The assumption hiding inside "25× your expenses"
Almost every retirement rule of thumb starts the same way: take what you expect to spend in a year, divide by a withdrawal rate, and that is your number. At a 4% withdrawal rate, $40,000 a year implies $1,000,000 — the familiar "25× your expenses."
That arithmetic quietly assumes something quite specific: that you will spend $40,000, in today's purchasing power, every single year from the day you retire until the day you die. Thirty, forty, sometimes fifty consecutive years of identical real spending.
Studies of what retirees actually spend do not find that pattern. They find a shape.
The shape
Spending tends to run at its highest in the first stretch of retirement — the years with the energy and health for travel, hobbies, family visits, the things retirement was being saved for. Through the seventies it typically eases: less travel, fewer big discretionary purchases, a smaller household, a slower rhythm. Then, in the final years, it commonly lifts again as health and care costs arrive.
High, then lower, then higher. Plotted against age it looks like a shallow smile, which is where the name comes from.
It is worth being clear about what this is and is not. It is a pattern observed across populations, not a law, and certainly not a prediction about any particular person. Plenty of people spend more later, not less. Someone who retires into a long illness, or who takes on care for a relative, may see nothing resembling a smile at all. What the shape gives you is a more defensible default than "flat forever" — not certainty.
What it does to the number
OptiAI models the smile as three stages, each holding a share of your base spending:
- Active years — from 55, at 100% of your base figure.
- Slower years — from 71, at 80%.
- Later years — from 83, at 85%, as care costs lift the curve back up.
Averaged across a retirement running from 65 to 95, those stages work out to about 86% of the base figure per year. So the $40,000-a-year plan above does not need $1,000,000 — it needs roughly $860,000. Retire earlier and more of the horizon sits in the expensive active phase: from 55 to 95 the average rises to about 89.5%, so a plan spending $360,000 a year needs about $8,055,000 rather than $9,000,000. On that larger plan the flat assumption overstates the requirement by around $945,000.
Those are not small differences. They are the difference between a target that is reachable and one that is not, or between retiring at 62 and retiring at 66.
Why the stages are anchored to your age, not your retirement date
A subtle but important detail: the boundaries above are absolute ages, not "twelve years after you stop working." The slowdown and the late-life care lift are driven by age and health, not by how long you have been retired. Anchoring them to the retirement date would put someone who retires at 40 into their care years at 68, which is not what the research describes.
The practical consequence is that if you retire late, you simply open in whichever stage you are already in. Someone retiring at 75 never gets an "active years" phase, because by the standard table that phase has already passed.
Flat is not the conservative choice
It is tempting to treat flat spending as the cautious assumption — assume the worst, and anything better is a bonus. That reasoning does not quite hold up. Flat spending is not a deliberate safety margin; it is a different estimate that happens to be larger, and it is not the estimate the evidence supports. A cushion you did not choose, cannot see, and cannot size is not really a cushion.
If you want margin in your plan — and there are good reasons to want some, given everything below — the honest way to get it is to add it explicitly and know exactly how big it is. That way you can reason about it, rather than inheriting it from a simplification.
What a spending curve still does not capture
A smile-adjusted number is a better estimate than a flat one. It is still an estimate, and a fairly coarse one. It does not account for:
- Healthcare costs, which frequently rise faster than general inflation and can dominate late-life spending.
- Long-term care, which is expensive, unevenly distributed, and difficult to plan for with an average.
- Taxes on withdrawals, which vary enormously by account type and jurisdiction.
- Sequence-of-returns risk — the order returns arrive in, which matters far more than the average once you are drawing down rather than contributing.
- The fact that a withdrawal-rate rule treats a dollar spent at 90 as equivalent to a dollar spent at 56. That simplification is inherited from the rule itself, not introduced by the spending curve.
- Your own life changing — a move, a health event, supporting a family member, a different retirement date than the one you assumed.
Any of these can matter more to your plan than the smile does. Treat the number as a compass bearing, not a destination.
See it with your own numbers
The Coast FI calculator, retirement calculator, and FIRE calculator all use the model described here, and each shows the flat-spending figure alongside so you can see the size of the difference for your own inputs. They use the same model OptiAI uses inside the app, so the number does not move when you sign up.
Disclaimer
This guide is for informational and educational purposes only and is not financial advice, a recommendation, or a substitute for professional guidance. Spending patterns, returns, inflation, and individual circumstances vary widely, and any projection is conditional on assumptions that may not hold. Consult a qualified financial professional before making decisions about your retirement plan.
How OptiAI helps
OptiAI keeps your retirement assumptions — including the spending stages, which you can edit — connected to your real net worth and goals, so you can revisit them as life changes and ask an AI assistant how you are tracking, instead of rebuilding this model by hand every year.