All about inflation-linked bonds

Welcome to the first post of 2022!

I hope you had a COVID-free Christmas and New Year.

Unfortunately, I wasn’t one of the lucky ones, and am writing this from the confines of my second bedroom, with three days left on my isolation.

I was a victim of the ’cron.

Luckily self-isolation has meant I’ve had no excuses not to focus on blogging. I’ve now made plenty of headway into my series on bonds, and am edging closer to my next series which, by popular demand, will be on cryptocurrencies.

On the subject of cryptocurrencies, I worked out my annual return for 2021 recently, which came out at a respectable 31%.

That was a combination of 25% from my index trackers (which make up the vast majority of my portfolio), 55% from my small allocation to individual stocks, and 137% from my even smaller allocation to cryptocurrencies.

I’m sure 137% would be a pretty pedestrian return for a hardcore crypto nerd – I’m certainly not going to be buying Lamborghini any time soon – but I’m pretty happy with it. At any rate, I’m looking forward to sharing how a boring ‘TradFi’ guy (Traditional Finance guy) approaches the insane world of crypto.

Before we get to that, though, I’m sticking to my word and will be continuing with the series on bonds.

The next few posts will be all about one particular type of bond – the inflation-linked bond.

These types of bonds tend to go a bit under the radar. I’m not really sure why.

But they’re an interesting asset class, with some extremely attractive features which many investors might find useful. At the very least, we should consider them for our own portfolios, if only to rule them out.

Unfortunately, the post ran a bit long – in the region of 10,000 words – so rather than dump it on you all at once (I’m sceptical whether anyone will read 10,000 words of anything on the internet), I’ve split the post up into four parts.

Part 1 will be an introduction to inflation-linked bonds, Part 2 will look at the arguments for owning them, Part 3 will look at the arguments against them, and Part 4 will be the summary and conclusion on who should own them and why.

I use the terms ‘bond’ and ‘gilt’ interchangeably here, because I’m specifically looking at government-backed inflation-linked bonds as they provide the highest level of protection. As a rule, I’m not a fan of corporate bonds – inflation-linked or not.

Inflation-linked bonds go by a few names. You’ll also see them called index linked bonds/gilts (in the UK), and Treasury-Inflation Protected Securities (TIPS) in the US. I’ve tended to stick with ‘inflation-linked bonds’ in these posts, but also call them ‘linkers’ whenever I’ve lost the will to type out ‘inflation-linked bond’ again.

Contents

Part 1

• The difference between nominal bonds and inflation-linked bonds
• How do they work?
• The breakeven inflation rate
• Can we use the breakeven inflation rate to forecast inflation?

Part 2

• Why you should own inflation-linked bonds

Part 3

• Why you shouldn’t own inflation-linked bonds

Part 4

• Summary and conclusion

The difference between nominal bonds and inflation-linked bonds

The main difference between inflation-linked bonds and nominal bonds is inflation-linked bonds (as the name suggests) are linked to inflation.

If you hold an inflation-linked bond and inflation rises, the value of your inflation-linked bond will rise. The coupon also increases with inflation since it’s based on a higher principal value. Unlike a regular bond, where inflation eats away at the value of the bond and its coupons, the real value of an inflation-linked bond will hold steady.

Here in the UK, inflation-linked gilts (aka index linked gilts) are backed by the UK government, just like regular gilts. Our inflation-linked variety have their value tied to the Retail Price Index (RPI), which is a measure of the UK’s inflation. As RPI increases, so does the value of your inflation-linked bond.

You’ll see the yields on linkers as being much lower than the yields on nominal bonds – this is because they’ve already had inflation taken off their yields. Because inflation-linked bonds are quoted in terms of a real yield (i.e. with inflation taken off), you can’t compare them to the yield from nominal bonds (which are quoted in nominal terms – i.e. before taking inflation off). A quick-and-dirty way to estimate a nominal yield for an inflation-linked bond, so you’re comparing apples with apples, is to add inflation on to the inflation-linked bond’s real yield. This does involve some estimation error, though, which I’ll go through in a second.

How do they work?

Let’s say we invest in a 10-year inflation-linked gilt, with a par value of £1,000 and a coupon of 1%.

If, after the first year, RPI inflation is 3%, the principal value of the bond will rise from £1,000 to £1,030 (£1,000 * 1.03). The interest payment would then be £10.30 (£1,030 * 1%).

If, over the life of the bond, RPI is 20%, then the principal repayable at maturity would be £1,200 (£1,000 * 1.20).

Both the par value and the coupons have been increased by the level of inflation.

And that’s it. They’re that simple.

The breakeven inflation rate

The difference between the yield on an inflation-linked bond and the yield on a nominal bond with the same maturity is known as the breakeven inflation rate. This is the inflation rate where owning inflation-linked bonds would provide you with the same real return as owning nominal bonds.

It’s probably best explained with some simple numbers:

The expected return for a nominal bond is simply its yield to maturity (we saw that in this post on working out bond returns). In this example, it’s 1%. For the inflation-linked bond, the expected return is the yield plus the expected inflation adjustment. Notice that the expected returns between the two bonds are the same.

In this example, the breakeven inflation rate is 3%. The reason breakeven inflation rates are important is if actual inflation turns out to be higher than the breakeven inflation rate, then the inflation-linked bond will return more than the nominal bond.

If inflation is less than the breakeven rate, then it’s the nominal bond which will have the higher return.

But we’ll get more into the differences in returns between inflation-linked bonds and nominals in Part 2.

Can we use the breakeven inflation rate to forecast inflation?

A bit of a digression to end this first part of the series.

If you believe markets are pretty efficient (which we, as passive investors, generally do), then it should be possible to figure out the market’s expected level of inflation using inflation-linked bonds. And given we trust the market’s opinion more than the experts’, getting the market’s opinion is quite a useful thing to be able to do.

Because the difference in yields between nominal and inflation-linked bonds is dependent on the level of inflation, then taking the difference between the two yields should give us a good estimation of what the market thinks inflation is going to be.

This is because in an efficient market, the prices of the inflation-linked and nominal bonds should be moved by investors in such a way that there’s no return difference between buying an inflation-linked bond compared to a nominal bond.

Using our example from before, if a 10-year inflation-linked bond has a yield of -2% and the market believes inflation is likely to be 3%, then the nominal yield on an inflation-linked bond is 1%. This means the prices of nominal bonds will be moved by the market in order for their yields to also equal 1%.

If the yields on nominal bonds were higher than 1%, then investors would sell their inflation-linked bonds and buy nominal bonds. This would reduce the price of the inflation-linked bonds (which increases their yield), and increase the price of nominal bonds (which decreases their yield). The end result is that inflation-linked bonds and nominal bonds should end up with the same real yields.

Because we know the real yields on both bonds should be equal, we can take the yield from a nominal bond, deduct the yield from an inflation-linked bond of equal maturity, and work out what the market believes inflation is likely to be over the life of the bonds.

Let’s take a real life example.

The yield on the Vanguard UK Inflation-Linked Gilt Index Fund at the time of writing (January 2022) is -2.8%. Vanguard have a nominal gilt fund with approximately the same duration (22 vs 21.7), which is the Vanguard UK Long Duration Gilt Index Fund, and it has a yield of 0.9%. By taking one from the other, we can work out that the market expects inflation to be roughly 3.7%.

But this approach isn’t perfect.

There are a few reasons why you can’t assume the inflation rate implied by the yield spread between inflation-linked and nominal bonds is the market’s best guess. Each of these three reasons affects the yield of either nominals or linkers, without having anything to do with inflation expectations:

1. Liquidity

The first is that nominal government bonds are more liquid than inflation-linked bonds. While linkers are still relatively liquid securities, and also carry the full faith and credit of the government, they aren’t as liquid as nominal bonds. Thus, investors in nominal bonds pay a premium to own them, which reduces their yield.

2. Premium for inflation protection.

Holders of inflation-linked bonds also pay a premium, but this premium is for the safety of being protected against any inflation.

We’ll get into the details of this in the second post, but that inflation protection means inflation-linked bonds offer unlimited potential outperformance over nominal bonds in inflationary environments, while regular bonds only offer limited potential outperformance over linkers in deflationary environments. The limited nominal outperformance during deflation is because linkers have a deflation floor. At maturity, an inflation-linked bond investor receives the higher of the adjusted principal value or the original principal value.

This is an attractive feature of inflation-linked bonds, and the holders therefore pay a premium for it in the form of lower yields.

3. Structural

The third non-inflation related force being exerted on nominal/inflation-linked bond yields is more structural in nature.

Legislation in 2004 obliged UK pension funds to match their assets to their long-term promises. This in turn spurred demand for long-dated inflation-linked bonds, and the debt-management office issued lots more of them. Despite this issuance, the demand for inflation protection has over time driven real yields down to unusually low levels and pushed up break-evens.

Taking these three factors together, we can see a couple of opposing forces at play.

The liquidity premium increases yields on inflation-linked bonds, whereas the risk premium for unexpected inflation plus buying pressure from pension funds both depress yields for linkers. Because these forces work in opposite directions, they may cancel each other out. If that is the case, the TIPS-to-nominal-bond spread is a good indicator of the market’s aggregate view of expected inflation – but it’s impossible to work out the exact effect on yields of each of the three components, so impossible to say whether they cancel out or not. Hence why the implied level of inflation from the breakeven inflation rate should be a rough estimate only.

—

That was a lot of information to take in. If you’re still with me, well done!

It might have come across as a bit of an information dump, but this first part has laid the groundwork for the series.

Now we understand what inflation-linked bonds are and how they work, the next post will focus on why you might want to include them in your own portfolio.