May 2026 : 7 min read
I put this exercise off for almost a year because I was afraid of what it might reveal. What it actually revealed was almost the opposite of what I'd feared.
There's a number most people planning early retirement are carrying around in their heads.
It's the number they think their retirement will cost. The annual figure they're planning around. The target their pension pot and ISA bridge need to sustain.
For most people that number is wrong. Not slightly wrong. Significantly wrong. And almost always wrong in the same direction - too high.
I know this because I was one of those people. For the better part of a year I was planning around a number I'd estimated rather than calculated. It felt roughly right. It was based on what I currently spent, adjusted down for the things I assumed would disappear when I stopped work.
The problem with estimated numbers is that they inherit all your assumptions. And the assumption most people make - unconsciously, reasonably, understandably - is that retired life costs roughly what working life costs. Maybe a bit less. But broadly similar.
That assumption is wrong for most people approaching early retirement from a demanding professional career. And the three-month spending reality check is the exercise that proves it.
Why I put it off
I want to be clear about this because I think the resistance is common and worth naming.
I knew this exercise existed. I'd read about it in multiple places - described in almost exactly the same terms each time. Three months of real statements. Every line. The honest question about what would survive the transition.
And I kept not doing it.
The reason, when I examined it honestly, was fear. Not the fear that the number would come back higher than I thought. Something more specific than that. The fear that doing the exercise properly - with real data rather than estimates - might reveal that the comfortable life I was imagining in retirement was more expensive than the plan could sustain. That the gap was larger than I'd been telling myself.
The estimate felt safe. It was my number. I'd constructed it carefully. It told me the plan worked. The real data felt like a risk - an opportunity for the plan to be less solid than it appeared.
This is exactly backwards. The estimate is the risk. The real data is the safety.
I did the exercise eventually - about ten months into the planning process, prompted partly by writing this site and partly by a conversation with someone who'd done it and described the result with a kind of surprised relief that made me curious.
What I found changed the plan significantly.
How to do it
The mechanics are simple. Three months of bank statements and credit card statements - real ones, not estimates. Every transaction. Categorised honestly.
I used five broad categories:
- Essential fixed costs - mortgage or rent, utilities (water, gas, electric), insurance, council tax.
- Essential variable costs - food, transport, household maintenance.
- Working-life costs - commuting, work clothes, hotels, lunches, the costs that exist specifically to support employment.
- Lifestyle costs - eating out, leisure, holidays, subscriptions (gym, phone, TV, music, medication).
- And one-offs - irregular large costs that don't recur monthly.
For each transaction the question is the same. Would I still spend this if I wasn't working?
Some answers are obvious. The commuting costs go entirely. The work wardrobe maintenance goes. The Thursday evening takeaway that exists because it's Thursday evening and you're exhausted - goes. The gym membership you use twice a week because it's the only window available in a working week - might go, replaced by walking that costs nothing.
Some answers are less obvious. The grocery delivery that exists because you can't face the supermarket after a long day - does that survive? The cleaner you employ not because you want to but because you simply don't have time to keep the house properly - does that continue?
The honest answer to these requires a genuine assessment of whether the spending is responding to the constraints of working life or reflecting something you'd genuinely choose. Many people discover that a significant proportion of what felt like genuine lifestyle spending was actually coping spending - rational responses to a life with too little time and too much stress, rather than genuine preferences.
What I actually found
Let me be specific, because the specificity is what makes this useful rather than abstract.
I went through three months of statements in a Sunday afternoon. I'd been putting it off. When I finally sat down with a coffee and printed everything out, it took about three hours.
The categories that disappeared almost entirely when I asked the honest question were striking in their volume.
The commuting. I was working away from home one or two nights a week for five years. Hotels, evening meals, the incidental costs of being on the road. None of it was expensed and there were always costs around the edges - fuel, parking, socialising. An entire category of spending closed overnight when I handed in my notice.
The convenience spending. I added it up. The takeaways, the deliveries, the bought lunches, the things I spent money on not because I particularly wanted them but because time was short and energy was lower. The figure was considerably larger than I'd estimated. And none of it - none of it - was something I'd choose to spend money on if I wasn't exhausted on a Thursday evening.
The work wardrobe. Suits, shoes, dry cleaning. Not extravagant. Just ongoing. And entirely work-generated.
The social spending around work. The leaving drinks, the birthday collections, the team lunches. Not things I resented at the time - but things that existed because of the job rather than because I'd have chosen them.
When I added up everything that was going to disappear, the number was larger than I'd assumed. Considerably larger.
The number that came back
I'm going to be specific here in a way that most posts on this topic aren't, because I think the specificity matters.
My estimated annual retirement spending figure - the one I'd been planning around - was approximately £42,000 per year for our household. That was my working assumption going into the exercise.
The figure that came back from the real data - stripping out everything working-life-related and honestly assessing what would and wouldn't survive the transition - was approximately £35,000 to £37,000 per year.
A gap of £5,000 to £7,000 annually. The difference, at a 3.5% sustainable withdrawal rate, between needing a pot of approximately £1.2m and needing a pot of approximately £1m. The difference between "this requires careful management" and "this is comfortably achievable."
That gap came from doing the real exercise rather than trusting the estimate. It was sitting there all along, invisible inside the assumption that working-life spending was a reliable proxy for retirement spending.
It isn't. For most people approaching early retirement from a demanding career, the two numbers are significantly different.
What tends to go up
I want to be open about the other direction too. Because the exercise isn't only about discovering what disappears.
Some things go up in retirement. Not by as much as the working-life costs come down - in my experience the net effect is almost always a lower number. But they go up.
- Travel. When you can go whenever you want rather than when the diary permits, you tend to go more. Off-peak pricing helps - the same trip costs considerably less on a Tuesday in October than on a Saturday in August. But the frequency may increase.
- Hobbies and activities. Things that were deferred during working life tend to get picked up in retirement. Some of them cost money. The honest exercise includes a realistic assessment of what those costs might be rather than assuming they'll be negligible.
- Food. When you have time to cook properly, the grocery bill sometimes goes up - better ingredients, more time spent on meals. In my case this partially offset the disappearance of the bought lunches and the takeaways.
- Home maintenance. With more time at home comes more awareness of the things that need doing. Projects that were deferred tend to happen in retirement. Worth budgeting for specifically rather than hoping they won't arise.
The exercise accounts for both directions. Not just what disappears but what emerges. Most people still find the net effect is a significantly lower number than the estimate. But the exercise should produce the real number rather than an optimistic one.
Why estimates are almost always too high
I've thought about why this pattern is so consistent - why the estimated retirement spending figure is almost always higher than the real one.
Part of it is the working-life anchor. We naturally estimate future spending based on current spending. And current spending includes all the working-life costs - the commuting, the convenience, the coping spending - that we don't fully register as work-related because they've become habitual.
Part of it is the lifestyle creep problem. Spending gradually expands to meet income over a career. What started as a choice becomes a habit. What started as a luxury becomes a baseline. The estimate inherits all of that accumulated expansion without questioning how much of it reflects genuine preference.
And part of it is the fear dynamic I described at the start. The estimate that tells you the plan works feels safer than the exercise that might tell you it doesn't. So the estimate stays comfortably vague and comfortably high rather than being tested against reality.
The exercise is the antidote to all three. Three months of real data, honestly assessed, strips away the anchoring, the lifestyle creep and the comfortable vagueness simultaneously.
When to do it
The best time to do this exercise is earlier than feels necessary. Not in the final weeks before you stop - at that point the number it produces has limited time to influence the plan. A year or two out gives you the real figure while there's still time to act on it.
The second-best time is now, wherever now is in your planning process.
I wish I'd done it at month two rather than month ten. The estimate I'd been planning around for eight months was wrong by enough to matter. Eight months of planning optimised around the wrong target.
Do it with real data rather than memory. Print the statements or export them from your banking app. Memory editing is real - we remember spending that felt significant and forget spending that felt habitual. The statements remember everything.
And share the results with your partner if you have one. The honest number - where the money is actually going, how much of it is work-related, what the real retirement cost looks like - is a conversation that should happen together rather than alone.
The relief
I want to end with what the experience actually felt like, because I think it matters.
The Sunday afternoon I spent going through three months of statements was not frightening. It was absorbing and then, gradually, relieving. Each category of working-life spending that I crossed off the list was a small reduction in the gap. Each hour that passed made the real number look more manageable rather than less.
By the end of the afternoon I had a number I trusted. Not an estimate hedged with uncertainty. A real figure, grounded in real data, with a specific set of honest assumptions about what would and wouldn't survive the transition.
The relief of that - of replacing vague with specific, estimate with real - was disproportionate to the effort involved. Three hours on a Sunday afternoon produced a clarity that eight months of spreadsheet-building hadn't.
That's the exercise. Do it with real data. Be honest about what's actually work-related spending rather than genuine preference. Account for what goes up as well as what goes down. And let the number that comes back replace the estimate you've been carrying.
It's almost certainly lower than you think.
Related posts
This post is part of the Anti-Panic Toolkit series. The main article - Ten Things That Actually Reduced the Stress of Planning Early Retirement - covers nine other practical techniques alongside this one.
For the Retirement Basics version of this topic - the mechanics of calculating your enough number: How Much Money Do You Need to Retire Early in the UK?
For what spending actually looks like in retirement - what goes and what stays: How to Cut Your Costs in Early Retirement Without Sacrificing the Things That Matter
For the broader planning framework this exercise feeds into: Building Your Own Early Retirement Plan - A Step-by-Step Framework
Tony writes about his personal journey to early retirement at freebefore65.co.uk. He is not a financial adviser. All content reflects his own experience and research and should be taken as a starting point for your own thinking, not as professional advice.
Add comment
Comments