The failure of index-linked bond funds to perform post-Covid has really been bothering me. What’s the point of these things if they don’t actually protect you from inflation? Meanwhile, individual index-linked gilts – correctly used – are meant to be a proper inflation hedge. But is that true?
Can we empirically prove individual linkers1 worked when inflation let rip?
First, some context. Our favoured linker fund holding at House Monevator prior to the post-pandemic price surge was a short-duration model. That’s because short-duration index-linked fund returns are more likely to reflect their bonds’ inflation ratchets, and are less prone to price convulsions triggered by rocketing interest rates.
Longer duration linker funds, meanwhile, got hammered in 2022 because they’re more vulnerable to rising interest rates. When rates soared, prices dropped so hard and fast that their bond’s inflation-adjustment element was rendered as effective as wellies in a tsunami.
Hopefully you at least avoided that fate…
The weakest link(ers)
So it’s October 2021, and you’re duly positioned on the coastline, scanning the horizon for inflation, with ample resources invested in short-duration linker bond fund units.
Here’s how our defences performed once the inflation Kaiju was unleashed:
Inflation versus short-duration linker fund
Index-linked bond fund is the GISG ETF. Data from JustETF and ONS. February 2025. NB. The linker fund trend line was corrected on 18 Feb 2025.
Oh. As that calm-voiced announcer-of-doom on Grandstand might have intoned: “Inflation One, Passive Investing Defence Force, Nil.”
Or, in numbers more appropriate to an investing article, the annualised returns from October 2021 (when inflation lifted off) to year-end 2024 are:
- UK CPI inflation: 5.9%
- Short-duration linker fund: 0.6%
Note: all returns in this article are nominal, dividends reinvested.
In other words, this linker fund fell far behind rising inflation and posted real-terms losses over the period.
Right-ho. So that was a learning curve.
Since then I’ve put a lot of time into researching individual index-linked gilts, commodities, gold and money market funds – all assets fancied as offering some degree of inflation protection.
The most reliable should be individual index-linked gilts. After all, they come with UK inflation-suppression built-in. Put your cash in, and it pops out at maturity, with a price-adjusted enamel on top. Purchasing power protected!
All you must do is not sell your linkers before maturity. Buying-and-holding prevents the kind of losses bond funds are vulnerable to realising. Funds’ constant duration mandates make them forced sellers when bond prices are down.
Excelente! But one thing was still nagging me. Did individual linkers actually deliver on their inflation-hedging promise during the recent price spiral?
Inflation versus individual index-linked gilts
To answer that question, I simulated the performance of a small portfolio of individual index-linked gilts using price and dividend data from October 2021 to year-end 2024.
Then I pitted the individual linkers against CPI inflation and GISG, the short-duration linker ETF discussed above.
Here’s the chart:
Data from JustETF, Tradeweb and ONS. February 2025. NB. The linker fund trend line was corrected on 18 Feb 2025.
Okay, the individual linkers (pink line) did better than the fund but they still lagged inflation. The annualised return numbers are:
- Inflation: 5.9%
- Individual linkers: 4.1%
- Linker fund: 0.6%
That’s still an unhealthy gap as far as I’m concerned – like buying a peep-hole bulletproof vest.
Proving a negative
Why did the individual index-linked gilts lose money versus inflation?
Because way back in 2021 they were saddled with negative yields. That is, the buy-in price for linkers was so high that their remaining cashflows were guaranteed to sock you with a loss, if you held them until maturity.
The best a linker portfolio held to maturity could do was limit the damage against inflation. But that negative yield drag meant it was always going to underperform.
But that’s a historical problem. Today index-linked gilts are priced on positive yields, so they can keep pace with inflation while sweetening the deal with real-return chocolate sprinkles on top.
The other point worth making is that my clutch of individual linkers were still susceptible to the downward price lurches that afflicted constant-duration bond funds.
The chart above shows a big dip in late 2022 when prices fell as interest rates took a hike, for instance. Think Trussonomics and other traumas of the era.
These are only paper losses to the individual linker investor who holds until maturity or death. Hold fast and eventually your bond’s price will return to meet its face value on redemption day (plus inflation-matching bonus in the case of linkers.)
Meanwhile, the bond fund is flogging off its securities all the time – profiting when prices rise and losing when they fall. That was a very bad design feature during the post-pandemic inflation shock.
My individual linkers’ price dip was smaller than the fund’s largely because I could choose to populate my modelled portfolio with shorter-duration bonds. Short bonds are less affected by interest rate gyrations, as discussed.
Still, I wondered if I was being unfair to the fund. After all, linker funds previously gained in 2020 as money flooded into the asset class.
One last chance for the linker fund
The next chart shows annual returns including 2020, the year before inflation ran hot.
Index-linked bond fund is Royal London Short Duration Global Index Linked M – GBP hedged.2 Data from Royal London, Tradeweb and ONS. February 2025.
Yep, 2020 was a good year for the linker fund. Interest rates fell and its price rose giving it a healthy lead over inflation, and the individual linkers. (Remember the fund profits by selling bonds as prices rise. Meanwhile, the longer average duration of the fund’s holdings meant that it enjoyed a stronger bounce versus my battery of gilts.)
There’s not much to see in 2021 – bar inflation engorging itself – but 2022 is the fund’s annus horribilis. It’s down 5.4% at face value and 16% in real terms. (Horrifyingly, the long-duration UK linker ETF, INXG, was down 45% in real terms that same year.)
Overall, incorporating 2020 does improve the linker fund’s showing. The annualised returns for the five year period 2020 – 2024 are:
- Inflation: 4.6%
- Individual linkers: 3.7%
- Linker fund: 2.2%
It’s still not enough. In my view, the best linker funds available were a fail when inflation actually came calling. I personally held both GISG and the Royal London fund at the time and became deeply disillusioned with them.
All change
The issue driving all this drama was that as inflation accelerated, investors demanded a higher real yield for holding bonds.
The average yield of the simulated linker portfolio above was -4.2% in October 2021. It had risen to 0.5% by December 2024.
When bond yields go up, prices go down. And that exposes the fatal flaw in linker fund design from an inflation-hedger’s perspective – the available products are always selling and even the short duration versions aren’t short enough.
Perhaps yields won’t surge as violently in a future inflationary episode.
But I don’t see why I’d take the risk when I can now buy individual index-linked gilts on positive real yields, hold them to maturity, and neutralise that problem. Individual linkers aren’t going to be slow-punctured by negative yields from here.
So I’ve ditched my index-linked bond funds. They were better against inflation than the equivalent nominal bond funds. But that’s not saying much.
There are other places to store your money so I’ll extend this comparison to the most interesting and accessible of those alternative assets in the next post.
Take it steady,
The Accumulator
Bonus appendix
If you’re interested in buying individual index-linked gilts then these pieces will help:
- Should you build an index-linked gilt ladder
- How to buy index-linked gilts
- Index-linked gilts: how to price, value, and track them in your portfolio
- How to build an index-linked gilt ladder
Are individual linkers better than linker funds?
At hedging inflation yes. At being more profitable, no.
For the avoidance of doubt, I’m not saying that a portfolio of individual index-linked bonds can magick up more return than a bond fund containing precisely the same securities.
What I am saying is that the individual linker portfolio is the superior inflation hedge when each bond is held to maturity. The design of constant maturity bond funds mitigates against matching inflation in the short-term, but should provide a similar overall return in the long run.
If you don’t care about hedging inflation then there’s nothing to gain by swapping your bond funds for a rolling linker ladder.
Fixed duration index-linked gilt funds could also hedge inflation effectively, but they don’t exist.
UK inflation versus globalised inflation
It’s worth mentioning that individual index-linked gilts are linked to UK RPI inflation (switching to CPIH in 2030). RPI was higher than CPI during the period so that’s helped my simulated portfolio claw back some ground against CPI.
By contrast, the short-duration linker ETF, GISG, currently allocates 14% of its portfolio to index-linked gilts. The rest is composed of other developed market, CPI-linked, government bonds: 56% US, 10% France, 7% Italy and so on. The point being that these other linkers don’t protect against UK inflation, though they do match related measures i.e. inflation in highly interconnected, peer economies.
As it was, inflation in these other countries was typically less than the UK’s post-pandemic. I haven’t attempted to calculate what difference this made but I think it’s another reason to favour an index-linked gilt investment product when you can get it.
Individual linker portfolio simulation
I didn’t want to bog the main piece down with a wander through the weeds (well, more than I already have) but for the record I’ll now show my workings.
The individual linker portfolio was constructed from three index-linked gilts, TIDM codes: T22, TR24, and TR26. Each gilt matures in the year indicated by the numbers in the code.
When each gilt matures, the redemption payment is reinvested into the next shortest gilt. For example, T22 is reinvested into TR24. I did not include trading costs for reinvesting dividends or redemption monies.
Relatedly, the performance figures for GISG and the Royal London fund are slightly affected by their OCFs of 0.2% and 0.27% respectively. But I don’t think these charges made a meaningful difference to the comparison over such a short time-period. The differential is too big to be explained by fund fees.
Thanks for the article. Somewhat depressing ! I have a fair amount in 2 short duration index linked funds. I was looking at the nominal returns and wondering as both show a slight loss and that’s in nominal terms.
Revisiting my logic I remembered that I was trying to derisk by moving more into bonds but having lived through the 70’s I have an inbuilt fear of inflation so went for IL funds rather than straight bond funds.
Perhaps time to shift some to individual linkers.
I hold the Royal London Global Short Duration fund in both my ISA and SIPP. It’s there alongside intermediate conventional bonds for the purposes of inflation hedging. I want to keep things simple, it’s the only active holding in both portfolios but the duration of the passive ETFs/funds was too high. What would be a simple set & forget allocation for true inflation protection, that doesn’t require me laddering in and out of index linked gilts?
Thank you very much for an excellent and really useful article. I sold my linker bond funds quite a long time ago after you pointed out their long duration and sensitivity to interest changes. As you say you intend I’ve been buying individual linkers since they moved onto positive real yields after the bond rout a few years ago. You must have saved me an arm and a leg and I am very grateful indeed for the fantastically informative articles you have written on the subject over the years
Do you really mean that you re-invest the maturing T22 into TR24. If it’s a rolling ladder, don’t you want to be re-investing into whatever comes after TR26 (T27, I think)? I need to get my head round this sharp-ish because I sold my Royal London linker fund units last week, with a view to buying that three-year ladder of gilts in Halifax’s next cheap-trading period. Which is tomorrow. (I’m slightly dicombobulated by the mention of “the most interesting and accessible of those alternative assets”, but maybe I’ll worry about that for next year’s ISA.)
Hello. Thanks for the clear explanation. I have a tangential question.
Given money market funds currently yield c5% with low risk and minimal volatility, could you argue they are superior to bonds?
@Chris
If you know when you want to use the money, you could just buy and hold an individual linker to maturity? That’s about as ‘set and forget’ as you can manage. You’d just need to ignore any changes in value in the meantime
Are you sure about the 0.6% annualised return for the linker fund in the first two graphs? It looks closer to 3% to me.
When I was learning about investment I was told that, for instance, the shares of an investment trust are equities and therefore behave broadly like the assets the trust holds i.e. other equities. Whereas, by contrast, a fund of bonds has characteristics substantially different from those of individual bonds.
I decided that any bonds that I could conveniently buy individually – e.g. ILGs – I would. I’d use a fund only for something too intricate or specialised for me to deal individually e.g funds of commercial and foreign bonds. Seem reasonable?
I liked your point “Why did the individual index-linked gilts lose money versus inflation? Because way back in 2021 they were saddled with negative yields. That is, the buy-in price for linkers was so high that their remaining cashflows were guaranteed to sock you with a loss, if you held them until maturity.”
So the price you pay matters and you’ve explained exactly how it matters for ILGs. Thank you.
We had some protection from the last bout of high inflation via ns&i Inflation-Linked Savings Certificates. They have disadvantages – such as no longer being on sale – but you are in no danger of getting negative real returns (where “real” refers to CPI inflation rates). Indeed if inflation turns negative you are guaranteed positive real returns because the nominal returns are collared at 0%.
I believe that they can be inherited; if so tell Granny not to cash hers in.
(Open to correction on the inheritance point: anyone?)
@Richard – great spot, thank you! The annualised return was correct but the trend line was wrong for the linker fund on the first two graphs. Pure cut and paste error on my part. Have replaced the charts, my apologies all!
I’m not sure it’s worth fighting the last war….
Index linked gilts were on deeply negative real yields in 2019 and are now gently positive to around 1.8% at 20+ years maturity.
Buying gilts on a negative real yield is always going to lose you money in the long run….unless they become more negative.
The answer in 2020 was to hold the shortest possible duration bonds.
In 2025 interest rates “appear” to have normalised and the difference between a set of individual gilts or a short duration rolling ladder ( which could be a fund) is unlikely to be dramatic.
I would be more concerned about the strength ( or weakness ) of sterling, which could dramatically affect uk inflation.
Perhaps hold some bonds not hedged to sterling ?
US Tips range from around 1% real yield to around 2.35% real yield, not unattractive.
Linkers protect you from unexpected inflation. If you buy linkers when the real yield on them is negative you are signing up for negative real investment returns no matter what happens to inflation.
You can infer the markets’ expectation for inflation by looking at the difference between the real yield on linkers and the yield on the conventional issuance of similar term (gilts in the UK).
Active antics I know but knowing how much inflation has to increase by before you are better off holding a linker than a conventional bond is useful.
The good news is that the real yield on inflation linked gilts is now positive, so you are no longer signing up to negative real investment returns by investing in inflation linked gilts.
@Hariseldon in addition to ILGs I hold unhedged long dated US TIPs but got in too early so still sitting on a capital loss.
My logic was over the long term the exchange rate should adjust to any relative difference in inflation between US and UK due to purchasing power parity.
So provided PPP does play out it should lead to higher expected returns long term due to the higher real yields on TIPS to begin with.
@Chris – there isn’t a set-and-forget that’s any better than the fund you have. There’s no simple solution. NS&I index-linked saving certs were once the way to go but they just don’t make them any more.
@David C – for a rolling ladder then yes, when one gilt matures, buy another gilt at the far end of the ladder to keep extending it, if you don’t need the cash. I wasn’t strictly simulating a rolling ladder for the purposes of the article.
@Simon – Money market funds beat inflation in 2023 and 2024 but lost every year before that going back to 2009. They’re not an inflation hedge but can win sometimes. More generally, I’d suggest money market is a separate asset class. The benefit as you say is low volatility, the long-run downside is lower returns than longer bonds.
@Hariseldon – the big difference between the individual linker portfolio and fund was duration not yield. If a future bout of inflation raises real yields significantly then the fund will lose again. Ultimately, if your objective is to hedge inflation then why take interest rate risk if you don’t need to?
Foreign bonds seem like an unnecessary complication to me. The bond that hedges UK inflation is the index-linked gilt not the US TIP. I get that you might want to hold TIPs if you want to diversify your bond holdings or speculate on currency.
The only scenario I can think of which would prove beneficial for funds is if a stagnant economy causes interest rates to be held low and inflation is allowed to potter along at a high but tolerable rate.
Like Chris, I crave a simple ‘set and forget’ solution for the risk off allocation of my portfolio. For risk on, any combination of HSBC FTSE All World, Fidelity Index World, SWDA or ACWI is there or thereabouts. By contrast, I fret constantly about the other part, looking at all these things I struggle to comprehend, especially as I move from accumulation towards decumulation. Help!!
Great article. I use my bond allocation to provide some stability to the portfolio and holding individual bonds to maturity means you know your cashflows in advance. At each maturity you can decide how to reinvest (obviously reinvesting the coupons along the way). Bond funds are just not that transparent and I think a lot of investors just hope they will perform in a certain way without understanding what’s really going on.
@prospector I also hold unhedged long duration TIPs , I think the prospects over time are attractive.
@accumulator
I hold bonds to guard against risks and cover my income needs.
By sticking to US and UK Government bonds I am covering credit risk.
By holding a mixture of different durations, linked to my needs for funds I have partially covered interest rates risk. I add to that a barbell of short and long duration bonds
Rather than speculate on currency there is a risk that the pound might significantly deviate in value against the dollar over time. To cover that risk I hold a mixture of pound and unhedged dollar assets.
To guard against unexpected inflation then short duration index linked bonds help but what about the risk of stagflation over a protracted period ? Long linkers could be attractive…
As a general rule my personal take is to split evenly between UK and unhedged US government bonds , split between long and short duration bonds , split evenly between inflation protected and nominal bonds.
In that way every decision about the unknowable future I’m going to be half right and half wrong !
The Market might be even more omniscient than we realize:
Before 2022, if an inflation spike happened, the RPI linked return was indeed what you needed. An inflation spike was coming. The market therefore offered a fixed% loss as the price of linker insurance.
Because it knew positive-real linkers would have really worked well and it knew what could happen. The expensive asset was the right one but it was over priced (like gold or AI now?).
Today we are offered a positive-real return on these same linkers. Why? What does the Market know? Why are they “cheap” now?
Is it that just a few years of low inflation among the: predicted 3.5%- or-whatever RPI supposedly is forever now :- will make linkers look poor?
Is it that governments really have to find a way to let prices rip up without the official stats recording it so linkers are now useless against real-life inflation?
Super good question @Meany #18
Diversification means always having to apologise for something, and in 2022/23 it was the turn of long duration ILGs and INXG to say sorry.
But they’re off the naughty step and sporting positive yields to maturity (“YTM”) now.
A rule of thumb which I’ve seen for fixed income is to have a max portfolio allocation % of the lower of either your age or, either:
a) 10x the % nominal YTM for conventional bonds; or,
b) 10x the sum of the break even rate plus the real terms YTM for linkers.
If the YTM is negative then that means a 0% allocation.
If you’re 60 and the YTM is 5% for conventional or say a 2% real yield plus 3% break even for linkers then the max allocation would be 50%.
Can’t find a way to buy individual US TIPS through the main platforms. Don’t want a HNW a/c with a ‘Wealth’ mgr (partly cos not HNW so too expensive!). Very frustrating: everyone allows US stock trades but not Treasuries!
@Hariseldon (#17):
Interesting approach, which AIUI uses no commercial debt. Have I got that about correct?
An interesting article, but I think it is a slightly flawed comparison between the individual bonds and the fund.
Firstly, and most importantly, the maturity range and duration of the fund (for former it is 1 to 10 years, while the latter is currently just under 5 years) is not the same as the maturity range and duration of the three gilts you’ve chosen (which in 2021 would have had an average maturity of about 3 years and a duration of around the same value). Choosing the lowest maturity gilt in which to reinvest coupons and maturing bonds would also had the effect of keeping the duration lower than the fund (the average maturity is gradually reducing with time). Price changes caused by the change in yields that were as important (for a duration of around 5) as inflation during this period. A fairer test would be to use individual gilts out to a maturity of 10 years.
Secondly, as you’ve pointed out the fund is an international fund the top 20% (at least) of which is invested in US TIPS and therefore dependent on changes in yields in the US (which excepting Truss, were similar in magnitude to the UK) and inflation.
The difference between an index fund and a DIY rolling ladder really comes down to holdings and reinvestment strategies. Typically the fund will hold the bonds in proportion to their issue weights and will reinvest maturing bonds and coupons to maintain this (and will also add new bonds being issued). Holdings and reinvestment strategies for a DIY rolling ladder will depend on the investor, but a fairly logical systematic approach is to reinvest all coupons and maturing bonds at the highest required maturity.
I also note that as a consequence index funds with fixed maturity ranges do not have fixed durations since a) the weighted maturity depends on the bonds in issue and b) the duration of individual bonds, and hence the fund, changes with yield and coupon (duration decreases with increasing yields and coupons). To take a nominal example, the duration of the under 10 year index ranged from 2.5 years in 1975 to over 4 years in 2021, while the nominal ‘all stocks’ index ranged from 4.5 years to 12 years over the same period.
There are several short term ILG indices defined by FTSE-Russell (e.g., under 5 years and under 10 years), but, as you say, funds that follow them are not widely available (e.g., L&G offer them but, AFAIK, only to institutional investors and they exist in the Blackrock Aquila series – not sure about the availability of that one). Like the recent additions of new index funds for nominal gilts (e.g., ishares under 5 and under 10 years), they might become available at some point.
@Alan S — Just to add to your thoughtful comment, there is an ‘up to ten years’ Index Linked Fund available now from iShares that retail can invest in:
https://www.ishares.com/uk/individual/en/products/331736/ishares-up-to-10-years-index-linked-gilt-index-fund-uk
The duration is about 5.5, so pretty low but not super dooper low.
@ Martin White – I use Charles Schwab to buy individual TIPS. Indeed that’s all I use Charles Schwab for. The platform is easy to set up and use (and, at the risk of comparing apples and oranges, compares favourably with the likes of AJ Bell and II platforms).
@al cam
No corporate debt , (it’s priced for perfection )
My split between long and short is not even, I juggle the balance to achieve the desired duration.
@alan s Absolutely right about highlighting rising yields reduce duration, the gilts index went from around 13 years to under 10 years driven almost solely by rising yields.
@Alan S – I take all that on board and it’s absolutely fair comment. My objective is to show what’s possible for a DIY investor holding individual linkers rather than to construct a copy of a short-duration linker fund portfolio that was a compromised choice from the beginning – from the perspective of a DIY investor. The funds are global because those were the only short duration choices available at the time.
@HS (#25):
Thanks for the reply.
I am even more intrigued by your definition of “perfection”. I am not really sure what you mean by that, but would have thought that the introduction of suitably graded, non-callable corporate debt would reduce the cost albeit by introducing some [possibly higher] default risk, but with more diversification.
AFAICT, most annuity providers fish in the corporate debt pond, although there are not enough govt bonds to fulfil all their needs anyway.
Just a thought attached to the observation that you are totally exposed to [two] governments being able to satisfactorily service their debt. So why not try and hedge that risk too? As things stand, I think you could be fully wrong in that scenario – a classic low probability/high impact risk.
Meany #18
I don’t pretend to understand market sentiment, but one of the reasons debt in general was expensive was because of QE – 10 year yields on nominal gilts in mid-2020 were 0.2% (there was still discussion about following Japan into negative yields), such that since implied inflation (the Bank of England term for the difference in yield between nominal and inflation linked gilts of the same maturity, AKA ‘breakeven inflation’, AKA ‘expected inflation’ which is the US treasury term for the same thing) was 3.2%, the real yields on 10-year inflation linked gilts were 0.2-3.2=-3%.
At the end of January 2025, the 10 year yield was 4.5% and implied inflation about 3.6% leaving real yields of about 4.5-3.6=0.9%. The market is expecting slightly more inflation over the next 10 years than it did back in 2020, which given recent experience compared to a decade of relatively benign inflation is understandable.
While we don’t yet know the outcome at maturity, if someone had bought both a 10 year nominal gilt in 2020 and a 10 year inflation linked gilt, so far they would have been happier with the performance of the inflation linked one. Of course, if the spike in inflation hadn’t happened then the outcome would have been different.
wrt your last sentence, the basis of all inflation series (i.e., the basket of goods and services used and the methods of calculation) are openly published, so (under a democratic government) would be difficult to fabricate and any market suspicion of such shenanigans would see a massive sell off. Interestingly (YMMV!), one of the reasons that ILGs were introduced back in 1981 was to reassure bonds markets that the government would not inflate away debt. There’s a paper “The capital market is dead’: the difficult birth of index-linked gilts in the UK” (available for download from the open university site) that is well worth a read for anyone wanting to know more.
@TA (#26)
No worries – IMV the huge advantage that DIY rolling ladders have over index funds is flexibility, tax treatment (outside of pensions and ISAs) and, possibly, lower costs (no management fees, but bid-offer spreads may be larger for the retail investor, e.g. as I write T27 has a spread of about 1%, with prices of 102.44/103.44). One disadvantage is slightly more complexity (particularly when rebalancing).
As for performance, I’m just writing up a study of the difference in returns between nominal DIY ladders and bond index funds in the UK since 1916. Over the entire period the annualised returns for the ladder exceeded those of the fund by somewhere between 0 and 8 bp depending on the sector (i.e., the range of maturities chosen) but the behaviour varies over shorter, 30-year, periods (i.e., sometimes the ladder ‘wins’ and sometimes the fund wins). The behaviour is also different between accumulation and withdrawal.
@ dearieme #8 – yes regarding the ability to inherit ILSC’s…..just request and complete a ‘Transfer Form’ from NS&I. (I happen to have one awaiting completion on my desk gathering much dust whilst i await a response from HMRC!).
@Alan S. Presumably only conventional sovereign bonds as IL haven’t been around long enough ?
@al cam
The yield on corporate debt is only marginally more than government bonds.
US investment grade debt has an average default rate of 2% annually ( although that is unlikely to be a total loss)
IE investment grade bonds don’t offer sufficient reward at present imho.
Regarding the default of UK and US government bonds …..I rather suspect the currencies would take a hit if there was problem rather than outright defaults and there would be other issues that might be concerning….
IE if you held a global portfolio of bonds hedged to sterling and UK government debt was an issue then you would still have problems …..
I believe there is one unhedged aggregate global bond ETF (GLAG) 40% USA and a yield around 1% less than treasuries…
No perfect solution but my approach is I think reasonable, it sits alongside other assets that as a whole seem to cover most events but the unexpected is all around us , all the time !!
@HS (#32):
Thanks for the further details.
That the differential is currently marginal is interesting; I was thinking along the lines of a differential of c. 1%PA in favour of corp. debt. And that the ETF yield is lower than treasuries is just odd to me. What is this telling us I wonder?
I think you once said something along the lines of more could happen than ever does happen – and that makes things tricky. On a similar theme, expect the unexpected covers a multitude of possibilities.
Go well!
JPGR – many thanks, I will check out Schwab. My only concerns would be their long term commitment to running a UK-accessible platform (to hold those long dated parts of the TIP ladder!) and tax: I really want to get those TIPs in a SIPP/ISA home (assuming HMRC don’t extend the ILG CGT rules to overseas sovereigns).
@al cam
The yield on a treasury ETF may be lower than the treasuries within that ETF but that reflects the pricing of individual treasuries that may be below par and the historic nature of the income flow.
Eg CBU7 ishares 3-7 has a yield to maturity of 4.41% duration 4.25 years with a yield of 3.94% on a trailing 12 month basis.
The yield on the underlying holdings of 3 to 7 year treasuries, range from 4.298% to 4.464% so the ETF matches the expected.
@martin white There are no TIPS maturing between 2035 and 2040 which hinders forming a ladder.
There is a 0-5 year TIPs etf from ishares and UBS offers a 1-10 year ETF and a 10+ ETF
all can be held in an ISA and are available hedged and unhedged.
@HS (#35):
Good point, thanks again.
If you look at the real yields on an individual gilt such as TR8F (maturity 22/11/2055) then you can see why it hasn’t kept pace with inflation from 1st October 2021 to 31st December 2024.
https://ibb.co/jXckkf0
At 1st October 2021 it’s clean price is 175.41, it’s dirty price is 431.71, and it’s real yield is minus 2.13%pa
And at 31st December 2024 it’s clean price is 86.15, it’s dirty price is 270.98, and it’s real yield is plus 1.84%pa.
That’s an almost 4% increase in real yield, a 37% reduction in the dirty price and a 51% reduction in clean price. So in real terms (against RPI) it has dropped 51% in value, and in nominal terms it has dropped 37% in value, ignoring the coupons paid from 2021 to 2024 which make only a small compensatory dent to that fall.
It’s price at 31st December 2024 is by a back of an envelope calculation about 28% of what it would have been had the real yield stayed the same at minus 2.13%. That’s based on 0.96^31 = 0.28 where 31 is the outstanding term to maturity. So it’s price fall is mainly down to the increase in real yield that it’s priced at. And of course it would have been purchased at a negative real yield of minus 2.13% at October 2021 so would have been guaranteed to provide a real loss if held to maturity.
If we now look at the spot yield curve (which is the hypothetical real yield that a zero coupon gilt of that term would be priced at) we see also at every term the real yield has increased from 2021 to end of 2024.
https://ibb.co/Pv4ZzFYb
For example at 30th September 2021 the spot real yield on a 5 year index linked gilt was minus 3.31%pa, and at 31st December 2024 this had increased to plus 0.56%pa. The change in real yield will have less of an effect on the price than TR8F because it’s 5 years and not 31 years, but it’s still in the wrong direction. So the real return from October 2021 to 2024 on say a bond fund with average duration 5 years is we’d expect going to be considerably worse than minus 3.31%pa.
Of course the main point as made in the article is that index linked gilts are now priced to provide a real return if held to maturity, whereas in 2021 they were priced to ensure a real loss. And that changes everything.
@Mr Optimistic (#31)
Yes you’re right, only conventional gilts (although ILG have now existed for 40 years, so a first stab could be made).
@Hariseldon (#35)
Re: The TIPS maturity gap. There’s an interesting (and very long!) thread at https://www.bogleheads.org/forum/viewtopic.php?t=432366 on using duration matching to cover the gap years in a TIPS ladder. The essence being that holding excess TIPS either side of the gap can provide reasonable protection against reinvestment risk (e.g., largish changes in yield).
@Snowman #37: it’s a terrific example (TR8F) and it points to an obvious issue that was somewhat under discussed IIRC from 2009 to 2021, the QE era if you will. Namely, whilst the first cover which I can remember to call a bubble in equities after the GFC was way back in November 2009 (!), and whilst for the next 12 years we were all incessantly warned of an equities’ bubble about to pop; far far fewer commentators stated the (blindly) obvious that if equities were indeed overvalued as claimed then that necessarily (by parity of reasoning) made long duration gilts, especially ILGs, and long duration Treasuries, especially TIPS, a screaming sell in terms of bubble fears.
Everything is relative in investing.
If long bonds have negative yields and cash near zero interest paid then they’re less unattractive than equities even if P/Es have expanded and the Earnings Yield has contracted.
Put another way, there’s a potentially steep price to be paid in choosing not to be able to have any flexibility to respond to large and long lasting changes in the relative attractiveness of assets by adopting a pure passive fixed allocation to equities and long bonds, including ILGs.
ILGs, TIPS and commodities (and even gold) can provide hedges against inflation, deflation and equities in the right circumstances.
However, they’re each less effective in this the more relatively highly valued they are compared to both their own price histories and as compared to the alternatives to them as assets.
QE operated through the bond market, not the equity market. It was always much more likely to really substantially affect the return profile of long duration bonds, especially long ILGs, than it was for other less ‘inflated’ assets, like equities (even though it probably had significant indirect effects on those too).
Today cash has yield. Bonds have yield. ILGs and TIPS have real yields. There are alternatives to equities that give you decent returns. The investment world has been utterly transformed since 2021, and (in terms of both choice and diversification) arguably much for the better. Sanity is restored!
Apologies. Too many double negatives there. Para three should say “less attractive” and not “less unattractive”.
“far far fewer commentators stated the (blindly) obvious that if equities were indeed overvalued as claimed then that necessarily (by parity of reasoning) made long duration gilts, especially ILGs, and long duration Treasuries, especially TIPS, a screaming sell in terms of bubble fears.”
Ahem, because he’ll be too modest to link to this himself (possibly), here’s TI-adamus on pricey bonds back in 2008:
https://monevator.com/government-bonds-an-exciting-new-way-to-lose-money-to-the-bear/
@Alan S – please post your paper around these parts when you get a chance!
@Snowman – really interesting as always.
@Snowman (#37)
To me your excellent post confirms that the price paid is important! I also note that the 1.25% coupon on TR8F could have been reinvested at a variety of yields since 2005 (and since 2019 coupon reinvestment would have been at much cheaper prices – one reason why future total return on a non-zero coupon bond is unknown) .
I note that, in retirement, an advantage of a collapsing ladder of ILGs is the relative certainty in income it provides compared to that from a portfolio. To take a simple example, the historical SWR for the UK was around 3% to 3.5% for a 30-year retirement, while a 30 year gilt ladder with a real yield of 0% will provide an SWR of 3.3% (even a real yield of -0.7% will give an SWR of 3%). Even at lower yields, a shorter ladder can help obviate a poor sequence of returns in the portfolio early on in retirement.
@TA (#41). Will do.
@Delta Hedge #39
Great post.
I agree quantitative easing (QE) is something that isn’t discussed enough. The deficits of the Asset Purchase Facility Fund (through loan interest exceeding bond interest and losses on QT sales) of billions of pounds a year are having to be covered by increased taxation (or borrowing). What proportion of the public are aware of this?
Although as I understand it index linked gilts didn’t form part of the bonds purchased though QE despite making up about a quarter of all gilts, clearly they are going to be amongst the assets most affected by the government buying up conventional gilts through the Asset Purchase Facility in QE.
The huge falls in real yields leading to negative real yields on index linked gilts happened after QE was announced/started. And real yields have now moved back into positive territory now QE has stopped and quantitative tightening began in 2022.
How much of the dramatic falls then increase in real yields priced into index linked gilts at the longer end can be assigned to QE and subsequently QT in your view?
@snowman #43, I think I saw a research piece that QE in the UK resulted in the BoE holding a third of the gilts in issue.
Prices are inversely proportional to yields and prices reflect the marginal trade. I’d say buying up a third of the market would have pushed the price of the marginal trade up substantially, perhaps 10-20%? And so yields down 0.5% to 1%?? So by no means explains all the fall in yields seen post 2008.
Much of the demand causing the remaining fall in yields I suspect came from insurance companies and pension funds due to regulatory and accounting requirements.
As market expectations of low base rates were priced in for longer and longer pension funds and insurance companies would be buying gilts to match their long dated pension and annuity liabilities. Thus bidding up long-dated gilt prices even further.
This all unwound spectacularly (at least for pension funds who bought the gilts on leverage) in Oct 2022. During the Truss premiership margin calls resulted in pension schemes having to sell gilts resulting to rising yields/falling prices, further margin calls and further gilt sales. Rinse wash repeat until the Bank of England had to step in. Meanwhile gilt yields had risen 1-2% or so which is probably a fair reflection of how much pension schemes bid up prices in the first place buying the gilts.
@Snowman #43. Thank you. I can’t improve on the answer of @Prospector #44 a moment ago.
I’ll just note here that the BoE brought £875 bn of conventional gilts (and £20 bn of CBs) in total:
https://www.bankofengland.co.uk/monetary-policy/quantitative-easing
And that, as of May 2024, it still held 80% (£700 bn). That’s a lot of unwinding and QT left still to do against a backdrop where UK CPI is at 3% and most recently rising.
@Delta Hedge #45
Unwinding of QE is something I tried to get my head around a few years ago. There is a strange world in which the BoE just holds onto the gilts to expiry.
US focused piece explaining the ins and outs here:
https://medium.com/new-york-fed/what-might-the-feds-balance-sheet-look-like-in-the-future-4bfa8292cd07
Thanks TA for sharing this analysis.
I thought I’d also share some analysis I did a couple of months ago now on the performance of various bonds/bond funds over 2021-24.
The quick summary is that UK linkers of appropriate duration seem to be the only way to have confidence in what real return you will get out of those I’ve considered, currency hedging doesn’t seem to work as intuitively expected (if someone could explain that for me, it would be very interesting), and very short-term nominal US bonds also did intriguingly well relatively.
UK inflation in the period was 24% (CPI) and 33% (RPI).
First, a fairly short-term UK linker, similar to the article’s analysis – I used the linker maturing in 2026. (This is one I might have held in 2021 if I were decumulating, and it’s the one with the nearest maturity date that I could easily fund data for.) This returned 13.9% by my calculations. So 19% below the RPI index that it is meant to track, partly explained by the increase in interest rates, with the bond having started at a price premium of 17% above its maturity value, which fell close to 0%, accounting for 15% of the underperformance. I’m not sure where the other ~4% went, but it seems close enough to consider the performance in real terms quite predictable in advance.
I also looked at the hedged international inflation-linked government bond funds listed on Monevator’s low-cost funds page – as found in the article, these did a lot worse e.g. “abrdn Short Dated Global Inflation-Linked Bond Tracker Fund B Acc” returned only 6.7%, 17% below UK CPI and 7% below the UK 2026 linker. Causes may be longer duration, and also perhaps that it tracks closer to CPI than RPI inflation, which would give ~9% underperformance compared to RPI-linked investments based on the UK inflation figures.
A US TIPS (0-5yr) $ ETF (STIP) did a bit better (+10.2%). The US CPI increase (21%) was ~3% lower than the UK’s. The performance relative to the UK 2026 linker (about -4%) is about what I’d expect given the difference in CPI inflation (-3%), the CPI-RPI inflation difference (about -9%) and the $/£ exchange rate improving (about +9%). Real interest rates increased similarly in both places, so had a similar effect on returns I think. I don’t see a good reason to use these rather than the internationally diversified funds, but it’s good to see the numbers work.
However, the hedged TIPS (0-5yr) version (T15G) performed very similarly (+9.9%) – I’d expect the relative performance to have been lower by the weakening of the £/$ exchange rate. This raises the question for me of what affect does currency hedging actually have? (Did hedging produce a positive return in this period that cancelled out the exchange rate change? It’s a bit concerning if it doesn’t dampen exchange rate volatility. It also raises questions about the hedged global funds.) It also makes it seem difficult to anticipate what returns in future would be, if the dependence on the exchange rate change isn’t removed.
I also looked at short-term nominal bonds, as suggested by the fellow at Portfolio Charts (https://portfoliocharts.com/2017/05/12/understanding-cash-will-make-you-a-better-and-happier-investor/ and https://portfoliocharts.com/2022/05/27/proven-ways-to-protect-your-portfolio-from-inflation/). A UK 0-5yr Gilt ETF (IGLS) returned 0.4% and a US Short-term treasury ETF (SHV; with bonds with <1yr maturity) returned 11.2%, slightly better than 0-5yr TIPS. However, it did ~6% worse in 2021 and then had to catch up. So this gives some partial support to PC's idea for using nominal bonds, for those with <1yr maturity at least, but 0-5yr seems too long for nominal bonds to be used this way. I don't know of any funds allowing investing in UK gilts with <1yr maturity or thereabouts, though.
Hopefully that's of some interest to someone. It would be great to see an explanation of what the effect of currency hedging in funds actually is.
Thanks @Prospector and @Delta Hedge.
By the way on quantitative tightening in the UK, there’s an absolutely brilliant highly technical FT article by Tim Young from January 30th 2023 (but none the less relevant for being 2 years old) titled “How quantitative tightening really works, the world according to BEAPFF”.
I can’t direct link to it but if you do an internet search based on the title on a desktop computer you should be able to access it and the article comments.
@Peter R – thank you for your thoughtful comment. Was there some difference in duration between those two US ETFs that might explain the puzzle?
For GBP bonds under 1yr, I’d think money market funds. They did outperform linker funds over the period, though I don’t think they’re a great anti-inflation option. Just better than longer bonds.
Also, you could invest in individual nominal gilts to create a DIY ultra-short gilt fund. I’ve been surprised to learn recently how many Monevator readers are closet gilt traders on the side 🙂
@Peter R (#47):
Very interesting post – thanks for sharing.
Slowly but surely I feel that collectively we are all edging towards a clearer understanding – and, as seems to be the norm – no free lunches/miracle products/strategies exist; no matter what one might think “it says on the tin”!
@TA If by “those two US ETFs” you mean the hedged and unhedged 0-5yr TIPS ETFs, both track the same index of 0-5yr TIPS and both have the same duration in Morningstar data (2.54 years).
Good point about money market funds – yes they look to have performed similarly to US <1yr treasuries. Though I don't feel sure about how their risk compares to government bond funds.
Maybe a DIY short gilt fund it possible, but it sounds like quite a bit of work!
@Al – thanks! Yes maybe we are edging closer to understanding bond-like investments, but I feel I still have a long way to go!
@Peter R – Sorry yes, I was talking about those two. That is a mystery.
I do think there’s more risk embedded in money markets than with the equivalent gilts:
https://monevator.com/money-market-funds/
I hear you on the work involved running a short individual gilts fund. That said, I think most of the effort is upfront – schooling yourself on the arcanery of it all. Once that’s done I don’t think it would be harder than maintaining a small portfolio of shares – provided you choose a broker that trades gilts online rather than over the telephone.
@TA Cheers. On short gilts, if the aim is to keep pace with inflation, I think our analyses indicate that using individual linkers with up to ~a few years until maturity would keep up with inflation fairly closely, if they don’t have very negative yields built in at the time of purchase (and even with negative starting yields on this occasion, they still did better than a <1yr bond fund). So it doesn't seem to me that there's a need for maintaining a set of <1yr gilts actually – it was just interesting I thought that a fund of very short duration bonds outperformed a fund of quite short duration inflation-linked bonds.
Totally agree. I was a bit non-plussed when I discovered I’d have been better off with a money market fund than a short duration linker fund in an era of high inflation!
As it is, I’ve turned to individual index-linked gilts to solve the problem.
Any thoughts on transferring away from bond funds (held inside a pension and ISA) into gilts whilst maintaining my asset allocation ratio?
If I switch my bond funds to shares my allocation will be off.
Should I hold individual gilts inside the pension and ISA? Most advice seems to say it’s pointless as it negates the no CGT benefit. But in this scenario it seems to make sense.
Anything else I should be considering?
Whisper it but will the CGT exemption survive? Devastating for individual linkers outside isa/sipps.
@Crazy Caps – There’s only an issue holding gilts inside tax shelters if they’re displacing something else that would be more efficiently held inside your ISA/SIPP. Otherwise, you don’t need the CGT benefit.
Also worth noting that the interest on gilts is taxable.
Great name, btw!