Why Buy an Index-Linked Gilt at a 3.3% Breakeven When a 5% Fixed Bond Is Guaranteed, FSCS-Protected and Simpler?

A linker only wins if inflation outperforms what the market has already priced. The 10-year breakeven is 3.28%, from the Bank of England's yield curve — but the actual number right now is 2.9% CPI and 3.2% RPI, both published by the ONS. To beat a 5.00% fixed bond, a linker at 1.9% real needs RPI to average above 3.1% for years. That is a coin flip, priced as if it were a certainty.

Read the bars. At July's RPI of 3.2% the linker pays 5.1% — a rounding error better than the bond. At anything below 3.1% it pays less. The bond pays 5.0% in every scenario. You are taking duration risk, liquidity risk and no FSCS cover to earn an extra tenth of a point in one inflation outcome and underperform in all the others.

The breakeven is not an abstraction. It is the average inflation rate you must believe will be exceeded, for a decade, to justify the linker over a nominal bond. Today you would be paying for 3.3% when the latest print is 2.9%. The market is asking you to bet on inflation that is not happening.

The linker's 1.9% real yield is only guaranteed if you hold for a decade, and only in RPI terms. RPI is not a stable yardstick — it swung from 4.1% in March to 3.0% in April and back to 3.2% in July. Your 'guaranteed' income swings with it. If you need the money before 2036, you sell at whatever price the market sets that day. Real yields have already climbed from about 1.7% in June to 1.85% in August; a continuation of that move is a capital loss on top of everything else.

Add the three-month indexation lag and you are being paid on inflation that happened last quarter, while your bills are priced off today.

A fixed-rate bond is the opposite. The 5.00% is contractual. The maturity date is short enough to plan around. The FSCS protects £120,000 per banking licence if the provider fails. None of that depends on what RPI prints next month.

To buy the linker you first have to work out which index it tracks. Most UK linkers track RPI with a three-month lag, not the CPI in the headline. Newer issues track CPI. Some carry an eight-month lag. There is a deflation floor you hope never matters, a real coupon that looks like a rounding error, and an inflation uplift that lands as capital. That is five moving parts before you have checked the price.

The Index-Linked Gilts explainer runs thousands of words just to make the product legible. The fixed bond needs one sentence: 5.00% for three years, money back. When a product needs a manual to know what you own, you are not the one being paid for the complexity.

Then there is the practical friction. Buying an individual linker means dealing with the DMO or a broker, waiting for an auction or the secondary market, and paying dealing costs on both ends. A fixed bond opens in five minutes on an app — no dealing charge, no spread. The 'saving' the linker offers is eaten by the effort of owning it.

The linker case leans on the Bank of England possibly hiking on 17 September — three MPC members voted that way in July, with Bank Rate at 3.75%. But a hike is a bet against inflation, not for it. If the Bank raises rates to drag CPI back toward 2%, the linker's inflation uplift shrinks and its real yield could rise, hitting the price from both sides.

A hike raises the discount rate on a 10-year linker's future cash flows and can lift the real yield — the same mechanism that knocked gilt prices through the summer. You would be holding the one 'safe' asset that falls when the Bank does the right thing.

The fixed bond needs none of this to work. It pays 5.00% whether the Bank hikes, cuts, or does nothing. A saver who wants to hedge a hike has a better tool anyway: easy-access cash that reprices up with Bank Rate, not a 10-year linker that reprices down.

Be fair. If you are a higher or additional-rate taxpayer with a large balance outside an ISA, the linker's CGT-exempt uplift under section 115 of TCGA 1992 is a genuine advantage — the same point made in the 17 August gilt-versus-cash debate. If you have a known liability in a decade and want to preserve its purchasing power, a linker held to maturity does that job.

That is a narrow group. Most savers in 2026 are protecting a lump sum for one to five years, they are not sure when they need it, and they want the £120,000 FSCS cover a bond provides. For them the linker is the wrong tool at the wrong price — and the Personal Savings Allowance means most of them never even hit the tax problem the linker is built to solve.

Run both outcomes on £50,000. In a 5.00% three-year bond you get about £7,900 of interest over the term, guaranteed, and your £50,000 back. In a 10-year linker at 1.9% real you get RPI plus 1.9% — at July's RPI that is 5.1% nominal, about £2,550 a year. Marginally more, and only if RPI holds at 3.2%, only if you can wait a decade, and with the price moving against you on every hawkish headline.

The line is the argument. A 'guaranteed real return' tied to a number that moves like that is not a plan; it is a forecast wearing a guarantee. Keep the savings hub and the gilts hub side by side if you want to check the numbers yourself — you will end up back at the same conclusion: take the 5% and the FSCS cover.

If rate risk worries you, ladder three bonds of one, two and three years and roll them as they mature. You get rate certainty, annual liquidity, and the option to reinvest at higher rates if the Bank hikes — none of which a single 10-year linker offers.