already losing capacity and the family unable to act. The right time to execute the LPAs is years before the household believes it is strictly necessary.

This is one of the few places in this guide where I am going to step outside the framework and speak personally; I want to make one point unequivocally. The paperwork is straightforward, the cost is small, and the protection it can provide at the worst possible moment is considerable. Alongside a full and regularly updated will, LPAs are arguably the most important - and most misunderstood - pieces of paperwork in family life. Of every practical action this guide recommends, this is the one I would not allow to drift.

Two shapes worth naming

Two shapes recur on this question. The first is the household where the parent's care decisions, when they came, were eased because the LPA was in place in good time. The attorney — usually a son or daughter, often working with the surviving spouse — could move the parent's banking online, manage the care home invoices, sell the family home when the parent moved into residential care, and consolidate finances without contested process. The decisions were difficult emotionally but procedurally clean.

The second is the household where the LPA was not in place. The parent lost capacity unexpectedly — a stroke, a sudden cognitive decline — and the family then faced a six- to nine-month wait for a Court of Protection deputyship, with bills going unpaid, a property unable to be sold, and pension and benefit administration stalled. The first household paid £164 for two LPAs registered in good time. The second household paid several thousand pounds in court fees and solicitor time, and the financial cost was the smallest part of what they paid.

A note for international readers

The forward note in §8 deserves its substantive answer here. UK retirees who move abroad commonly assume that destination-country provision for care in later life will resemble the UK's local-authority safety net; the assumption is wrong in many of the destinations covered in §8. Spain, France, Portugal, and Cyprus all operate care systems with quite different structures from the English one, some with strong family-care expectations, some with means-tested systems calibrated very differently, some with private-pay markets that move at quite different price points from the UK. A household planning a retirement abroad should treat destination-country care provision as a separate piece of due diligence and should commission specific advice on it before the move, not after. The cross-border POA position needs separate attention as well — UK-issued LPAs do not always transfer cleanly to other jurisdictions, and a host-country equivalent is sometimes required.

Before the next section

The single most important piece of paperwork in this guide is the Lasting Power of Attorney. If the household does not have both LPAs in place — Property and Financial Affairs, and Health and Welfare — for both partners, the practical action this section asks for is to put them in place this quarter. Not next year, not when the household feels older, this quarter. The OPG forms are publicly available, the registration fee is modest, and the process is straightforward in most cases. Where the family or asset situation is more complex, a solicitor's hand is worth the additional cost. Care planning of any meaningful kind sits downstream of the LPAs being in place; without them the rest of the conversation cannot proceed past the moment capacity is lost. This is the floor item; everything else this section described builds on top of it.

The guide has been honest about its limits throughout. Every section closes on a practical action; several sections name decisions that benefit from a planner who can see the household's actual numbers; §3, §6, §7, and §8 all make the call explicitly. The reader who has worked through nine sections of structural material has a clearer picture of their financial life than they did when they began, and probably also a clearer sense of which of their decisions they cannot reasonably make alone. This section is for the reader at that moment. Not a sales pitch — the guide has earned the right to land calmly here, after nine sections of substantive material. As a final piece of orientation for a reader who is ready to move from reading to deciding, and who may need someone in the room with them at the moment they actually decide.

Five signals that point past the guide's range

Five signals, named cleanly. They are not exhaustive, and they do not all need to be present; any one of them is reason enough to commission personalised work.

The first is a material estate position — total household wealth above the level where the frozen IHT thresholds bite hard, where the post-2027 pension change has structural consequences, and where structured gifting and trust patterns earn their cost. §7's £3.08 million worked example sits in this band, and many readers of this guide are above it rather than below.

The second is a cross-border element — a destination country in view, a non-UK pension in the picture, a non-British spouse, or a recent move in either direction. §8 covers the substantive country-side mechanics; the household-specific reading sits with a planner who can hold both jurisdictions on the same piece of paper.

The third is a live annuitisation question — a household near the edge where the bucketing structure does not entirely cover the floor, where sequence risk feels acute, and where a partial annuity's role in the long-term layer warrants serious modelling rather than a textbook answer.

The fourth is a 2027 pension-into-estate optimisation that requires modelling — the withdrawal-sequencing rebalance §3 introduced and §7 returned to substantively. The right pace at which to draw the DC pot, given the household's other tax position and the surviving spouse's long-term position, is a numbers question and not a framework one.

The fifth is an imminent life change — retirement itself, a property move, a death or divorce in the family, a non-resident return, a sudden inheritance — anywhere the existing plan stops fitting the household's actual position.

What a good adviser does that a guide cannot

Five things, in escalating order of why they matter.

First, a good adviser sees the client's actual numbers. The guide has worked from rounded composites — Susan and Peter at £1.8 million, David and Helen at the Spanish position, the worked examples in §§4, 5, and 7 at illustrative wealth levels. The household's real numbers are different from any composite, and the right answer for the household sits inside its own balance sheet rather than inside any illustrative one. A planner can see the difference; a guide cannot.

Second, a good adviser tests scenarios against those numbers. Cashflow modelling, withdrawal-sequencing variants, stress tests against a 2008-shaped market, what-if-one-of-us-needs-care arithmetic — the household's plan only earns its credibility once its actual numbers have been run through the actual scenarios. The reader who finishes this guide knowing the framework still needs the modelling, and the modelling is the planner's substantive work.

Third, a good adviser coordinates. Tax advisers, solicitors, accountants, mortgage brokers, and occasionally specialist consultants on niche cross-border or estate matters all have a part to play in a serious household plan. A planner who acts as the coordinator for the household's professional cast keeps the parts working with each other rather than past each other. This is unglamorous and structurally invaluable.

Fourth, a good adviser sits across the desk for the decisions that benefit from someone else in the room. The annuitisation question, the 2027 pension restructuring, the residency-reform planning, the trust conversation, the move-abroad call — households consistently make these decisions better with a calmer, more structured outside voice on the other side of the table. Reading is not the same as deciding; the desk matters.

Fifth, a good adviser is in the room when the news is bad. This is the part that does not show up in qualifications, fee schedules, or first meetings, and that earns its keep across the decade or two that follow. Markets fall; family circumstances shift; the household's plan meets a year it did not expect. The household with a planner whose number it has, whose voice it knows, and whom it trusts to take a call at the moment the temptation is to sell or to stop, is in a structurally different position from the household acting alone with the same plan on paper. The behavioural part of the work — staying the course when the course is right, adjusting when it is not, not confusing market noise with structural change — is rarely done well by reading. It is done by the relationship, and the relationship is what the previous four points combine into.

How to test an adviser

The reader who is ready to commission personalised work is not always ready to choose well between advisers. Three tests that genuinely sort the field.

First, the questions worth asking on the first call. Where are you regulated, and what is your firm's permission scope? What qualifications do you hold for the substance of my situation, and how do those qualifications map to the specific decision you will be advising on? How are you paid, and how does the fee structure work over the lifetime of an engagement? What does your typical engagement look like over the first six months? Who else will I be working with, and how is that coordinated?

Second, the red flags. A first conversation that is mostly about the firm's model portfolio service and barely about the household's actual position. A pension-transfer recommendation arriving on the second meeting before the household's full income picture has been mapped. A qualifications wall that displays many letters but does not include any that are specifically relevant to the substance of the decision in front of the household.

Third, the fee. Whatever the structure — fixed fee, time-cost, retainer, percentage of assets, or some combination — the test the household should apply is the same. Fees should be transparent, understandable, and representative of the service being provided. The right structure varies with the household's complexity and the planner's offering.

Two journeys worth naming

Two journeys recur, and they are worth naming because the difference between them is mostly the household's preparation rather than its wealth or sophistication.

The first journey is the household that arrived already well-prepared. They had read the guide, mapped their pots, sketched their withdrawal pattern, opened the conversations they knew they had been postponing, and arrived at the first meeting with a working draft of the questions they wanted answered. The first meeting moved quickly because the framework was already in place; the planner's work could focus on the modelling and the modelling-led decisions, and the implementation followed the modelling efficiently. Six months later the household was structurally settled and the relationship had moved into steady-state.

The second journey is the household that tried to do the whole thing alone. They read the guide, took the framework on board, executed parts of it, but stopped short of commissioning the modelling on the decisions that needed it. The decisions got made on instinct rather than against numbers. A year later they came back, a substantial avoidable tax bill in hand from a withdrawal-sequencing pattern that had not been stress-tested, and the planner's work was harder and more expensive because the corrective work sat on top of an already partially-implemented plan. Both households got there in the end. The first got there cheaper, faster, and with materially less worry along the way.

Before you close the guide

If you recognise yourself in any of the five signals named here, the next step is to commission a piece of personalised work from a qualified planner who specialises in retirement-income planning and, where the household has cross-border elements, in the international position. Look for the qualifications and the fee structure described above; ask the questions that test alignment of interests; treat the first conversation as your due diligence rather than as the planner's. The household has done the harder half of the work by reading nine sections of structural material. Finding the right planner, asking the right questions, and beginning the modelling is the lighter half, and it sits squarely inside the household's control.

That is where this guide closes. If any part of it has helped you see your household's picture more clearly, named a question that had been sitting unasked, or simply made the conversation with a partner or planner easier to start, the guide has done what it was written to do. The work that follows the reading is yours to do; my hope is that the framework above is the right one to do it against.

The DrawdownAtlas.

A working reference, not a book.

The ten sections

Three short questions.

Start with the frame, then the sections built for your situation.

A note from Jessica

The Shift

Mapping the Full Picture

Welcome.

A note on what this Atlas is.

What this Atlas is and is not.

All ten sections

Decumulation is a different discipline

The first twelve to twenty-four months

Shift one: from income thinking to sustainability thinking

Shift two: from saving discipline to spending permission

Shift three: from single-horizon investing to multi-horizon investing

A pattern from practice

A note on patience

Before the next section

Why the picture comes first

The Different Categories

Susan and Peter's picture

For readers with foreign-held resources

The forgotten pot

Your own pot map

Before the next section

The two-stage framework

The three layers, in brief

The three lenses on the wrapper question

What each wrapper costs in tax

Susan and Peter, two phases

UFPLS versus flexi-access drawdown

The 2027 change to pensions in the estate

A practitioner observation

A note for international readers

Closing the section

A note on the word "risk"

Why the first years weigh more

Bucketing: the structure many of my clients use

Dynamic withdrawal: the discipline that goes with the structure

What this looks like at £1.68 million

How the structure played out in March 2020

A note on my qualifications

Before the next section

Where the rule came from

The assumptions, and how they land for a UK retiree

Inside the bucketing structure

Guardrails: the discipline that replaces rigidity

Three approaches compared

What this looks like at £1.68 million

A pattern from 2020 to 2022

Before the next section

The question inside the §4 structure

How annuities work

The three options side by side

What the decision looks like at £1.68 million

Typical shapes of household

Before the next section

The IHT architecture

The 2027 pension change

What this looks like at Susan and Peter's wealth level

Lifetime gifting strategies

Trusts

Insuring the IHT bill, written in trust

Estate planning sits on top of the bucketing structure

Wills and powers of attorney

Residency, the spousal exemption, and what changed in 2025

A note for international readers

Before the next section

The three layers

Portugal

Spain

France

Cyprus

The four destinations side by side

David and Helen, in numbers

Returning to the UK

Two shapes worth naming

The 2025 reforms, as they apply on the international side

The Drawdown
Atlas.