Short answer: once you have decided to hold gilts rather than a bond fund, there is still a shape to choose. A ladder spreads the money evenly, one gilt maturing each year. A bullet puts it all in one maturity. A barbell splits it between very short and very long gilts with nothing in the middle. Built to the same average duration, all three gain or lose about the same when every yield moves together — so the choice is not about how much interest-rate risk you take, but about when the cash comes back and which part of the curve you are exposed to. The live example below shows the three side by side, priced off today's curve.
The three shapes, in plain English
Ladder. Ten gilts, one maturing in each of the next ten years, roughly £10,000 in each. Every year one repays and you either spend it or buy a new ten-year gilt at the far end. Nothing has to be sold; the income is known; the average yield is the curve's average over the ten years. This is the shape the gilt ladder guide and calculator build, because it is the one that pays out on a schedule.
Bullet. The whole £100,000 in one gilt, maturing on one date — five or six years out in the example. Simplest to buy, one dealing charge, one line on the statement, and exactly the right shape when the money is needed all at once: a house move, a lump sum to pay off a mortgage, a known bill. Between now and then its price will move with the yield for that maturity and nothing else.
Barbell. Half in a gilt maturing within a year or two, half in a long one — fifteen, twenty, thirty years out — and nothing in between. The short half is nearly cash: it repays soon and can be rolled or spent. The long half carries almost all the interest-rate exposure. Bond managers like the shape because a barbell has a little more convexity than a bullet of the same duration: it gains slightly more when yields fall and loses slightly less when they rise. For a private investor the more important feature is the cash at the short end.
£100,000 today — the live example
Ladder — £10,000 in 4¼% Treasury Gilt 2027 (matures 2027-12, yield 4.52%, duration 1.1); £10,000 in 1 5/8% Treasury Gilt 2028 (matures 2028-10, yield 4.72%, duration 1.9); £10,000 in 0 7/8% Treasury Gilt 2029 (matures 2029-10, yield 4.83%, duration 2.9); £10,000 in 0 3/8% Treasury Gilt 2030 (matures 2030-10, yield 4.91%, duration 3.9); £10,000 in 4% Treasury Gilt 2031 (matures 2031-10, yield 4.97%, duration 4.4); £10,000 in 3¼% Treasury Gilt 2033 (matures 2033-01, yield 5.08%, duration 5.5); £10,000 in 0 7/8% Green Gilt 2033 (matures 2033-07, yield 5.13%, duration 6.4); £10,000 in 4½% Treasury Gilt 2034 (matures 2034-09, yield 5.20%, duration 6.5); £10,000 in 4¾% Treasury Gilt 2035 (matures 2035-10, yield 5.28%, duration 7.0); £10,000 in 4 7/8% Treasury Gilt 2036 (matures 2036-07, yield 5.34%, duration 7.6). Yield on the £100,000: 5.00% a year; modified duration 4.7.
Bullet — £100,000 in 4 5/8% Treasury Gilt 2032 (matures 2032-03, yield 5.00%, duration 4.7). Yield on the £100,000: 5.00% a year; modified duration 4.7.
Barbell — £50,000 in 4 3/8% Treasury Gilt 2028 (matures 2028-03, yield 4.59%, duration 1.3); £50,000 in 4 5/8% Green Gilt 2037 (matures 2037-03, yield 5.39%, duration 8.0). Yield on the £100,000: 4.99% a year; modified duration 4.7.
| If, the day after buying… | Ladder | Bullet | Barbell |
|---|---|---|---|
| All yields fall 1.5 points | +£7,088 | +£7,035 | +£7,025 |
| All yields rise 1 point | −£4,725 | −£4,690 | −£4,683 |
| Curve steepens: two-year yields fall 0.5, ten-year and longer rise 0.5 | −£455 | +£355 | −£1,675 |
First-order arithmetic: each gilt's gain or loss is its modified duration × the move in its own yield × the amount held. Convexity would make every gain a little larger and every loss a little smaller, most of all for the barbell's long half. Held to maturity, none of these marks is realised — each gilt still repays £100.
Read the table across a row first. When all yields move by the same amount, the three shapes move together, because they were built to the same duration — that is the point of the exercise, not a coincidence. Then read the last row: when the curve steepens — short yields down, long yields up, pivoting around six years — the shapes part company. The bullet, sitting near the pivot, is roughly untouched; the barbell's long half is out where yields rose and takes the hit while its short half gains a little; the ladder, holding every rung, lands between them. Reverse the move (a flattening) and the ordering reverses too.
What each shape is really a bet on
A bullet is a bet on one point of the curve: only the yield for its own maturity matters, and it is the cleanest way to match a single known date. A ladder is a bet on the level of the curve spread evenly across ten points, with a cash flow every year — it is the shape that turns gilts into an income, and the shape that cares least about the curve's twists because it holds every part of it. A barbell is a bet on the shape: it profits when the curve flattens (long yields fall relative to short ones) and loses when it steepens, and it keeps cash coming back at the short end while the long half does the work. All three, held to maturity, deliver the yields they were bought at; the table above is what shows on the statement in between, and it only becomes real if you sell.
Two calculators do this arithmetic on your own figures: the gilt ladder prices a ladder of any length off today's curve with the after-tax comparison against savings interest, and what does a 1% rate rise cost me? turns any single gilt's duration — or a bond fund's — into pounds for rises and falls of ½, 1 and 2 points.
Which is right depends on…
- When the money is needed. All on one date: a bullet. Some every year: a ladder. Not sure, but possibly soon: the barbell's short half is the answer to "possibly soon", and its long half is a separate decision.
- Whether you would ever sell early. If not, the scenario table is noise and only the yields matter. If you might, the long half of a barbell is where the large losses (and gains) live; a ladder's are spread and shrink every year as the rungs shorten.
- Tax. Outside an ISA or pension only the coupon is taxed and the gain to £100 is not, so low-coupon gilts win in every shape; which gilts happen to be low-coupon at each maturity is on every gilt.
- Effort. A bullet is one purchase; a ladder is ten and a decision every year; a barbell is two but with a long gilt whose price will swing more than most people expect.
None of this ranks the three. The planner's cash bucket and the ladder calculator assume the ladder because it is the shape that produces an income; the other two are for other jobs.
Questions people ask
Do the three really have the same risk? The same duration, so the same first-order sensitivity to a parallel move in yields — the table shows it. They differ in convexity (the barbell has slightly more), in exposure to the curve's shape, and in when cash returns. "Risk" for a gilt held to maturity is neither: it is only the risk of needing the money before the date.
Why not just buy a bond fund? A fund holds hundreds of gilts and never matures, so it is a perpetual ladder that keeps rolling: convenient, but its value moves with rates for ever and there is no date on which you are made whole. The three shapes here all end.
What does "curve steepens" mean? Short yields fall relative to long ones — typically when Bank Rate is cut while long-term inflation expectations stay put. Flattening is the reverse. The scenario in the table moves two-year yields down half a point and ten-year and longer yields up half a point, with a straight line between, so the pivot is at about six years.
Where do the figures come from? The gilts are the DMO's register of gilts in issue; the yields and durations are computed each weekday off the Bank of England's spot curve, the same figures as every gilt; the scenario outcomes are modified duration × the move × the amount, first-order, with convexity left out and said so.