A holding of individual inflation-tracking UK government bonds is the way forward if you want an asset class to hedge UK inflation. But how do you actually buy index-linked gilts?

Thankfully, your neighbourhood investment blogger is here to clear that up.

I’ve personally been pushing this task around my own plate like a seven-year-old told to eat his greens, due to…

• Telephone dealing: “I hate you!”
• Dirty pricing versus clean pricing faffology: “Stoopid!”
• Accrued interest deductions: “Don’t wanna!”

In short, buying individual index-linked gilts meant dealing with the unfamiliar and, as far as I could tell, deeply sucky.

I put the task off for months. Yet now I’ve done it, it doesn’t seem so bad after all.

I suspect I’m not the only one discouraged by mental barriers when looking to buy index-linked gilts.

And so today I’ll walk you through my recent index-linked gilt transaction to demystify the process. I’ll explain any important mechanics as we go, and we can sort any remaining bafflement in the comments.

Missing link(er)

First challenge: not every broker allows you to trade individual gilts.

Of those that do, some enable you to trade at the click of a button, others make you speak to another human at the end of a telephone. (What is this? The Dark Ages?)

Even then your broker may not trade every bond you want, or it may not trade every bond online.

I diversify across two brokers. Of those, only AJ Bell lets me invest in individual gilts.

Thankfully, AJ further enables me to click-to-buy all but two of the UK linkers currently on the secondary market.

No humans required!

If you’re building an index-linked gilt ladder, know that only the 2033 and 2054 rungs are missing from AJ Bell’s roster. (And it might let you buy these by phone too. I’m not sure.)

As it is, I’m building a short-dated rolling linker ladder as modelled in the No Cat Food decumulation portfolio.

How to buy index-linked gilts, step by step

My objective is to keep a portion of my SIPP in a very low-risk, inflation-hedging asset. Three years’ worth of index-linked gilts fits the bill nicely.

Let’s get on with it!

Step one: free up some cash

I flogged off my incumbent global inflation-linked bond ETF (GISG). It’s the best passive short-dated linker fund available in my view, but it still suffered a real terms loss in 2022.

Step two: choose your individual linkers

My rolling linker ladder will consist of three index-linked gilts, ideally maturing in one, two, and three years.

Assuming I don’t need the dosh, then I’ll annually reinvest the cash I get from the latest maturing gilt into a new linker with three years left on the clock.

The snag is there isn’t a linker maturing in 2025. So my first three picks will redeem from 2026 to 2028.

With that decided, the choice is simple as there’s only one linker available per year:

No two organisations label their linkers exactly the same way. Search for – and double-check you’ve found – the right security by using its EPIC or ISIN code.

Once surfaced, you can click-through to trade your gilt – assuming your broker is on the grid.

Otherwise, it’s the telephone, or postal order, or semaphore trading for you m’lad / lass.

Brokers who facilitate online gilt trading

^{Disclosure: Links to platforms may be affiliate links, where we may earn a small commission. It doesn’t affect the price you pay. Your capital is at risk when you invest.}

AJ Bell lists its gilt line-up on a specific page.

Hargreaves Lansdown also has a dedicated linker page. Click the Maturity header to place them in a sane running order. But beware, most of HL’s linkers apparently require an expensive telephone trade. See this super-helpful comment from reader Delta Hedge.

iWeb lists linkers too. (This page appears organised by the Muppet Show. Click through on the names to trade.)

Halifax and Lloyds use the same platform as iWeb but in nicer colours.

Interactive Investor trades linkers online but I can’t find a public-facing page. Individual conventional gilts are listed though. You can find index-linked gilts on ii by searching using EPIC codes.

Charles Stanley trades gilts but it looks like a telephone-only service.

Fidelity is an obvious absentee here. Sort it out Fidelity!

Let us know of any other brokers you use in the comments.

Step three: understand how individual linkers are priced

Things can get pretty confusing because of the way index-linked gilts are priced.

Most brokers and online data feeds show each linker’s clean price before you order.

The clean price is typically the nominal price for each gilt.

That isn’t much use because the price you pay is the dirty price.

The dirty price is typically higher than the clean price. That’s because the clean price excludes:

• Inflation-adjusted principal and accrued interest for three-month indexation lag linkers.
• Accrued interest for eight-month indexation lag linkers.

Bear with!

Inflation-adjusted principal

Inflation-adjusted principal is the bond’s original £100 nominal value modified by the change in the RPI index since it was first issued.

In other words, if RPI inflation has increased by 10% since the linker hit the market, the value of its principal will have increased to £110.

It’s this inflation tracking property that makes linkers so valuable in the first place! (Along with their inflation-adjusted coupon or interest payments)

The clean price does not include inflation uplift on principal for most linkers, whereas the dirty price does.

While we’re here, I’ll just mention that all index-linked gilts are due to switch their link from RPI inflation to CPIH inflation from 2030.

Also while we’re here, bonds are a psychological hellscape of impenetrable jargon. Take the edge off it with our bond terms pain relief.

Also this Debt Management Office (DMO) glossary is a godsend.

Accrued interest

The dirty price includes inflation-adjusted accrued interest. Accrued interest is interest you’ve earned from owning the bond since its last coupon date.

By rights, that accrued interest belongs to the seller who held the bond until you swooped in.

Paying the dirty price (pumped up by the accrued interest) means you compensate the seller for the interest payment they won’t receive – because you now own the bond.

It’s a bit like pass the parcel. The previous owner handed the bond on to you while the music still played. And if you’re still in possession when the music stops, you scoop the whole prize – a semi-annual interest payment no less.

Thankfully, bond traders recognise that a children’s party game is no basis on which to build a thriving capital market. Thus accrued interest keeps everything fair and avoids foot-stamping temper tantrums.

This is also why bond trader parties are no fun.

Your broker will show accrued interest as a cost when you buy a gilt. You’ll make it back next time your linker deposits sweet, sweet income into your account. If you decide to sell a bond early, then someone will pay you any accrued interest in return.

Ownership of the gilt is determined seven business days before each coupon payment date. That seven day stretch is the ex-dividend period – beginning with the ex-dividend date.

If your purchase settles during that period (but not including the ex-dividend date itself) then you don’t pay accrued interest. Instead, you’re entitled to rebate interest. This will show as a Brucie bonus on your contract note.

What’s actually happened is that the seller has already been declared the winner of the next coupon. So if, for example, you take ownership of the gilt on the first day of the ex-dividend period, they owe you for the seven days of interest earned before the coupon paid out.

Just like accrued interest, rebate interest is a ‘fair’s fair’ mechanism. It ensures each party earns the right amount of interest for their period of ownership, regardless of where the coupon apples actually fall.

Fun fact: if your trade settles on the coupon payment date then there is no accrued interest (or rebate interest). Yin and Yang are in balance on this day.

Indexation lag

Eight-month indexation lag linkers upweight principal and coupon using RPI readings from eight months ago. For example, a coupon paid out in December is inflation-adjusted according to the previous April’s RPI index.

Eight-monthers are very much an endangered species. They were issued before 2005 and as mentioned only two remain in circulation: T30I maturing in 2030 and T2IL maturing in 2035.

Three-month indexation laggers represent the latest in UK linker engineering. They only trail inflation by three months.

Under the pricing bonnet, eight-month clean prices include inflation-adjusted principal and three-monthers do not. That’s why eight-monthers look more expensive at first blush.

In reality, it makes no difference. All gilts are bought at the dirty price and if you want a linker that matures in 2030 and 2035 then it’s an eight-monther for you.

Why don’t they show the dirty price?

God knows. It’s not as if they don’t calculate it when you make a purchase. Perhaps someone who knows about the live price plumbing can supply an answer. But it’s an annoying omission.

It’s also the reason why some brokers ask you to state a cash amount when ordering linkers rather than a unit number.

If you’re building a non-rolling linker ladder predicated on buying a certain number of gilts then it’s probably best to over-egg it.

That said, here are three sources of dirty price information:

• Tradeweb – Sign up for a free account. Select Index-linked in the Security Type menu and press Submit. Set the Page size to 50 to see every linker on the market.
• YieldGimp – Dirty price = Net Price (inc. Accrued) column on the spreadsheet.
• LateGenXer – Scroll down and switch on the Index-linked toggle in the left-hand column. Enjoy dirty prices!

Tradeweb is the official supplier of gilt stats to the DMO. However, it only provides the closing dirty price, which it publishes around noon the following day.

YieldGimp updates its dirty prices throughout the day, so this is your go-to source if you want a rough and ready take on how many gilts you can expect to purchase. It won’t be spot-on, as we’ll see shortly. But it’ll be pretty close.

LateGenXer has developed a superb app to help UK investors build linker ladders. The dirty price is updated towards the end of the day. Extend the ‘Number of years’ in the left-hand column to see more linkers.

You can also calculate the dirty price from the clean price on the fly. Updated clean prices are available from the London Stock Exchange. Search using EPIC or ISIN codes.

It requires some spreadsheet kung-fu to beat the dirty price out of the clean price, so we’ll save that for the next thrilling episode of Arthur C. Accumulator’s Mysterious World (of linkers).

Units vs gilts

Okay, one last point on the linker pricing imbroglio.

Gilt prices are typically displayed in pounds not pence. If you see a two or three figure price then that’s the price in pounds per gilt unless it says otherwise.

Tradeweb, YieldGimp, and the London Stock Exchange display prices like this.

The brokers generally do the same. Until they don’t.

Now, just in case you were finding all this too easy, you don’t buy gilts in handy bundles of gilts.

You buy them in units. Each unit is worth a hundredth of gilt.

So if a gilt has a nominal value of £100 then each unit has a nominal value of £1.

Which sounds simple enough but we’re all busy people and it’s easy to forget.

Especially when your broker mixes unit values with gilt prices!

Here are the crazy scenes in my account:

I’ve bought 14,850 units of mystery brand linker A. But my mischievous broker displays the gilt price not the unit price.

• 14,850 x £148.8817 = £2,210,893

I’m rich! Oh balls, I’m not rich. I just put the decimal place in the wrong column again.

The unit price is £1.488817 because each unit is worth one-hundredth of a gilt. Which explains why the value column is £22,108.93 and I haven’t bought a one-hundred bagger linker.

A single-figure price typically indicates a unit price. A two- to three-figure price suggests gilts, unless some eejit is showing you the price in pence, which some brokers randomly do. Good to keep you on your toes!

If you track your linker winnings on a spreadsheet and something isn’t adding up, then this units/gilt farce will often be the reason. At least it is for me.

Coupons, accrued interest, you name it – the amounts are typically quoted in pounds per gilt, so should be multiplied by your units / 100 when you’re totting them up.

Step four: lose the will to live

Revive with a coffee, a beer, or a fortifying hot chocolate to suit.

Step five: submit your order

I can’t believe it! I’m submitting my order already. So soon?

As I mentioned, nobody knows what the hell price they’re be paying so you’ll be asked to put cash on the table.

Once I did that with my trade, I was treated to this quote screen:

The clean price is just so much screen clutter. Fuggedaboudit.

Although that said, the £1 difference between the clean buy and sell price shows that you may pay a spread of about 10p per unit.

The indicative price is per unit and wasn’t too far out. I’ll explain more about this price in a sec as it’s dirty-ish but not strictly dirty.

The dealing charge was a fiver and very reasonable too. It works out at less than 0.023% of the transaction.

The order type was a market order or a limit order. In the end, I went for a market order.

Notice the small print that says: “Accrued interest payments will also be applied to the estimated total.” The bill for that is coming right up.

Anyway, dear reader, I submitted my order.

Telephone orders

I have not made a telephone order, but Monevator readers Mark Dawse and Sleepingdogs, among others, have reported on the process:

• Know which index-linked gilts you want to order in advance.
• Identify each one by their EPIC code. It’s much easier than using the longer ISIN number, and will knock precious minutes off the call!
• The broker’s agent will repeat back the gilt’s code and other identifying details to ensure you’re both talking about the same thing.
• They should quote the fee and an indicative price. You then confirm whether you wish to proceed.
• The agent is likely to put you on hold while their team places the trades.
• Once all trades have gone through, your agent will list your purchases and the actual prices paid.
• Set aside plenty of time for the call, especially if you’re placing several orders in one go.

Step six: “Congratulations on your purchase of UK government debt”

Here are my contract note highlights (never thought I’d find myself saying that):

After the sale, you’ll finally know how many units you’ve bought.

That’s 14,850 – or 148.5 gilts – in this case.

The price per unit is higher than indicated on the quote screen. No biggie.

The consideration number tells me I’ve bought £22,094.15 of linker TR26. (Our mystery brand revealed!)

And I owe 23p in accrued interest. Could be worse.

Notice how accrued interest is a cost on top, like the dealing charge.

Step seven: incur an immediate loss

Every time you buy individual gilts, your broker is likely to show you’ve made an initial loss (unless the price moves sharply in your favour).

Here’s the losses weighing on my three linkers shortly after purchase:

There’s nothing like getting off to a great start, right?

The loss comprises:

• A £5 dealing charge
• Accrued interest
• Unfavourable price moves since purchase

By the time I took this shot, the 2026 linker (TR26) was down 56p price-wise. Meanwhile, the prices of the other two were up £6.59 and £4.54 respectively, reducing my initial losses.

Fam, it’s a rollercoaster.

You may show a much worse loss if your broker values your linkers using the nominal clean price.

If AJ Bell did that then my valuation would have looked approximately like this:

My holding would have appeared down by nearly £25,000 if it was valued by the clean price. (Remember the clean price is divided by 100 to get the unit price).

If you are seeing massive losses like that then there’s almost certainly no cause for alarm. (Assuming they’re caused by the clean price method which they probably are.)

Your index-linked gilts are actually valued by the dirty price. This includes all that lovely inflation uplift and accrued interest.

I’ll include a spreadsheet in the next part of this series so you can properly track the value of your holdings using the intra-day dirty price.

Inflation-adjusted clean price and this accrued interest business

Although AJ Bell isn’t valuing my linkers by the nominal clean price I don’t think it’s using the dirty price either.

If it was, then my portfolio wouldn’t show a loss due to accrued interest – because accrued interest is included in the dirty price.

So it must be valuing my units by the inflation-adjusted clean price. That is:

The dirty price minus accrued interest. Or, in other words, the clean price incorporating inflation-adjusted principal. 

Thus my linkers should be worth a little bit more than shown in the last screenshot above. Because if I sold them immediately after purchase, I’d be due the accrued interest I’d bought, but never received, because I sold out before the next coupon payment.

It’s all relatively easy to calculate but let’s leave it for the spreadsheet episode to come. Tradeweb also publishes accrued interest figures per gilt (see the link waaaaay above.)

Step eight: stop writing about index-linked gilts

Don’t mind if I do.

Hope this all helps someone.

Take it steady,

The Accumulator

When someone hears I’m currently writing the authorized biography of William (Bill) Sharpe, the most frequent question I get is, “Is he still alive?” Sharpe is the 1990 recipient of the Sveriges Riksbank Prize in Economic Sciences in Memory of Alfred Nobel, commonly known as the Nobel Prize in Economics. And, yes, in September 2024, he is still alive and well. He lives in Carmel-by-the-Sea in California. Every Thursday morning, he meets with his coffee klatch. He can often be seen walking his bichon-poodle near Carmel Bay. In June 2024, he celebrated his 90^th birthday.

And September 2024 was another Sharpe milestone: the 60th anniversary of his seminal capital asset pricing model (CAPM) paper in The Journal of Finance. It is extremely rare for research to remain relevant after a decade let alone six. I’ll explain what the paper is about, how it impacted the investment industry, most likely including your own portfolio, and why it still matters.

Photo by Stephen R. Foerster

The C-A-P-M

Let’s talk about the model’s name, common acronym, and what it’s really about. First, Sharpe never called it the “capital asset pricing model.” As the title of his seminal article indicates, it’s about “capital asset prices.” Later researchers referred to it as a model, adding the M. Second, once it became known as the capital asset pricing model, it was referred to by the acronym CAPM, pronounced “cap-em.”

Virtually every finance professor and student refer to it as “cap-em” — everyone except Sharpe himself. He always uses the initialism C-A-P-M. (So, if you want to honor the creator of the model, you can refer to it as the C-A-P-M!) Third, the focus isn’t really about prices of assets, but rather their expected returns. One of the key insights of the CAPM is that it answers an important investment question: “What is the expected return if I purchase security XYZ?”

Key Assumptions

Sharpe had written a paper published in 1963, “A Simplified Model for Portfolio Analysis,” that presented some of the same key concepts as in the seminal 1964 paper. There is an important difference between the two papers. As Sharpe later described it, in the 1963 paper, he carefully “put the rabbit in the hat” before pulling it out. The 1963 paper also answered that key question, “What is the expected return if I purchase security XYZ?”

But the rabbit he put in the hat was a preordained relationship between a security and the overall market — what I’ll describe later as beta. Andrew Lo and I interviewed Sharpe for our book, In Pursuit of the Perfect Portfolio: The Stories, Voices, and Key Insights of the Pioneers Who Shaped the Way We Invest. “So, I spent several months trying to figure out how to do it without putting the rabbit in the hat,” he said. “Was there a way to pull the rabbit out of the hat without putting it in to begin with? I figured out yes, there was.” In the 1964 article, Sharpe didn’t put a rabbit in the hat but rather he derived a market equilibrium based on theory.

With any theory, you need to make assumptions, to simplify what happens in the real world, so that you can get traction with the theoretical model. That’s what Sharpe did. He assumed that all that investors care about are expected returns and risk. He assumed investors were rational and well-diversified. And he assumed investors could borrow and lend and the same rate.

When Sharpe initially submitted the paper for publication in The Journal of Finance, it was rejected, mainly because of Sharpe’s assumptions. The anonymous referee concluded that the assumptions Sharpe had made were so “preposterous” that all subsequent conclusions were “uninteresting.” Undeterred, two years later Sharpe made some paper tweaks, found a new editor, and the paper was published. The rest, as they say, is history.

The CAPM in Pictures

Much of Sharpe’s classic paper focuses on nine figures or graphs. The first seven are in two-dimensional space, with risk — as measured by the standard deviation of expected returns — on the vertical axis and expected return on the horizontal axis. (Any finance student will quickly note that the now-common practice is to flip axes, which is represent risk on the horizontal axis and expected return on the vertical axis.)

On his horizontal axis, Sharpe began with the return on a special security that he called the “pure interest rate” or P. Today, we would refer to that special rate as the Treasury Bill return, or the risk-free rate, commonly represented as Rf.

The curve igg’ is Harry Markowitz’s efficient frontier: the “optimal” combination of risky securities such that each portfolio on the curve has the highest expected return for a given level of risk, and also the lowest risk for a given level of expected return. Sharpe’s model essentially looked for combinations of the risk-free security, P, with each portfolio on the curve igg’ that would provide the optimal risk-expected return. It is clear from the graph that the optimal mix is formed by a line from P that is tangent to curve igg’ — in other words, the mix that combines the risk-free asset P and portfolio g.

In Sharpe’s world, we can think of the investor as essentially having three choices. She can invest all of her money in risky portfolio g. If that’s too much risk for her, she can divide her portfolio between combinations of risk-free P and risky g. Or, if she wants even more risk she can borrow at the risk-free rate and invest more than 100% of her wealth in risky g, essentially moving along the line toward Z. The line PgZ is Sharpe’s famous Capital Market Line, showing the optimal combination of risk-free and risky investments, including either lending (buying a Treasury Bill) or borrowing (at the Treasury Bill rate).

The Footnote that Won a Nobel Prize

After presenting a series of graphs, Sharpe showed how this could lead to “a relatively simple formula which relates the expected rate of return to various elements of risk for all assets which are included in combination g. He then refers the reader to his footnote 22, an extensive 17 lines of equations and text that may be one of the most consequential footnotes in all of finance and economics literature.

That last line of the footnote may not look familiar, but with a bit of sleight-of-hand it will come into focus. Sharpe gave the left-hand-side a new name: Big, with “ig” as the subscript. In technical terms, Big is the covariance of the return on security i relative to security g, divided by the standard deviation of g. When creating the manuscript, Sharpe used a typewriter, with standard keys. What he really meant by B was the Greek letter b or beta. And as we’ll see, that has become one of the most used measures of risk today.

What Drives Expected Returns?

One of the key insights from Sharpe’s model is that when it comes to a security’s expected return, all that matters is Big, or beta.

In Sharpe’s final graph, expected return is still on the horizontal axis, but his new measure of risk, Big or beta, is on the vertical axis. Now the line PQ is actual the CAPM equation. What it powerfully shows is that, assuming an investor holds a well-diversified portfolio, the only measure of risk that matters is beta, or how risky the security is relative to the overall portfolio g. Since all investors want to hold g, then it must contain all assets. In other words, it must be the market portfolio. Today, we call that portfolio M.

We can now re-write Sharpe’s original derivation of the CAPM to the more-familiar version: E(Ri) = Rf + b x [E(Rm) – Rf] or E(Ri) = Rf + bi x MRP, where i represents security i and MRP is the market risk premium. Here’s the intuition. Let’s suppose you’re considering investing in a stock for the next 10 years — or maybe not. Alternatively, you could invest in long-term Treasuries and secure a return of Rf. Or you could invest in the market as a whole and get an expected return of E(Rm). That works out to be the same as Rf + MRP. Or finally, you could invest in security i. Your expected return, E(Ri) would be driven by how much market risk you are exposed to, bi.

Beta has a simple interpretation: how risky a particular security is relative to the overall market. In terms of benchmarks, by definition “the market” has a beta of 1.0. For a particular security, beta suggests what the particular return change is for every 1.0% change in the market. For example, for a low-risk stock with a beta of 0.5, if the market (often proxied as the S&P 500 Index) goes up by 1.0 percent, we would expect stock i to go up by 0.5 percent; if the market is down by 1.0%, we expect stock i to go down by 0.5 percent. The same logic holds for a risky stock, say with a beta of 1.5. If the market goes up by 1.0%, we would expect stock i to go up by 1.5%. If the market is down by 1.0 percent, we expect stock i to go down by 1.5%.

Why the CAPM Still Matters

Sharpe’s seminal 1964 paper matters for three reasons.

1. Beta is the appropriate measure of risk for a stock that is part of a diversified portfolio. It is also a widely available measure, on sites such as Yahoo!Finance. All that matters is risk relative to the market. If you have a diversified portfolio, it doesn’t matter how volatile a stock is on its own.
2. Sharpe’s model, and in some sense Figure 7, shows us a way to measure performance across well-diversified portfolios such as mutual funds. We can measure a fund’s performance or return, say over the past five years, in excess of what a risk-free investment would have returned. That’s the return measure. If we compare that to the fund’s risk, as measured by the standard deviation of the fund’s return over that period, we have a return-to-risk measure. That’s what Sharpe described in subsequent research papers and became known as the Sharpe ratio. It’s probably the most common measure of performance today.
3. In Sharpe’s CAPM paper, he defined his special portfolio, g, the one that everyone would want to hold, as one that represented “all assets.” That’s why we call it the market portfolio. In a narrower interpretation, it should at least contain all stocks. Specific to the United States, that implies buying an index fund like one that replicates the S&P 500 Index. We have Sharpe’s model to thank for the multi-trillion-dollar index fund that has emerged over the past 50 years. Chances are that you’re invested in an index fund, either directly or indirectly, say through a pension fund.

Of course, the CAPM has its critics. There are some competing models of expected return that capture additional factors beyond the market. There are some questionable empirical test results. And yet, the model is still front and center in finance courses and still used by practitioners. And it’s a very intuitive model. It has stood the test of time.

So please join me in wishing the CAPM a happy birthday, with many more to come!