Running out
    31 October 2026
    6 min read

    Is it a plan, or a hope?

    One assumption in our retirement plan is worth 14 of 100 futures. I went looking at what it rests on.

    The usual disclaimer

    I am not a financial adviser and this is not advice. It is our own plan, run through my own calculator, one assumption at a time. Your plan will give different numbers, and the point of this post is that you should go and get them.

    Our retirement plan works in 86 of 100 futures. That is the calculator's count of how many simulated futures pay every year of spending to age 85 without running short. Delete one assumption and it works in 72.

    The assumption is a pair of spending steps. From 70, we spend £1,000 a month less than we do at the start. From 80, another £1,000 a month less. Both in today's money. I put them in because studies of real pensioners say spending falls through the seventies and eighties, and because it felt right: fewer long trips, fewer big purchases, a quieter life. I had not asked how much of the plan was standing on them.

    The answer is 14 points. So I spent a week moving the steps around to see what they are actually made of. Five things fell out, and the last one is the one I did not want.

    The table

    Same plan, same markets in every run, 1,000 futures, works to 85. The bracket is the change from the plan as it stands.

    Spending stepsFutures that last to 85
    As planned: £1,000 a month less from 70, another £1,000 from 8086
    No step-downs at all72 (-14)
    Only the 70 step84 (-2)
    Only the 80 step76 (-10)
    One £1,000 step, from 60 / 65 / 70 / 75 / 8091 / 88 / 84 / 80 / 76
    £250 a month more now, £1,250 less from 7084 (-2)
    £500 a month more now, £2,000 less from 7085 (-1)
    No steps, £7,000 a year less from day one87 (+1)
    £500 a month off at each of 65, 70, 75 and 8090 (+4)
    The whole £2,000 a month off from 6597 (+11)

    1. Timing beats size

    Take the plan with no steps (72) and add one £1,000 a month cut. Start it at 60 and the plan works in 91. At 65, 88. At 70, 84. At 75, 80. At 80, 76. That is close to a point for every year earlier, all the way down the line. Starting the same cut five years sooner is worth about as much as doubling it.

    The reason is the one that sits under most of retirement planning. The futures that fail do not fail at 85. They fail early, when a bad run of markets meets full spending and the plan sells shares at the bottom to pay for the same life as before. Every pound not spent in those years is a pound that did not have to be sold cheaply. A pound not spent at 80 arrives after the damage.

    2. The 80 step is nearly decorative

    The two steps together are worth 14 points. The 70 step on its own is worth 12. The 80 step on its own is worth 4. In the worst tenth of futures with no steps, the money first runs short at 73. By 80, the futures that were going to fail have already failed, and the ones that were going to work would have worked anyway.

    I had thought of the two steps as equal halves of one idea. They are not. The 70 step is the plan. The 80 step is a footnote to it.

    3. You cannot borrow from your older self

    The tempting version of this is: spend a bit more in our sixties while we can enjoy it, and cut harder at 70 to pay for it. The model will not have it. £250 a month more now against a £1,250 cut at 70: 84, down 2. £500 a month more now against a £2,000 cut at 70, a £2,500 a month swing: 85, still down 1. No version of that trade gets back to 86.

    It is the same mechanism as finding 1 seen from the other side. Extra spending early is paid for at the worst prices, and no amount of later frugality buys those shares back.

    4. The promise has a price today

    Our two steps are worth the same to the odds as spending £7,000 a year less from day one with no steps at all: 87. So "we will spend £24,000 a year less by 80" and "we will spend £7,000 a year less from the day we retire" are the same plan to the model. One of them is a choice we make when we retire. The other is a bet on who we will be at 70.

    5. A slope beats two cliffs

    Spread the same total cut as £500 a month off at each of 65, 70, 75 and 80 and the plan works in 90, up 4, because a quarter of it arrives five years earlier. Take the whole £2,000 a month off from 65 and it is 97. Gentle and early beats steep and late, every time I tried it.

    So, a plan or a hope?

    Our plan assumes we spend 40% less at 80 than at the start. The research I have read says spending through the seventies and eighties does fall, often by a quarter or more in real terms, before care costs arrive for some. A quarter is not 40%. So some of those 14 points may be a hope dressed as a plan.

    What I take from it is not to delete the steps. It is to stop counting the 80 step as load-bearing, and to treat the 70 step as a decision we make at 70 with our eyes open, not a number the plan quietly relies on. And finding 1 says the honest version of "we will spend less when we are older" is "we will spend a bit less from the start", because that is the only cut the odds actually feel.

    What did you assume you would spend at 80, and where did the number come from? If the answer is "I did not think about it", your plan has a step in it too. It is just set to zero.

    To try this on your own numbers: in the planner, add the card "Will your spending change at set ages?" and put in a step. Then move its age and its size and watch the count. The story page tells you what each step did.

    New on the site this week

    Nearly all of this came from readers' emails, which is how most things here get built.

    • A ceiling on each person's pension withdrawals. A couple can now keep the smaller pension inside the personal allowance so it comes out without income tax, or inside the basic-rate band, and the other person covers the rest. On the planner's tax card and the calculator, and the story page says how it played out.
    • A correction: the first year of State Pension. On a plan with a retirement month, a State Pension starting part-way through a tax year was paid for the whole year. It now pays the part after it starts. A reader with a February start date caught it; his 2029/30 showed £12,548 where it should have been about £2,200.
    • "Any other income?" is now question 6 on the planner. A final salary pension, rent, part-time work, each with its amount and when it starts and stops. It used to hide among the optional extras and people kept missing it.
    • The year by year page, tidied. Whole pounds in every opened year, the tax working shows each year's standard bands, the allowance line no longer reads as a negative number, and on a plan that stops part-way through a tax year the charts label each bar with its tax year.
    • The accumulation table says how many months each row covers, and the contribution boxes say the figure is the amount that lands in the pot, tax relief included.

    Our own plan as of September 2026, works to 85, 1,000 futures, the same market sequence in every run. The count moves three or four points with the number of futures you run, so treat the figures as a shape rather than a decimal. Not advice. I am not a financial adviser.

    Hear when something changes

    One short email when something changes

    A new feature or a new article. A couple a month at most. No tips, no selling, unsubscribe in one click.

    You will get one email asking you to confirm. Addresses are held by Buttondown, not on this site. How your data is handled.