Orthodox investing advice has always been that UK investors should hold either gilts or high-quality global government bonds hedged to UK pounds (GBP) for the main part of their defensive asset allocation.
I’ve long subscribed to that advice myself.
Highly-rated government bonds provide the defensive ballast for your investing strategy. Sticking with your local currency – in our case GBP – avoids adding additional riskiness into what’s meant to be the steadier portion of our portfolio.
But (heresy alert!) recent evidence suggests that unhedged US Treasuries could be a better choice.
Why? Because the dollar has risen against the pound in the majority of stock market slumps since the turn of the century. When it did so, it bestowed a welcome FX bonus for UK-based investors who owned US government bonds – juicing up their defensive returns at just the right time.
The upshot is holding US bonds can protect your portfolio from equity losses better than home-grown gilts – when it works.
But what if that trend is a reversible historical anomaly, and not a bankable portfolio hack?
To approach that question we need to ask some others.
For starters, did US Treasuries dominate gilts for more than just the past couple of decades?
A second salient question: how does this strategy impact long-term returns? Because if US bonds actually deliver lower returns than gilts over time then they become much less attractive, even if they do better for a bit in a crunch.
Maybe those currency gains quickly unwind once market jitters subside, exposing UK passive investors to FX blowback and potentially extra-nasty losses if they’re caught sitting in US Treasuries?
It’s quite the conundrum. But if I’m better off holding US Treasuries instead of gilts though then I’d really like to know about it. So let’s dive in.
UK vs USA: enter at your own risk
Before getting to the good bit, we need to repeat that holding unhedged US Treasuries ahead of UK gilts means adding currency risk to your fixed income asset allocation.
Currency risk can work for you or against you:
- A UK investor – with assets valued in US dollars (USD) – benefits when the pound weakens against the dollar. A rising dollar means USD-priced assets are now more valuable in GBP terms.
- Equally, USD assets sink in value in GBP terms when sterling strengthens. The dollar price of such US assets is now worth less in pounds.
These gains or losses from currency risk are grafted on top of the asset’s underlying return.
If you invest in unhedged US Treasuries, you’re hoping for two things to happen:
- Your government bonds spike as global equities tumble.
- You additionally profit from the surge in demand for dollar-priced assets.
Hence this ploy adds an extra risk to your collection. Namely, that the dollar doesn’t live up to its reputation as a safe-haven during a market tailspin.
If the USD falls against the pound in an “adopt the brace position!” scenario then the currency knock-back could swamp any bond bounce you hoped to gain.
All of which tells us that playing FX roulette with your defensive allocation is like releasing a predator into the environment to wipe out a pest species.
It’s inherently risky and it may not work as advertised.
UK vs USA: battle of the government bonds
To discover how frequently US Treasuries beat gilts during sustained stock market falls, I calculated the annual total returns of unhedged US Treasuries in GBP from 1971 to 2022.
We’re looking at GBP returns throughout because we’re interested in this substitution from the perspective of a UK investor.
And those dates were selected because they span the entire floating exchange rate era for currencies, up until this year.
Next I compared the GBP returns of Treasury Bonds against gilts in every year when UK equities registered a negative annual return and/or the UK stock market fell 10% or more, for a period of at least one month, regardless of whether that loss is revealed by the annual returns data.
Against that backdrop, US Treasuries beat gilts in 15 years out of a sample of 21:
Nominal total return data from JST Macrohistory1, FTSE Russell, and Aswath Damodaran. Annual exchange rate from Measuring Worth2. February 2023.
Gosh, that’s quite the thumping. Not as bad as our record in the America’s Cup, but still a comprehensive win for US Treasuries.
Is that it in then? Is it time to ditch our gilts? Do we never need to worry about a mad Prime Minister ever again?
Not so fast…
US Treasuries vs gilts: overall annualised returns
Next I looked at the match-up between gilts and US Treasuries over the entire period, in terms of their annualised returns.
And oh my, the plucky Brits have won something!
George Washington, John Bogle, Beyonce, are you watching? Your bonds took a helluva beating! 
Okay, sorry about that – I may have got carried away and slightly exaggerated.
The bond scores are:
- 8.3% gilts
- 8.0% US Treasury bonds
Those are nominal, average annualised returns across the entire 52-year period, for an investor operating in UK pounds.
And there’s essentially nothing in it. Regardless of whether you bought and held gilts or Treasuries, your overall returns were much the same after 52 years.
US government bonds actually bested gilts, by 28 years to 24. But much of the gain made in US Treasuries during down periods was later undone by the strengthening pound when market confidence was restored.
US Treasury Bonds vs gilts: across the decades
Next question: are there distinct eras when owning US Treasuries worked best for UK investors?
The table below shows how many years per decade that US government bond returns exceeded gilts when UK equities fell (same criteria as before):
| Decade | US Treasuries | UK gilts |
| 1970s | 4 | 1 |
| 1980s | 1 | 1 |
| 1990s | 1 | 3 |
| 2000s | 4 | 0 |
| 2010s | 3 | 1 |
| 2020s | 2 | 0 |
By this reckoning, the 1990s was the only decade when US Treasuries didn’t counterbalance sliding stock prices at least as well as gilts.
However even this data hides decent periods for our boys versus US Treasuries.
Most notably, gilts made a comeback versus US Treasuries in the late 1970s, held their own in the 1980s, and then actually outperformed in the 1990s during those down years.
So preferring US bonds didn’t benefit UK investors for about a quarter of a century.
How bad are US Treasuries when they don’t perform?
When equities caved but gilts outperformed Treasuries, the average nominal annual return for each government bond for UK holders was:
- US Treasuries: -2.3%
- Gilts: 12.3%
Which is a painful showing for the US asset – one that would probably leave you ruing the decision to go off-piste if it happened to your portfolio.
As mentioned at the start, the problem with adding a currency play to the bond side of your portfolio is that FX volatility can swamp the asset’s typically more modest underlying returns.
Hence my biggest fear with this strategy is that an adverse currency move could cause US bonds to inflict large negative returns upon investors who are already buckling under the strain of watching their equities nosedive.
The worst GBP annual return for Treasury bonds was -13.2% during 1987 – the same year as the Black Monday Crash. In contrast gilts were up 17.9% that year.
That said, when gilts fell -16% in 1974 and -24% in 2022, US Treasuries were up 7% and down only -9%, respectively.
America the Beautiful
How do things look when US Treasuries beat gilts during stock market losing streaks?
Well, under these conditions, average nominal annual returns for the two government bond types were:
- US Treasuries: 12.3%
- Gilts: 2.3%
Meanwhile, across all 21 of the down years we looked at earlier, the average annual returns gap narrows to:
- US Treasuries: 8.1%
- Gilts: 5.2%
It’s still advantage US Treasuries, but the picture is more mixed.
Which leads me to wonder: which bond is the better option during a proper nightmare?
Which bond works best during the worst bear markets
The stiffest tests of investor nerve this past half century were the 1972-74 stock market crash, the Dotcom Bust, and the Global Financial Crisis.
US Treasuries beat gilts 3-0 during these utter meltdowns.
Here are the average returns:
- US Treasuries: 13.2%
- Gilts: 1.5%
That’s a big performance gap. US bonds potentially bucked up your portfolio just when you needed it most.
However, there’s one final and important check we need to make.
What difference does replacing UK government bonds with US Treasuries make to the overall returns for a globally diversified portfolio?
US Treasuries vs gilts: diversified portfolio returns
I compared the long-term results of two diversified portfolios. Both feature 60% MSCI World equities, with the remaining 40% devoted to either gilts or US Treasury Bonds.
And it’s a photo-finish!
Here are the nominal, annualised returns for the two portfolios (1971-2022):
- World / US Treasuries: 9.88% annualised
- World / Gilts: 9.94% annualised
(Equity returns are in GBP from the MSCI World index. Portfolios rebalanced annually.)
However, it wasn’t so close over the entire time frame. The portfolio with gilts was actually an annual percentage point ahead by the end of the 1990s.
It was still almost half a percentage point ahead before the Brexit Referendum.
US Treasuries vs gilts: bet now!
We’ve seen then that US Treasuries can indeed cushion your portfolio better than gilts when equity confidence crumbles. Not all of the time, but the majority of the time, at least historically. And especially in the worst crunches.
However just to keep things interesting, gilts edged the win when it comes to overall portfolio returns.
To me this strongly suggests the strategy is only worth considering if you’re a particular type of investor – one who is hands-on with your portfolio, enjoys managing extra complexity, and understands the extra currency risk may not pay off (and might even backfire) at the worst possible time.
So if you want to keep things simple, then you can happily leave this ploy alone.
Despite 50 years of relative UK economic decline, you’d still have been better off owning gilts all told.
Cool Britannia revisited
Don’t fall for gloomy geopolitical narratives that the UK is destined for the international knacker’s yard.
Plausible-sounding storylines of doom are the stock-in-trade of financial punditry. But they’re no basis for a long-term investing strategy.
To give you but one counterpoint: nobody would have predicted gilts would buck the trend against US Treasuries after Britain lurched from crisis to crisis during the 1970s as ‘the sick man of Europe’.
Sure, we’re in a mess right now. But our current national conversation could as easily signal a turning point as herald further decline.
Finally, if you are tempted by the idea of adding unhedged US Treasuries then consider dipping a toe in the water, rather than entirely ditching gilts (or GBP hedged global bonds).
You could split your nominal bond allocation fifty-fifty, for example. Or you could instead buy a slug of gold, given it’s a non-correlated defensive asset that also boosts UK investors when the dollar rises.
Take it steady,
The Accumulator
A few other points:
– Taxation (0% CGT for gilts) strongly tilts the scale in favour of gilts.
– FX hedging costs for USTs negate the yield difference, so the debate should be unhedged USTs vs Gilts, which really comes down to whether you think GBP will continue depreciating against USD. But if you do, you are proposing an FX trade, which is not what the bond allocation is meant for, regardless of past performance.
– To partially protect from GBP depreciation, index-linked gilts help, as FX depreciation feeds inflation. Even more so considering the tax issue mentioned above, as index-linked gilts can have almost their entire returns tax free as coupons (the only taxable element) are small.
Good to see a detailed piece on government bonds; the platforms have seemed strangely quiet on this topic, given the opportunity to buy into this stuff at what have looked like decent yields after years of drought.
I have been building up positions in these, although it has been sometimes tricky with yields cavorting all over the place. I was fortunate to get a longer-term chunk of gilts when Truss was briefly causing chaos as PM. Overall, gilts has gone into a mix of the Legal and General All Stocks index OEIC and the Vanguard VGVA ETF. I have gone for a 50/50 mix of gilts and GBP-hedged US Treasuries. While the FX hedging mechanism reduces any yield advantages, I have not gone onto unhedged US Treasuries, the point being that gilts and UST’s remain quite distinct risks. For hedged US Treasuries, I have used the Vanguard US Government Bond OEIC (GBP-hedged class) and the Ishares TRGB ETF.
A timely article! For simplicity’s sake, I am moving forward with a simple allocation of:
80% Vanguard FTSE Global All Cap fund
20% iShares Global Government Bond ETF GBP Hedged (IGLH)
I am not in a place where I can spend time trying to capitalize on yields of short term government bonds and miss out on gains, or take the duration risk on long term government bonds in the hope we enter a recession and there are interest rates cuts. I also don’t know how I can decide between US vs UK Government bonds and if I mix, what allocation to give to each, and what durations to blend.
So much like with my equities, I am outsourcing this. I will use the IGLH ETF, which is an intermediate bond ETF that has exposure to 7 developed countries government bonds.
Hopefully this makes sense and pans out for me!
It seemed sensible to me that if I went global for equities that I did the same for bonds
I had done short term gilt ladders in my time but it was a lot of work
So killed 2 birds with one stone ie went global and used one Vanguard fund only for bonds(and equities)
Been in this position for many years and it seems to have worked
The parallel fall of bonds and equities last year was a new one for me but it has happened historically -knowing your investment history is important!
Bonds did not fall as far as equities so arguably did their job albeit more poorly than one would have liked
Bonds now seem to be reverting to their usual pattern of moving out of sync with equities
Everything changes but everything stays the same!
Staying the course seems to still be the best policy for amateur investors provided your Asset Allocation has initially been set satisfactorily
xxd09
Thank you, TA, for the analysis. I was wondering myself if I should use Treasuries for a better rebalance bonus. However, I was wondering if you have analysed drawdown of 60/40 portfolio for each of them (as that is what I want to from bonds to smooth the line)?
I always considered it sensible to hold non-uk denominated items because what we buy is not solely priced in pounds.
Sure when you go to the petrol station, the price you pay is in pounds, but the price the companies are paying for it is in Dollars. So if there are big FX fluctuations, the companies will adjust their prices as required.
This goes for much of what we consume. For this reason, I would want a significant portion of my ‘safe’ assets unhedged.
So in share-heavy portfolios there might already be plenty of unhedging (i.e. dollars) in the shares. But in a defensive portfolio with a lot more bonds and so on, is this the situation where increasing the amount of dollar protection by
using more unhedged treasuries is likely to be more helpful?
Thanks very much for the article. I am planning on going for the 50/50 split in my bond allocation. I don’t think it will do much harm, a little more diversification is how I see it. Certainly not expecting it to blow up the portfolio.
I’ve been sitting on the sidelines though, watching and waiting for the exchange rate to improve since the Truss crash.
The OECD Purchasing Power Parities report (if I’m reading it correctly) estimates parity at ~1.5 dollars to the pound. Trouble is, it looks like GBP has been at a ‘discount’ for some time, and perhaps it will stay that way too. I guess that highlights part of the problem, when and how to judge an entry point.
https://data.oecd.org/conversion/purchasing-power-parities-ppp.htm
I have noticed in the past when comparing hedged and unhedged US bond etfs that the performance numbers seemed to be much better for the unhedged version.
However, the period of comparison covers just the post 2008 period and I wondered if perhaps this was ‘non-representative’, low interest rates QE and so on. However you have to wonder on what basis you should try to paint such a long and recent interval as somehow a distortion of some other more true reality.
If the bond holding is meant to be defensive, a tanking UK economy relative to the US would lead to reduced bank rates, lower pound and lower gilt yields so higher gilt prices. At the same time the lower exchange rate would boost an unhedged US bond holding. I assume a hedge would be ineffective or just a cost drag. A surging UK economy would lead to higher bank rate reducing gilt prices and raising the exchange rate so reducing the value of an unhedged US bond fund. Sounds like a currency hedge would help.
Is my thinking correct? If so hedging looks valuable if you believe the UK is going to out perform the US but not otherwise.
I wonder what gilts look like in a taxable account vs in a sipp or isa. Is tax on the income offset by no CGT?
Lovely article TA, I’m still deciding which three names I’d have chosen for a ‘helluva beating’ 🙂 although your three were great choices.
I think you make a great argument for diversification and inclusion of treasuries, gilts and gold in a portfolio, certainly a retirement portfolio. Perhaps you should include all three in your planned educational portfolios. Gold certainly took the sting out of the recent gilt massacre for me.
On the topic of new portfolios I’d certainly like to see accumulation and decumulation portfolios and their accompanying treatment of volatility. Also, (while I’m on) perhaps splitting Vanguard Small Cap into Mid and Small Cap would be useful too.
Is cash (not under the mattress) in simple savings accessible accounts a viable alternative to bonds?
(may have discussed previously).
Many thanks TA, another enlightening article. Ref new portfolios, It would be good to see something for the “oldies” following the combined bond and equity route.
I have been considering this for a while and the extent to do this in my own portfolio. I’ve was working through the following:
1) Have you considered (or factored in) your portfolio behaviour due to the different average maturities and durations of the US Treasury and US Gilts indices. I’ve been modelling off the Bloomberg US Treasuries Gross Return in USD (LUATTRUU) and the Bloomberg Sterling Gilts Total Return Index (LSG1TRGU). Aggregate Gilt indices have a materially longer average maturities and durations, with 13.7 years average maturity and 9.8 years duration vs 7.9 years average maturity and 6.3 years duration (using the above Bloomberg indices). Hence, considering the Treasury and Gilt portfolio characteristics may be important in the current cycle.
2) Have you looked at impact of the cyclicality of USD over different horizons? Over long horizons, like 50 years, the currency impact is limited when compared to the significant impacts over shorter horizons of say 10-15 years? The dollar index (DXY) had three major cycles over the last 50 years. Over even shorter periods the challenge is greater. The USD spot rally to Q3 2022 against all currencies was close to a 2-sigma event (a once in every 44 years event!) vs average USD annual returns over the last 50 years.
3) To align with the post, the key may be the appetite/extent to hold USD as a diversifier within the total portfolio. Over the 50 year long term, whilst the UK had a positive interest rate differential to the US (1.3%), the Dollar has a significantly negative correlation (-0.4) to global equities. Whilst Sterling is positively correlated (+0.3) to global equities.
Personally, I’m still working through it and so far I’ve aligned my portfolio of Gilts to be similar to the UST index maturity/duration by using a mix of the All Gilts index and a shorter maturity/duration index. I expect to move a portion to an unhedged treasury index. My other holdings of an unhedged Global equity index, hedged gold and a hedged commodites index provide me with other diversifiers for of a stressed scenario – e.g. as a sterling based investor with an intrinsicly high dollar denominated equity holding.
Thank you for the interesting post. A new subscriber 🙂
@ mp – Good point about checking any available tax advantages. I’d add that gilts tax edge only applies to individual gilts (not gilt funds) if you hold your bonds outside of tax shelters. I’d guess that most Monevator readers hold gilt funds and have a significant portion of those tucked away within ISAs / SIPPs.
Love your point about linkers.
@ CM – I think you describe the dilemma perfectly. On balance, this doesn’t look worth losing sleep over to me.
@ xxdo9 – you’re spot on – there was nothing unusual about equities and bonds falling in tandem last year, except that it hadn’t happened for a long time. What was unusual (though not unprecedented) was the scale of bond losses. Nonetheless, it’s not surprising 2022 was shocking to many given bonds are marketed as a relatively low risk asset class. Sadly low risk doesn’t mean no-risk and comes with a ton of small print.
@ Captain – I haven’t analysed US treasuries as part of a drawdown portfolio. But I agree it would be interesting to see, especially as they significantly outperform during the three worst bear markets of the past 50 years. Thank you for the idea! I’ve put it on my ‘to do’ list.
@ Meany – You could argue that in a 60% bond portfolio or similar then adding a slug of US treasuries increases diversification. OTH, if you’ve chosen that portfolio because you prize low volatility then you could shoot yourself in the foot when US treasury returns are swamped by adverse exchange rate moves.
@ Mr Optimistic – yes, there tends to be lots of talk about hedging dollar holdings when the pound is apparently weak and dropping the hedge when the pound seems strong – on the assumption that mean reversion will follow. It’s tactical asset allocation – which is fun – but I suspect is about as successful as trying to pick stock market winners and losers.
@ Nearlyrich- that is a fun game! I think ‘helluva beating’ has to be my favourite piece of football commentary ever.
I am tempted to test US treasuries in a retirement portfolio but wary of overcomplication. Maybe I’ll start with the SWR analysis suggested by the Captain and see what difference it makes using historical data.
@ Oldie – you surely can replace bonds with cash and a number of veteran Monevator commenters have done exactly that. It probably means giving up on return and recession protection. My own take is I want to be diversified across both asset classes.
@ Howdy – Thank you for taking the time to post and for your considered response. Yes I did consider the maturities angle – and happily the available long-run data sets are approximately comparable.
Average maturities are as follows:
1971-1989
UK and US – 20yr
1990-2015
UK 15yr
US 20yr
2016-2020
UK 10 yr
US 10yr
2020-22
UK 13yr
US 10yr
Obvs you could repeat the analysis using different maturities in the hope of discovering a sweet spot, but that would beg the question, ‘Would it replicate ex-ante?’
As for the impact of shorter-run exchange rate patterns, I think the onus would be on proponents to demonstrate that these are signals not noise. i.e. are they predictable ex-ante and actionable by everyday investors?
I completely agree with you that reallocating to US treasuries is dependent on each investor’s willingness to diversify. I am personally attracted to the idea but so far have done nothing about it – I think my inertia is trying to tell me something 🙂
I recently discovered this report titled “Currency Exposure in Fixed Income: To Hedge or Not To Hedge?” – not specifically US vs UK, but interesting all the same.
https://www.pinebridge.com/_assets/pdfs/insights/2019/10/currency-exposure-in-fixed-income.pdf
Gilt yields obviously move when base rates change but how close is the correlation?
I’m looking at building a gilt ladder (probably 12 years to bridge the gap to state pension) but given further base rate increases seem nailed on, is it prudent just to wait until the next MPC meeting or are rate increases baked in to an extent?
Presumably it’s not as simple as a 25 basis points increase in base rate leading to a 25 basis points increase in gilt yields, but is there any measure of what can be expected in such a case?
There is no formal mechanism I am aware of for translating a BofE rate increase to a 10 year gilt yield increase. The 10 year rate will reflect medium term expectations rather than the immediacy of a rate hike. For myself (but there are no guarantees given in offering this view), I suspect the already elevated 10 year yield has a bit further to go sometime in the course of this year. I have been buying gilts and GBP-hedged US Treasuries in chunks as this year progressed, not wishing to miss out on what seem to be attractive yields for the medium-term.
Thanks for your reply Hospitaller.