There is a standard assumption buried inside almost every retirement calculator. It is rarely stated explicitly, which is part of why it persists. The assumption is this: that your spending in retirement will decline gradually and predictably over time, as age reduces your appetite for activity and your body imposes its own economy on your ambitions.
It is a reassuring model. It suggests that the hardest part of retirement — the first years, when you are adjusting from a salary — is also the most expensive. Things get easier from there. By the time your health begins to deteriorate, the financial pressure has already eased.
The data do not support this picture. What researchers consistently find instead is a pattern shaped like a smile.
1. The shape of a retirement
The smile pattern in retirement spending was identified and named by financial planner David Blanchett in a 2014 paper that analysed actual spending data from thousands of retirees across multiple decades. The finding was counterintuitive and, once seen, difficult to ignore.
Spending starts high. In the early years of retirement — the go-go years — retirees spend actively. Travel, leisure, home improvements, gifts to children and grandchildren. Many people treat the first years of retirement as a long-deferred reward, and they spend accordingly.
Then spending falls. From roughly the mid-60s into the 70s, real (inflation-adjusted) spending declines. Retirees settle into routines. The novelty of free time has normalised. Physical energy begins its quiet negotiation. This is the slow-go phase.
Then spending rises again. In the final years — the no-go years — healthcare costs climb steeply. Long-term care, medical procedures, home adaptations, in-home support. Spending on experiences may have collapsed, but spending on survival and dignity has not.
The shape is a smile: high at the left, dipping in the middle, rising at the right. And it changes everything about how you should plan.
2. Three phases, three different problems
Understanding the smile is not just an academic exercise. Each phase presents a distinct financial and psychological challenge that a linear planning model is poorly equipped to handle.
| Phase | Typical age | Spending profile | Planning challenge |
|---|---|---|---|
| Go-go | 60s – early 70s | Active leisure, travel, dining, gifts. Can match pre-retirement levels. | Sequence-of-returns risk is highest here. Large early withdrawals permanently impair portfolios. |
| Slow-go | Mid-70s | Discretionary spending falls. Routines established. Travel and activity decline naturally. | Risk of under-spending — leaving go-go experiences on the table due to excessive caution. |
| No-go | 80s+ | Healthcare, care facilities, in-home support. Discretionary spend near zero; medical spend sharp. | Longevity risk combined with care cost inflation. The tail risk is both long and expensive. |
The slow-go dip is particularly important because it creates a false sense of security. A retiree who has managed comfortably through their late 60s and mid-70s may conclude that their financial plan is working. What they may not have modelled adequately is the cost curve that begins its upward climb in the no-go years.
“The slow-go years do not mean the financial pressure is over. They mean it has temporarily moved offstage.”
3. The connection to Die With Zero
Bill Perkins’s argument in Die With Zero focuses primarily on the go-go end of the smile: the failure of people to spend actively and deliberately during the years when experiences are most available to them. His framework of time buckets maps almost perfectly onto the go-go years.
But the smile pattern extends Perkins’s argument in an important direction he underweights: the no-go end of the curve.
Perkins’s model sometimes implies a gentle decline — spend well in the go-go years, reduce in slow-go, arrive at death with near-zero. The smile suggests the arithmetic is more complicated. The no-go rise is real, often underestimated, and frequently catastrophic for plans that assumed a gentle downward glide.
| Risk | What it means |
|---|---|
| Go-go risk | Perkins’s concern: you defer experiences that belong to this decade and miss them entirely. The memory dividend never gets invested. |
| No-go risk | The smile’s second peak: healthcare and care costs rise sharply, often exceeding what retirees projected during the comfortable slow-go dip. |
4. What the smile means for your withdrawal strategy
The 4% rule was derived from historical portfolio data and assumes, implicitly, a relatively stable or gently declining real withdrawal rate. It does not model the smile. A spending-smile-aware withdrawal strategy looks meaningfully different.
Go-go years: Budget for a higher withdrawal rate. This is when the return on spending is highest. A dynamic withdrawal approach that permits higher early draws better matches reality.
Slow-go years: Lower actual spending provides a rebuilding window. The temptation is to treat this as proof that you have over-saved. It is not — it is the cushion that funds the no-go spike.
No-go years: Dedicated healthcare and care reserves — not drawn from your general portfolio — provide the most reliable protection. Long-term care insurance or ring-fenced capital are worth serious consideration.
70% of people over 65 will require some form of long-term care at some point in their lives, according to US Department of Health data. The average duration of care need is around three years — but the distribution has a long tail. Planning for the median is not the same as planning for the risk.
5. The psychological dimension
The smile pattern is not only a financial phenomenon. It reflects something deeper about how spending preferences shift across a life.
In the go-go years, retirees are spending on experiences — on doing. The memory dividend Perkins describes is most actively generated here. The psychological return is high. Money converts efficiently into satisfaction.
In the slow-go years, the appetite for doing diminishes. People find satisfaction in simpler, cheaper pleasures. The psychological need for novelty and adventure has, for many, been substantially met. Spending less is not deprivation — it reflects a genuine shift in what constitutes a good day.
In the no-go years, spending becomes largely involuntary. It is not about experiences at all — it is about maintenance, dignity, and care. It does not generate memory dividends; it sustains the conditions for whatever quality of life remains.
The planning question most people avoid: Your retirement model probably has a number for healthcare costs. Does it have a model for what kind of care you want, where you want to receive it, and how much autonomy you are willing to trade for cost savings?
6. The honest limits of the model
The smile is a population-level pattern, not an individual guarantee. Some people spend actively well into their 80s. Others experience a compressed decline that combines slow-go and no-go into a shorter, more intense final phase. Healthcare costs vary by condition, country, and care setting in ways that make any single projection inadequate.
The smile also describes nominal spending patterns. When you adjust for the higher inflation rate of healthcare goods relative to general consumer prices — in most countries, medical inflation runs significantly above headline CPI — the no-go spike looks even steeper in real terms.
What the smile gives you is not precision. It gives you the right shape of expectation. Planning for a straight line produces plans that are wrong in predictable directions. Planning with the smile in mind produces plans that are at least oriented toward reality.
What this means for your FIRE plan
Plan explicitly for three phases, not one. A single withdrawal rate applied to a single time horizon ignores the structurally different spending profiles of each phase. Even a rough three-bucket model — go-go, slow-go, no-go — is more honest than a linear projection.
Front-load your experience budget. The go-go window is finite. Experiences deferred to slow-go or no-go often do not happen. The memory dividend from go-go spending compounds; the memory dividend from experiences that never occurred is zero.
Do not mistake the slow-go dip for surplus. The fall in discretionary spending during the middle phase is real, but it does not indicate over-accumulation. It is the cushion that absorbs the no-go rise. Spend it early and you may find the second spike has nowhere to land.
Make care decisions now. The most under-planned dimension of the no-go phase is not the money — it is the decision architecture. What kind of care do you want? Where? On what terms? These conversations, deferred until they become urgent, are made under the worst possible conditions.
Use our Retirement Calculator to model your FIRE number with a realistic spending curve, and our Life Expectancy Quiz to understand how long each phase might last.
This article draws on research by David Blanchett and others on observed retirement spending patterns. It is intended as general educational content and does not constitute financial advice. For guidance specific to your circumstances, consult a qualified financial adviser.