What is a gilt ladder, and why haven't I got one?

One of those things everyone in the forums assumes you already have.

The usual disclaimer

I am not a financial adviser and this is not advice. It is me running my own plan through a model I built, and showing you the numbers that came out. Your plan will behave differently. Every figure here comes from the simulator on this site, and the assumptions are stated as I go.

Until I started delving deep into my own pension I had never heard of a gilt ladder. Not once, in twenty-odd years of paying into the thing. Then I built a drawdown model, started reading properly, and found it everywhere: forums, adviser blogs, comments under my own posts. Always mentioned as though everybody already had one and I had somehow missed the assembly.

So I did not know what it was, let alone whether I should have one, where it should sit, or when I should build it. This post is me working all four out on my own figures.

Some context on my own position, because it shapes the answer. I have stacked my savings very heavily into the pension, and not out of strategy. It is because I cannot be trusted with an ISA. Money I can see is money I will find a use for. The pension is the only wrapper with a lock on the door, so that is where nearly everything went. The result is a plan with a decent pot, almost nothing outside it, and a lump of tax-free cash arriving on the day I stop work. If you are reading this thinking your own savings are lopsided in the same direction, this is not too late and the numbers below are for you.

Run this on your own plan, not mine

Every table below is my plan with one thing changed. There is now a page that does the same on yours: what protects this plan? It runs each defence against each of the crashes that actually broke retirements, twenty full runs in a couple of seconds, and lets you move any single assumption through a range. My answers depend on my pots, my spending and the gilt yield I can get. At least two of those are different for you.

What a gilt ladder actually is

A conventional gilt is a loan to the UK government. You buy it, you receive its scheduled coupons along the way, and you get its face value back on a fixed date. Its price moves in between, sometimes sharply, but if you hold it to maturity those movements do not change the cash you receive. That is the whole product, and the certainty is the point.

A ladder is several of them, maturing one after another. Buy one that pays out next year, one the year after, one the year after that. Each year a rung matures, hands you cash, and you live on it. That is it. There is nothing clever going on. It is money you have deliberately parked somewhere boring, arranged so a known amount lands each year.

There are two shapes, and the difference matters later:

A one-off bridge. Set aside a sum at the start, it drains over the years you chose, then it is gone and you are fully invested again.
A rolling buffer. Always hold, say, three years of spending. Each year one rung matures and you replace it, so the buffer is permanent.

What it is for, and it is only one thing

A ladder will not make you richer. Money parked in gilts earns the gilt yield, not what the stock market earns. Over thirty years that gap costs you real money.

What it buys is protection from one specific disaster: a bad run in the first few years, while you are drawing an income. This is what people mean by sequence of returns risk, or SORR if you have met it in a forum thread, and it is the thing that actually sinks retirements. A pot that falls thirty per cent in a good year for your health is an annoyance. A pot that falls thirty per cent while you are pulling sixty thousand pounds a year out of it may never recover, because you sold the shares at the bottom to pay for the shopping.

To make that concrete, here is what the decade from 2008 actually did to a UK investor:

Year one of your retirement: shares down 29.9 per cent, inflation 4.0 per cent. Year two, up 14.5. Year three, down 3.5. Year four, up 12.3. Year five, up 20.8. It recovers, and handsomely. The question is whether your pot is still there to enjoy the recovery, or whether you spent the bottom of it on groceries.

A ladder means the first three years of spending are already sitting in cash. You do not sell a single share in the crash. That is the entire pitch.

The part nobody mentions: where the money comes from

Every article I read described the ladder itself in detail and skipped the question of what you buy it with. On my own figures that question turned out to matter more than the gilts did.

You can build one inside the pension, by moving part of the pot into gilts. The money stays in the wrapper, but every pound you eventually spend from it is a pension withdrawal and is taxed as income. On my plan, building and refilling a three-year buffer this way costs around £49,000 in tax over the life of the plan, and it shrinks the pot that later buys my annuity.

Or you can build one outside the pension, from money you already hold: savings, an ISA, or the tax-free cash you take when you stop. Spending from that ladder is tax free, because the tax was already dealt with on the way out.

Same three years of gilts. Completely different trade.

My own plan, four ways

My plan: two pots, the larger one mine, retiring at 58, sixty thousand a year of take-home spending, half the pot buying a fixed-term annuity at 62, and sixty thousand pounds of tax-free cash available on day one. Almost nothing outside the pension, for the reason given above. I ran four versions, four thousand simulated futures each, measured to age 85.

Assuming five per cent growth with ten per cent swings, which is my cautious setting:

What I do with the tax-free cash	Works in	Median left at 85, today’s money
Leave it in the pension, no ladder	82 of 100	£265,000
Take it, hold it as savings, no ladder	82 of 100	£267,000
Take it, put it in a three-year gilt bridge	83 of 100	£278,000
Leave it in, build a rolling buffer from the pot	84 of 100	£274,000

Read the first two rows together. Taking the tax-free cash, on its own, does almost nothing. It moves the money from one place to another and the plan barely notices. That surprised me, and it is worth saying out loud to anyone agonising over whether to take their twenty-five per cent: on my plan, the taking is not the decision. What you then do with it is.

The gain shows up in the third row, where the same cash is arranged as a ladder instead of left as savings.

It depends entirely on what you think markets will do

Here is where I nearly wrote a confident conclusion and got it backwards. Run the same four plans at a more optimistic growth assumption and the answer flips:

	At 5% growth	At 9.8% growth
No ladder	82 of 100	92 of 100
Tax-free cash as a bridge, outside	83 of 100	93 of 100
Rolling buffer, built from the pot	84 of 100	90 of 100

At the optimistic setting the buffer built from the pension makes my plan worse, and not marginally. It drags the expected legacy down from about £1.62 million to £1.20 million in today’s money, because for thirty years it holds money in gilts that would otherwise have compounded, and pays tax to do it.

This is the honest shape of the thing. A ladder is insurance. The premium is real and you pay it every single year. In the futures where markets behave, you have paid for cover you did not need.

The gilt yield is the entire argument

If the premium is the gap between what gilts pay and what your investments would have earned, then the gilt yield decides everything. So I held my plan still and moved only that one number.

Gilt yield	Works in	Median left at 85, today’s money
1 per cent	75 of 100	£162,000
2 per cent	78 of 100	£200,000
3 per cent	80 of 100	£242,000
4.5 per cent, roughly today	84 of 100	£314,000
5.5 per cent	86 of 100	£370,000

That table is one click on your own figures, and I would run it before taking anyone's word on ladders, mine included. Details at the end.

And there is the answer to the title. I have not got a gilt ladder because for most of my adult saving life it would have been a bad idea. In 2021 a ten-year gilt paid under one per cent. Parking three years of spending in that was setting fire to money for the privilege of sleeping well.

At today's yields it is a different proposition entirely. The thing did not change. The price did. That is also why the idea has suddenly reappeared in every forum after fifteen quiet years, and why anyone repeating advice they formed in 2015 is answering a question nobody is asking now.

When should you build one? Now, later, or never?

This was the question I could not answer, because until this week my own tool insisted a ladder was built the day you stopped work. So I added the option to choose the age, and asked it properly.

When I build it	Rolling buffer, from the pot	Bridge, from the tax-free cash
Never	82 of 100	82 of 100
Day one, at 58	84 of 100	83 of 100
At 62	84 of 100	81 of 100
At 67	83 of 100	82 of 100
At 70	83 of 100	82 of 100
At 75	82 of 100	82 of 100

Two clear findings, pulling in opposite directions.

The rolling buffer is just as good built at 62 as at 58, then fades, and by 75 it is worth nothing at all. That fits the theory rather than contradicting it. A buffer only protects the years it covers, and the danger is concentrated at the start. The practical version: I can leave the pot fully invested through my late fifties and still get the protection when it counts. I did not expect that and it is the most useful thing in this post.

That question is now a setting rather than a rebuild: you can tell the model the age the ladder is built, and sweep it across a range to see the shape above on your own plan.

The bridge funded by tax-free cash behaves the other way round. Waiting actively hurts it, because the cash sits there earning market returns and then gets moved into gilts after the risky years have already passed. Build it at the start or do not bother.

What happened when I pointed it at 2008

Averages hide the thing you are insuring against, so I ran the plan into the real decade from 2008 and let the model use the actual returns and inflation for the first ten years.

Retiring into 2008	Works in	Median left at 85, today’s money
No ladder	82 of 100	£222,000
Tax-free cash as a bridge, outside	86 of 100	£263,000
Rolling buffer, built from the pot	75 of 100	£172,000

The bridge does exactly what it promises. In the one scenario built to break a retirement, it is the best of the three by a distance.

The buffer built out of the pension is worse than having no ladder at all. It pulls taxed money out of the pot in 2008 and 2009, which is precisely when those pounds were about to be worth the most, and misses the recovery with them. The insurance and the thing it was insuring cancelled out, and I paid the tax anyway.

What it does not protect you from

I ran 1973 as well, the oil shock and the stagflation that followed. Shares fell 28 per cent, then 52 per cent, while inflation peaked above 24 per cent.

My plan fails in every single future, with a ladder and without one. The ladder makes no difference whatsoever.

That is not a flaw in the model, it is the point. A gilt pays a fixed number of pounds on a fixed date. If those pounds have lost a quarter of their value by the time they arrive, the ladder has protected you from the wrong thing. It is crash insurance, not inflation insurance. If inflation is your fear, the conversation is about index-linked gilts and rising guaranteed income, and it is a different post.

You can run all of this on your own plan

Every table in this post came from running my plan over and over with one thing changed. Until this week that meant rebuilding it by hand a dozen times, which is fine for me and absurd to ask of anyone else.

So I built the thing I had been doing manually. What protects this plan? takes the plan you have already made and does two jobs.

The grid runs every defence against every crash at once. No protection, a ladder inside the pension, a ladder outside it, a spending guardrail, and both together, each against markets as you assume and against 1973, 2000 and 2008. Twenty full runs of your plan, all on the same simulated futures so a difference between two boxes is the decision and not luck. It takes a couple of seconds.

The sweep holds your plan still and moves one number through a range: growth, the gilt yield, inflation, what you spend, the age you stop, or the age the ladder is built. That is how the yield table above was made, and it is the honest way to find out how much of your plan rests on a number you cannot actually know.

Run it on your plan →

You do not enter anything new for either. It is your own plan with one thing changed, which is the only kind of comparison worth making.

I would rather you ran it than took my conclusions. Mine come from my pots, my spending, my assumption about growth and the fact that I have almost nothing outside the pension. Two of those are probably different for you, and the gilt yield you can actually get is the one that decides it.

So what am I actually going to do?

On my figures, and mine only: take the tax-free cash, put roughly three years of spending into gilts held outside the pension, and leave the rest of the pot alone. Not build a buffer out of the pension itself, which on my plan costs tax, shrinks the annuity I buy at 62, and in the one scenario I most want protection from makes things worse.

And the thing I nearly missed. My plan already cuts its spending automatically after a bad year. That guardrail is doing more of the protecting than any ladder, and when I turn it off every ladder in this post becomes roughly twice as valuable. If you would genuinely tighten your belt after a crash, you need less of this than the forums suggest. If you know you would not, you need more.

So: not too late, and not a thing I should have had all along. It is a trade that only became worth making at the yields we have now, and only in the version where the money comes from outside the pension.

If you have a ladder, or have decided against one, I would like to know which and why. Use the Contact button at the top of the page. Corrections and disagreements go in with credit, as they always do. The plan itself lives in the calculator, and the scenarios in what protects this plan?

All figures from the simulator on this site, version 1.45, four thousand simulated futures per scenario measured to age 85. Money figures are the median across those futures, converted to today’s money at the plan’s own inflation assumption, with one conversion per table so the rows are comparable. Historical sequences use FTSE All-Share total returns and ONS RPI. Gilt yield assumptions are stated in each table. Every correction made to the model is listed on the changelog.