There is something slightly contradictory about the way retirement planning has developed. Our ability to model the financial consequences of retirement has become increasingly sophisticated, while our understanding of the retirement being modelled can remain surprisingly superficial. We can project investment returns, inflation, pension income, taxation, withdrawals and expenditure over several decades. We can introduce different scenarios, model adverse conditions and illustrate the effect of decisions that have not yet been made. None of this is unimportant. Good modelling can materially improve financial planning. But the quality of the output still depends on the quality of the assumptions we put into it, and some of the most consequential assumptions in a retirement plan are often established remarkably quickly.
A client tells us they would like to retire at 60 and believes they will need £60,000 a year. Perhaps they want a few more holidays, a new car every few years and some money for the house. We can take those assumptions and construct an impressive projection of the next thirty years. What the sophistication of the model can disguise, however, is how little work may have gone into establishing whether £60,000 bears any meaningful relationship to the life the client will actually lead.
This matters because retirement is not simply a financial event. Indeed, as I argued recently in What if retirement isn't the destination?, for some people conventional retirement may not be a particularly appealing objective at all. Some people genuinely enjoy their work. Work may provide challenge, purpose, structure, relationships and a sense of continuing contribution. The right later-life plan for them may involve working less, changing what they do, taking extended periods away or retaining professional involvement for as long as it remains worthwhile. There is little merit in financial planning helping somebody achieve an objective that was accepted without ever being properly examined.
Even where stopping work is genuinely what the client wants, the retirement date tells us remarkably little about what follows it. That, for me, is the more interesting planning problem.
Imagine somebody retiring at 60 and living into their nineties. We would never sensibly describe the previous thirty years of their life as a single phase, yet we routinely make assumptions that come surprisingly close to doing that with the next thirty. In reality, later life may contain several quite distinct periods. There may be a transition away from work rather than an abrupt stop. That may be followed by years in which time, health and financial independence coincide and the client has greater freedom than at almost any previous point in adult life. Much later, the pace may change. Travel may become less important, home and family more so, and eventually questions around support or care may become relevant. Throughout all of this sits another question: how much of the wealth accumulated during life is intended to be consumed, how much is being retained for security, how much might be used to help other people and what, intentionally or otherwise, is likely to remain?
That is a much richer planning problem than determining an appropriate retirement income.
It is also why I have developed Map 5, the Later Life Map. The framework moves through Transition, Active, Less Active, Care Reserve and eventually Estate/IHT, but those headings should not be mistaken for a prediction of how somebody's life will unfold. In particular, Transition, Active and Less Active are not age bands. The intention is to allow the client to give those phases meaning and, where appropriate, timing. The framework is there to improve the conversation, not to replace one set of arbitrary assumptions with another.
The distinction becomes important when we begin discussing expenditure. Financial planning has an understandable tendency to move towards numbers because numbers are ultimately required. Yet I increasingly think the sequence matters. The principle underpinning Map 5 is story first, daily life second, spending third, modelling later. That is not an argument for avoiding financial analysis. It is an argument for giving the analysis something better to analyse.
If a client says they intend to travel extensively in retirement, the planner could simply increase the holiday budget. But “travel more” can describe profoundly different lives. Three months each winter somewhere warm is not the same as several expensive long-haul holidays. Visiting children overseas is not the same as wanting to explore the world. Nor does either tell us what happens during the other forty-eight weeks of the year. The more revealing conversation may concern ordinary life: where the client expects to live, what a normal week contains, which interests will occupy their time, who they expect to spend that time with, and what parts of working life will need to be replaced when work is no longer there.
Only then does expenditure begin to acquire context.
We should not overstate what this achieves. A 55-year-old cannot reliably tell us what their 78-year-old self will want, and a framework does not remove that uncertainty. People change. Families change. Health changes. Interests disappear and others emerge. Some of the things that appear critically important before retirement may prove surprisingly unimportant once it arrives. A credible later-life plan therefore cannot mean producing a more elaborate prediction of an unknowable future.
What we can do is make today's assumptions more explicit and better understood.
There is a considerable difference between saying that expenditure falls by 20 per cent at 75 because that is the convention built into the model, and discussing how the client imagines life changing as they become less active and deriving an assumption from that conversation. Neither will necessarily prove correct. The second at least has an intellectual basis that the client and planner can revisit. When circumstances change, we know what assumption has changed and why.
The same principle becomes more important as planning moves further into later life. Care is an obvious example. It would be disproportionate to turn every retirement discussion into detailed care planning for an event that may never arise, but ignoring the possibility altogether can be equally unsatisfactory. Map 5 therefore treats care primarily as a question of optionality: whether the client wants capital notionally available should additional support eventually be required, rather than pretending that we can forecast the eventual form, timing or cost of care decades in advance.
Eventually this leads into the estate, and here the distinction between different parts of financial planning becomes increasingly artificial. Money retained for later-life security cannot simultaneously be given to children. Capital deliberately preserved for an inheritance may restrict expenditure during the client's healthiest years. Lifetime gifting changes the eventual estate and potentially the inheritance tax position. Conversely, a client who accumulates a substantial estate without ever deciding that leaving one was an objective may simply have spent substantially less during life than they could have afforded.
Those are not principally tax questions. They are questions about the purpose of capital and the competing claims upon it.
This is also where the Later Life Map connects with the wider planning process rather than operating as an isolated retirement tool. The people identified through the family conversation matter. Decisions about lifetime gifting matter. The existing assets and liabilities matter. So does the client's view of what constitutes sufficient security for themselves and sufficient provision for others. The Map 5 process explicitly connects these areas because later-life spending, care, gifting and inheritance are different uses of the same finite resources.
Only after developing that picture do I think the cash-flow model can do its most valuable work. The purpose is no longer simply to test whether retirement at 60 is affordable at £60,000 a year. We can start asking what happens if work becomes optional earlier, whether the client could spend more during the years when they are most likely to enjoy it, what giving money away now does to later security, how much capital genuinely needs to remain available and whether the eventual estate is being created deliberately or merely emerging as the residual consequence of cautious assumptions.
At that point modelling stops being merely a test of affordability and becomes a way of exploring choices. That matters because clients can arrive with unnecessarily pessimistic expectations about what their resources will allow, and modelling can reveal alternatives and trade-offs that are difficult to appreciate from asset values alone. It should be capable of changing the conversation, not simply confirming the assumptions with which the conversation began.
Map 5 is therefore not an alternative to cash-flow modelling, nor is it an attempt to turn retirement into a sequence of five predictable stages. It is intended to address the problem immediately before the modelling begins: whether we have understood enough about the client's later life to know what we are asking the model to test.
There will always be uncertainty. A plan stretching several decades into the future that does not acknowledge uncertainty is probably less sophisticated, not more. The answer is not to abandon long-term modelling, but to become more thoughtful about the assumptions underneath it and more willing to revisit them as the client's actual life replaces the life we once imagined.
Perhaps that is the more useful definition of retirement planning. Not determining the number somebody needs at a particular age, but understanding what their capital is intended to make possible across the rest of their life, testing the choices available to them and continuing to adapt the plan as reality unfolds.
Before asking whether the money works, we need to understand what we want the money to do.
Map 5 starts there.