≡ Menu

Should you invest in short or long-term government bonds?

Should you invest in short or long-term government bonds? post image

This article about time horizon and asset allocation is by former hedge fund manager Lars Kroijer, an occasional contributor to Monevator. He also wrote Investing Demystified.

Previously we’ve discussed how you should choose the minimal risk asset that will form the bedrock of your portfolio. The short version: If you have high-rated government bonds available in your base currency, they will typically be the best choice.1

Today we will look at the time horizon of your minimal risk investment, what returns you can expect to make, and how best to buy it.

Match the time horizon

In most discussions about minimal risk investments, the assumption is you’re talking about short-term bonds. Longer-term bonds have greater interest risk (i.e. they fluctuate in value with changes in the interest rate), and it’s not clear that all investors need to take that risk.

Consider a one-month zero-coupon bond and a 10-year zero-coupon bond that both trade at 100. (Zero-coupon bonds are a particular kind that don’t pay interest, only the principal back at maturity).

Suppose market interest rates suddenly go from zero to 1%. What happens to the value of these bonds?

A table showing how short-term zero coupon bonds are not vulnerable to interest rate risk, compared to longer-dated equivalents.

As you can see, the one-month bond declines only a little in value to reflect the higher interest rate of 1%. This bond matures in just a month, and at that point the bond holder will be able to reinvest their principle at higher prevailing rates if they want to.

In contrast, the 10-year bond declines to a value of around 90.5 as it reflects the higher interest rate. Clearly something that can go from 100 to 90.5 fairly quickly is riskier – even if the probability that you will eventually be paid in full has not changed.

The time horizon for the vast majority of investors exceeds the maturity of a short-term bond, however. Also, someone who is interested in maintaining a position in the minimal risk asset for five years will be taking an interest rate risk over that five-year time horizon, whether they’re buying new three-month bonds every three months, or buying a five-year bond and holding it to maturity.

Finally, the situation is dynamic. If your time horizon is such that you think five-year bonds make sense, you shouldn’t necessarily just buy a five-year bond and hold it to maturity. A year hence, your bond would only have four years to maturity and therefore no longer match your time horizon.

The best way to address these issues is not for you to constantly buy and sell bonds to get the right maturity profile (in my example selling the now four-year bond and buying another five-year bond), but rather to invest in the bonds through a product like an ETF or investment fund that trades the bonds for you.

Such funds offer exposure such as ‘Germany 5–7 years (to maturity) government bonds’, ‘UK 10–12 year government bonds’, and so on. Buying one or a couple of these products to match your desired minimal risk asset and maturity profile is a cheap and easy way to ensure you always have your chosen minimal risk exposure.

Investors with a longer time horizon should buy longer-maturing minimal-risk bonds. As a reward for taking the interest rate risk associated with longer-term bonds, you’ll typically enjoy a higher return than you’ll get from short-term bonds.2

Go longer to reflect your investment horizon

If you need a product that will not lose money over the next year, then you can pick short-term bonds3 that match your profile.

However if – like most people – you want a product that will provide a secure investment further into the future, choose longer-term bonds and accept the attendant interest-rate risk.

Here you should consider the time horizon of your portfolio and select the maturity of your minimal risk bonds accordingly.

If you are matching needs far in the future (such as your retirement spending) then there is certainly merit in adding long-term bonds or even inflation-protected bonds to your portfolio.

Long-term bonds compensate investors for interest-rate risks by offering higher yields. You also have the further benefit of matching the timing of your assets and needs.

You can also mix the maturities of your minimal risk assets. You may have some assets that you won’t need for decades, and others you think will be needed in five to seven years, say. In that case, there is nothing wrong with picking a couple of different products with different maturities to match that profile.

What will the minimal risk bond earn you?

Even people with only a peripheral interest in finance know that interest rates are near historical lows. Nobody should expect to make a lot of money investing in the minimal risk asset in any currency right now.

In fact, with nominal interest rates near zero, inflation means that investors in short-term government bonds will experience negative real returns.

While your $100 invested in a government bond will almost certainly become $105 in five years’ time, the purchasing power of that $105 will be less than that of your $100 today. This is, of course, still better than if you had held the $100 under the proverbial mattress for five years – in that case the purchasing power would be even lower.

There’s not a lot you can do about this. If you are after securities with minimal risk then the yields are just very low at the moment. At the time of writing, cash with FSCS protection may be a better option for private investors, as we discussed last time. (Over the long term, returns from cash have historically been lower than from bonds, on a market wide view).

Elsewhere, instruments that offer much higher returns come with more risk of not getting paid. Anyone who tells you otherwise is not telling you the whole story.

A charting showing US Government Bond Yields and Real Yield

The chart above shows what US government bonds (so in $) will currently earn you, by maturity, both in real and nominal terms (you can easily find it for other currencies if you Google ‘£ Government bond yield curve’, and so on.).

You can see, for example, that if you buy a 20-year US government bond, you can expect to earn just under 1% real return per year . Likewise, for a five-year bond you can expect just over a zero percent real return per year. (Five-year US government bond returns yields have actually been negative for much of the recent past.)

While the outlook for generating very low-risk real returns is fairly limited at present, these are continuously moving markets. Keep an eye on them as rates change.

Nominal yields, real yields, inflation, and taxes: The previous chart also shows you the current market expectation for future inflation. (It’s the difference between the two lines.) If the markets are assuming there will be roughly 3% annual inflation in the US for the next 25 years but only around 2% for the next five years, this suggests higher inflation in the longer term. Inflation is bad for many things, one of which is tax. While the benefits you get from your investments are based on real returns and the future purchasing power of your money, you pay taxes on the nominal return. Suppose you invest $100 for a nominal return of 2% the following year. You could be liable for tax on your $2 gain, even if 2% annual inflation had eroded the real gain4. Compare that to a zero inflation rate environment. With a 0% nominal and real return, your $100 would still be $100, both in real terms and nominally, at the end of the year. And there would be no gains to be taxed on!

If the previous chart gave you the sense that the return you’ll get in any one year from owning US government bonds is stable, reconsider.

The next chart shows the annual return from holding very short-term (less than one year) and long-term (more than ten years) US bonds since around the Depression.

A chart showing Inflation adjusted US government bond returns since 1928

As you can see the annual returns move around a fair deal for both kinds of bonds – but far more for the long-term bonds. This is because such bonds will move in value much more as the interest or inflation rate fluctuates.5

Some takeaways from the two graphs:

  • Real return expectations from these minimal risk assets are currently near historic lows.
  • Returns from these minimal risk assets have fluctuated quite a bit, because of changes in inflation and real interest rates, and can reasonably be expected to do so in the future.
  • You can generally expect higher returns from investing in longer-dated bonds. If that matches your investment horizon, then hold your minimum risk bond portfolio through an ETF or index fund. But be prepared that particularly for longer-dated bonds, the yearly fluctuations in value can be significant.

Buying the minimal risk asset

Because of the costs involved in trading bonds, most investors in short-term bonds have to accept that in most cases the bonds in their portfolios will not be super short term6, and that you will be taking a little bit of interest rate risk as a result.

The most liquid short-term bond products like ETFs or index funds have average maturities of 1–3 years. The slight interest rate risk that comes from holding such bonds is a reasonable compromise between the theoretical minimal risk product and one we can actually buy in the real world.

For most investors with longer-term investment horizons, there are funds with different ranges of maturities like 5–7 years, 7–10 years, and so on, to suit your preferences.

How much of the minimal risk asset you should have in your portfolio and what maturities it should comprise depends on your circumstances and attitudes towards risk.

If you’re extremely risk averse, you might put your entire portfolio into short-term minimal risk assets, but you should not expect much in terms of returns. As you add more risk – mostly by adding equities – your potential returns will increase, and vice versa.

Varying the amount of minimal risk asset you hold in your portfolio adjusts the risk profile.

In the simplest scenario, where you only choose between the minimal risk asset and a broad equity portfolio, you could weigh the balance of those two according to the desired risk. The minimal risk bonds would have very little risk, whereas the equities would have the market risk. How much risk you want in your portfolio would be an allocation choice between the two (in a future post I will discuss adding corporate bonds).7

For some investors, putting 100% of their money in the minimal risk asset is their optimal portfolio. This would be appropriate if you are unwilling to take any risk whatsoever with your investments – and if you accept this means very low expected returns!

Summary

  • If your base currency has government bonds of the highest credit quality (£, $, €) then those should be your choice as the minimal risk asset.
  • If your base currency does not offer minimal risk alternatives, you have the choice of lower-rated domestic bonds where you take a credit risk, or higher-rated foreign ones where you take a currency risk. Keep in mind that any domestic default would probably happen at the same time as other problems in your portfolio, and your domestic currency would probably devalue. That would render foreign currency denominated bonds worth more in local currency terms.
  • If you want the lowest risk you should buy short-term bonds. If you have a longer investment horizon, then match the maturity of your minimal risk bond portfolio with your time horizon. You will have to accept interest rate risk, even if you avoid inflation risk by buying inflation-linked bonds.

Video on your low risk portfolio allocation

Here’s a video that recaps some the things I’ve discussed in this article. You will also find other relevant videos on my YouTube channel.

Lars Kroijer’s book Investing Demystified is available from Amazon. He is donating any to medical research. He also wrote Confessions of a Hedge Fund Manager.

  1. Private investors may also want to consider cash as part or all of their minimal risk asset allocation. Please see my previous article for more details. This article will focus on the traditional minimal risk building block – high rated government bonds. []
  2. There are cases where the yield curve is reversed and shorter-term bonds yield more than longer-term ones, but these cases are less frequent. []
  3. Or cash. See the previous article on the minimal risk asset. []
  4. i.e. The purchasing power a year hence would still be $100 in today’s money even if the nominal amount had become $102. []
  5. The market view of credit worthiness will also have played a role. []
  6. Imagine the scenario where you want to hold one-month government bonds. Tomorrow the bonds are no longer one-month to maturity, but 29 days. Is this ok? How about 2 days hence? How much you are willing for the maturity to deviate from exactly 30 days is up to you, but in reality there is a trading and administrative cost associated with trading bonds. It would simply not be feasible to stay at exactly 30 days to maturity at all times. []
  7. Certain corporate bonds trade with lower risk premiums than many governments. The view is that these corporate bonds are lower credit risk than many governments – not hard to believe – and although they do not have the ability to print money, nor do governments in the eurozone. The reason I believe that you should not consider these bonds as the minimal risk asset is more practical. Compared to government bonds, the amount of corporate bonds outstanding for any one company is minuscule and you would probably not be able to trade them as cheaply and liquidly as government bonds. []
{ 38 comments… add one }
  • 1 Walter @ Walbrock Research April 25, 2017, 4:46 pm

    Excellent article Lars,

    I always thought the reason longer-term bonds returns were positive was that interest rates have tracked lower since the 1980s.
    One reason was lower inflation, as Western companies seek cheaper labour costs aboard.
    At the same time, the median household income stagnated, especially in the U.S.
    When China joined the WTO in 2001, that brought in more benefits for Western businesses as Chinese exports helped to temper inflation concerns.

    Now, Western economies are over-leveraged and rely on cheap financing costs and further borrowings from creditor nations like China.
    There is no way interest rates will manually go up because wages in the West can’t handle the increase in mortgage payments. While at the same time, the creditor’s nations can’t keep lending to the West.

    China got their credit-fuelled bubble.

    But the disadvantages of lower interest rates are causing asset bubbles in Western Markets. And one way or the other, the people will feel the pain if interest rates stay low or go up.
    I’m leaning towards investing in shorter-term bonds, in the case of interest rates suddenly spiked higher. At the same time, if there is a financial crisis bond rates are near historical lows.

    One would assume bond yields can go negative like in Japan and Switzerland.

  • 2 dawn April 25, 2017, 6:05 pm

    Good lord its hard to get my head around bonds
    Especially when hes talking from a US standpoint
    Im 40% in cash. I fix it in lumps… . short term.
    Yes ,with inflation im probably loosing in real terms but only 1 – 2%
    Equities can fall 50 %
    I have 2 portfolios now…
    1] Is for 15 years time when i reach 67 and hopefully get state pension its 100% Equities
    2] Cash plus a smaller equity allocation for up to 67

  • 3 Naeclue April 25, 2017, 6:35 pm

    This may seem pedantic, but the example of a 9.5% loss on a 10 year bond if the interest goes from 0 to 1% is misleading. That would only happen if the yield to maturity of the bond went from 0 to 1%. In other words a parallel shift of 1% in the yield curve, or if the bond carried 0% interest, the 10y rate went from 0% to 1%. Parallel shifts like that are very rare and this is the sort of statement that puts people off investing in bonds. They mistakenly assume that the next time interest rates rise, the value of their bonds will automatically fall.

  • 4 The Investor April 25, 2017, 7:05 pm

    @Naeclue — Hi, I don’t think you’re being pedantic exactly, but to put the other side the article is 2,238 words (!) long and already full of caveats and explanations (look at the footnotes alone). I already suspect many people will skip it.

    The illustration was meant to show how the price of longer duration bonds will be more impacted by rising rates than short-term bonds, which I believe it does.

    @Dawn — Keep at it! 🙂 Eventually it will click. Easiest to get your head around individual bonds first. With high-rated government bonds you know exactly what you’re going to get in nominal terms over the life of your investment when you buy (catastrophic government default not withstanding — pedantic caveat for @Naeclue 😉 ) which is why they are so much lower-risk than equities, despite people who don’t understand that saying they’re very high risk.

    What they should say is the risk is very low, as always — but right now the likely return is very low, too.

    @Walter — Yields have been around zero in Japan for decades. It could happen here too, though I think it won’t. Personally I suspect we’ll be at about 3-4% on the US 10-year Treasury in five years. But such speculation isn’t really the point of this article so I’ll leave it there. 🙂

  • 5 Naeclue April 25, 2017, 7:08 pm

    Lars, I am surprised that you think it best to buy an ETF rather than a bond ladder. For example, if the maturity someone wants is about 5 years, they just have to buy gilts maturing in 3, 4, 5, 6 and 7 years. Then each year (coinciding with annual rebalancing) they sell the short dated gilt and buy another 7 year gilt. Costs are low as only a fraction of the ladder is traded each year.

    Sometimes the ladder can get unbalanced, for example after a very large rise or fall in share prices, which might require selling and/or investing in more than one bond, but on the whole I have found that to be unusual. I have not really cared too much if the weighted maturity is out by a couple of years though.

    I do agree that an ETF would be simpler and I might go over to that at some point now that ETF and bond funds are available for reasonably low charges. We do still have some difficulty in the UK though as the more specialised ETFs you talk about (5-7 year, etc) have not been available, or not at low cost, so 2 ETFs of short and long duration might be needed, weighted to get the desired duration. This then starts getting complicated again.

  • 6 Naeclue April 25, 2017, 7:12 pm

    @TI, yes fair enough. I just wanted to try and dispel the myth that bond prices are hard-wired to Bank of England interest rate changes. Something that most financial media repeatedly like to imply.

  • 7 Naeclue April 25, 2017, 7:56 pm

    Being pedantic again I am afraid, but this is not true “you pay taxes on the nominal return”. In the UK, for unsheltered gilts and some corporate bonds, you pay income tax on the interest, not on any capital gains. Foreign currency bonds are subject to income tax on the interest and capital gains tax on any profit after disposal or redemption. Buy bonds in a fund or ETF though and any profit made on disposal of the fund does become liable to capital gains tax – even for gilt funds/ETFs.

    This is quite important as the income tax on a gilt with an 8% coupon will be much more than the tax on an equivalent duration gilt with a 2% coupon, even though nominal returns will be very similar.

  • 8 The Investor April 25, 2017, 9:40 pm

    @Naeclue — Hmm, again not really pedantic, it’s a fair point, it’s just when I subbed the piece I read that box as usefully talking more generally about investment theory than about UK bond returns per se. (But again you can see the difficulty? Do we add another two paragraphs saying “but this isn’t true for individual gilts” etc).

    Maybe I can add a few words to that effect, I’ll wait for Lars to offer some input as the email has already gone out now so no rush to fix.

    I’ve mentioned no capital gains on gilts in articles on CGT, but don’t think I have a specific bond tax article to refer to. (Perhaps I could will link to your comment. 🙂 ) Also I’m up-to-speed with gilt taxation but not Treasuries.

    (Instinctively I’d think any sovereign bond would have to have no capital gains tax for that market to be efficient but I don’t know and stand to be corrected).

  • 9 John B April 26, 2017, 9:42 am

    With a bond fund would I expect capital gain after market fluctuations, as I would for a income equity fund, or would I expect it to be flat? I know that individual gilt values trend towards their redemption value, but then you buy them at a discount/premium. For the ladder effect in a fund, does that mean capital gains or losses, allowing for inflation or not?

    (I wish I had a feel as to the dividend:capital growth ratio for equity too, its so hard to find someone unpicking the total return index compared with the FTSE index and inflation to say: “Over 50 years you’d have got 3.5% dividends and 2% capital growth over inflation”, or whatever the real numbers are)

  • 10 The Rhino April 26, 2017, 1:14 pm

    @NC – when you say ‘I just wanted to try and dispel the myth that bond prices are hard-wired to Bank of England interest rate changes’ is that different to saying existing bond prices are hard-wired to newly issued bond prices of a similar duration?

    My understanding is that an existing bond’s price has to fluctuate to make its yield the same as any newly issued bond, i.e. the price must change to arbitrage away any difference in yield of old bonds compared to new bonds (of similar duration)? Otherwise you’ve got two identical products on sale at different prices, which can’t last too long in an efficient market? Or have I missed something?

    I note the maths is missing from the OP on how the values change with respect to duration and change in interest rate, and my teacher said you should always show your working out. Then again, Hawking did say book sales half for every equation you include so I guess its swings and roundabouts?

  • 11 zxspectrum48k April 26, 2017, 1:28 pm

    @Rhino. You can easily have two government bonds which have the same modified duration and yet have different yields for a whole host of reasons. There might be a differing liquidity premia due to say one being the 10-year on the run benchmark but another being an old less liquid off the run 10-year bond (say a 20-year that was issued 10 years ago). The two bonds might repo at different funding levels which will impact the forward breakeven yield (and carry). One bond may be the cheapest-to-deliver in the basket of bonds which is used for a bond future. They might have different coupons which may make them less/more attractive to certain investors from a tax perspective (i.e you currently have 10-year gilts with coupons from 1.25% to 6%).

  • 12 The Investor April 26, 2017, 2:57 pm

    I note the maths is missing from the OP on how the values change with respect to duration and change in interest rate, and my teacher said you should always show your working out.

    Now I’ve started on this subject (see my comments to @naeclue) I may as well continue.

    I really can’t exaggerate to you guys how Monevator is being paralyzed at the moment by the enormous size of the posts we’re writing, partly to try to cover every last eventuality, without being just a run of bullet points.

    Personally I have about half a dozen posts half-written because of this (including that still not done Lifetime ISA follow-up). The Accumulator is still writing his book 2.5 years on because rather than the original ‘let’s just collate a lot of our articles in one handy format’ he’s ended up writing the whole thing from scratch! 🙁

    True, this is my problem as the site’s publisher, not yours. But readers need to understand not everything can be in every article we publish, sadly. Every post — even a 2,200+ word one — is not a book! 🙂 This is what links and Search tools and the rest of the web is for.

    How I hanker for the days when we could post a 700 word article, get a cheery thumbs up in the comments from our three readers, and get on with the next one. Besides posting paralysis, I suspect we’re actually losing a lot of readers trying to cater for the completists (which include ourselves, true) and boring most people to death.

    #firstrankbloggerproblems

  • 13 Comet April 26, 2017, 2:59 pm

    I’m in a similar position to Dawn. I’m currently 45% cash, with only a small % of bonds through my Lifestrategy holdings.
    I’m trying to improve my understanding of bonds, but all I see is a smaller yield than cash, and capital risk if and when interest rates rise. I do understand the value of diversification, and the role bonds are supposed to play in a portfolio, so I’m planning to add some corp bond funds and also LS40.

  • 14 The Investor April 26, 2017, 3:06 pm

    @Comet — Lars is going to do a post on corporate bonds soon, with a bit of luck. We’ve also written the article below, which might help. It’s good to try to understand bonds, but at the moment there’s not much penalty for retail investors sitting in cash IMHO, provider you’re getting an above average interest rate. Good to get knowledge ready for when that situation changes, though.

    http://monevator.com/bond-asset-classes/

  • 15 The Rhino April 26, 2017, 4:14 pm

    @TI – fair point. Maybe keep articles brief then allow any blanks to be filled in via the comments? I would wager most newcomers wouldn’t bother reading the comments so it shouldn’t cause to much confusion? I did investigate the maths over on investopedia and it got involved quite quickly, prob right call not to cover it in the OP, i.e. take Hawkings, not my maths teachers advice.
    @ZX – its fair to say i have no clue what you’re talking about. Sounds like you know your onions though.

  • 16 The Investor April 26, 2017, 4:56 pm

    @Rhino — You see, I do understand what @ZXSpectrum is saying, and Lars would enough to actually write it without having to double check. But as I said, it’s a never-ending rabbit hole from the perspective of bringing more into an article.

    The key thing to understanding what he’s saying is that expected returns from individual bonds comprise both capital returns (gains/losses) and the fixed coupon you’re regularly paid, which in most cases won’t fluctuate (another rabbit hole there, because there are special kinds where it does. You see? 🙂 )

    The yield to maturity — known at purchase, presuming you hold to redemption — reflects both the nailed-on capital appreciation (/depreciation) as the bond returns to its face value, and the stream of coupons you know you’ll get paid until then.

    http://monevator.com/how-to-calculate-bond-yields/

    So depending on the initial coupon and the price you pay in the secondary market for your bond, some/or most of the return may come from capital, and some/or most may come from income via those coupons. And as you know capital and income have different tax implications.

    Hope that’s slightly clearer. (If it’s not that is basically why Lars and @TA suggest most everyday investors just buy a bond ETF to match their time horizon… 😉 )

    The stuff about liquidity is interesting to nerds like me, but very much in the weeds for anyone not sitting on a bond desk in the City.

  • 17 SemiPassive April 26, 2017, 5:33 pm

    Thanks for the article. I get the logic – and when the crash comes you’ll be glad – but still really struggling to buy gilts or a gilt ETF at the moment. I think it would be easier as a US investor (buying US treasuries obvs – at least you get something back to partially counter inflation) but with 0-5 year gilt ETFs yielding close to zero after expenses, and even holding an individual 5 year gilt gives you just 0.5% yield to maturity.

    So I might just use a 50/50 split between cash and corporate bond ETFs for my “low risk” segment. Look forward to the next article on corp bonds.
    I’ve mentioned before that this is more of a problem within SIPPs than ISAs, at least with the latter you can use separate Cash ISAs for your risk free portion whereas you don’t get any interest on cash held in a SIPP (typically*) – you even have to caveat comments here in case some smart arse corrects you 😉

  • 18 John B April 26, 2017, 6:17 pm

    I can see that if I buy a 2% £1000 10 year gilt with 4 years to go I might pay £1050 or £950 for it, so I can predict my capital gain. But if I buy a gilt fund, how would I know the equivalent discount/premium? Do bond funds advertise book values like investment trusts do?

  • 19 Steve April 26, 2017, 7:34 pm

    I do agree with you that it can be easier to buy a fund and let them do the work on maturities. I have eg the ishares “IGLS” ETF (straight gilts up to 5 years remaining maturities) which just rolls along happily. I also use an Axa OEIC fund for sterling corporate bonds up to 5 years maturities where the management fee is not outrageous compared to the yield (which it probably would be on up to five year gilts). Once/if US Treasury 7-10 year yields get to 3% (now around 2 1/2%) I will probably buy some of that stuff via a GBP-hedged fund to diversify my risk.

  • 20 SurreyBoy April 27, 2017, 12:16 am

    On the risk of too much detail – in posts and responses: There is a risk of the good being sacrificed on the alter of perfection, if you get my drift.

    The good thing about the site is that most articles try and capture the main themes, or main ideas, or key drawbacks of a particular issue or investment approach etc. I have noticed they are getting more detailed, and a danger is this could deter newer readers who happen across the site and who really benefit from being able to easily pick up the basics.

    If this site serves any sort of higher purpose, its taking the likes of me from knowing not very much at all to having at least a decent idea of the nuts n bolts of investing.

    It free of charge, its responsibly written, and is highly informative. When people start banging on about where an article may be “wrong” or “incomplete” because gilt yields will converge differently in 30 years time to what the article suggests – then they are probably on the wrong site. For a reader like me its easy – i just skim over these sort of comments. If id spent hours and hours writing the article i expect id be somewhat irritated or demoralized.

    This site is all about learning, and the authors are the first to call out a useful comment but lets rein back on the technical autopsy. If I spent hours writing an article and my reward was a bunch of armchair warren buffets dissecting every paragraph, i dont think id bother.

    Keep up the good work.

  • 21 Hospitaller April 27, 2017, 7:49 am

    On duration risk, one odd thing I have noticed is that pension companies do not all seem to have wised up to the risk of duration in the present low rate environment when someone is getting nearer retirement. They tell us proudly that they are managing the risk of equity downturns by putting us into “lifestyle” funds where the bond to equity allocation increases the closer we get to retirement. Fair dos. What they don’t tell you is that they (well some of them) are putting us, inside the lifestyle fund, into seriously long-dated bonds (maturities way past the retirement date) where we could suffer a significant drop in value if there was an interim interest rate rise. When I saw this happening in my own case, I moved management of my defined contribution fund out of their hands and into my own via a SIPP.

  • 22 The Investor April 27, 2017, 7:52 am

    @SurreyBoy — Thanks for your comments and words of support. I just want to clarify that I do appreciate the detailed comments / extra information people have left here, and in other articles — and I understand where those who want more are coming from. Everyone is at a certain level, and some Monevator readers (well, commenters) are at that level. Which is fair enough and we’re pleased to have them reading. But as you’ve detected, the can’t please everyone always syndrome is causing some tension on the other side of the screen.

    Also, I think we’ve got the blogger blues. But please let’s not comment further on that here (I know I started it!) as it’s not really fair on Lars nor helpful to readers who want to know more about bonds. Off-topic here. I’ll post something on Saturday. Cheers! 🙂

  • 23 Lloyd April 27, 2017, 11:45 am

    I get the idea of buying into a bond fund that has an average duration that matches your time horizon, accepting that the longer the duration the more volatility there will be along the way. Why do so many financial commentators however say that the interest rate risk inherent in anything other than short term bonds is a risk not worth taking?

  • 24 Naeclue April 27, 2017, 11:46 am

    @Hospitaller, DC pension funds do this because they want to reduce risks in the pension someone will receive. If towards the point of retirement long gilt prices fall, then the size of the pot will fall, but this is mitigated by a rise in annuity rates. Conversely if annuity rates fall, long gilt prices will rise so you have more money with which to buy an annuity.

    If the fund was heavy with equities and a stock market crash happened before retirement, you would be faced with a smaller pot AND the likelihood of lower annuity rates.

    If I had been forced to buy an annuity from my SIPP I would have done exactly the same thing as your pension fund is doing.

  • 25 Naeclue April 27, 2017, 11:50 am

    @Lloyd, I think the answer is that most commentators are clueless about risk.

  • 26 Naeclue April 27, 2017, 12:55 pm

    I hold a long dated US Treasuries ETF in my SIPP and just looked up the returns for the last few years. In 2016 US rates went up by 0.25%, yet the ETF returned 1.33% in dollar terms, even though the average duration of the fund is 17 years and average maturity 25 years. In 2015, with no rate changes the ETF lost 1.36% and in 2014 it gained 25% with no changes in central bank rate. I am not criticising the article, just pointing out that bond prices can move in unexpected ways that override changes in short term central bank rates. I think the key word Lars used was “Unexpected” when referring to the influence of rate rises on bond prices.

  • 27 The Investor April 27, 2017, 1:39 pm

    @Naeclue — Yes, the US yield curve has been flattening, as you probably know. 🙂 Also I am on my phone and can’t check right now, but I’m pretty sure Lars said “a move up in interest rates” not “a move up in the central bank rate”. Market interest rates are not the same as central bank rates. (It’s a bit of a tautology in this context, admittedly).

  • 28 Naeclue April 27, 2017, 1:56 pm

    Quite right, he did. I am not accusing Lars of saying anything wrong here, but the difference between market rates and central bank rates will be lost on many people.

  • 29 The Investor April 27, 2017, 2:33 pm

    @Naeclue — Perhaps, but that brings me back to the “is 2,200 words enough” for one blog article point. (Not having a go, just explaining). Perhaps Rhino’s suggestion that the comments can thrash out the in-the-weeds stuff is the right one. I guess that’s what we’re doing here! 🙂

  • 30 Naeclue April 27, 2017, 6:05 pm

    I appreciate the difficulty. In principal bonds are really simple, but articles and discussions can get very complicated, especially when you get into areas such as duration matching as Lars has done. Ironically equities are far more complicated than gilts, but people seem to have fewer problems thinking about them 🙂

    One difficulty people have I think is simplistic thinking along the lines “Interest rates will rise, so bond rises will fall, so I should sell my bonds now” and forgetting that government bond markets are incredibly efficient. Some people think their small knowledge of bonds and the state of the world enables them to outwit bond markets even though they would not dream of making that assumption about equity markets.

  • 31 grislybear April 27, 2017, 7:19 pm

    Read this post about bonds and other posts on your blog. I thought bonds were supposed to be simple. Take a common etf such as VGOV its got a yield to maturity of 1.08% and a duration of 12.2 years. Now if interest rates go up by 1% over the next year (by whatever means Market, Bank rates) My capital in VGOV will decrease by 12% however the yield will increase a bit because the new bonds purchased by the fund will have higher interest rates. Say the yield to maturity increases to 1.5% (thats a guess). It will take about 7 years to get back to my original investment. Meanwhile assuming inflation over the next 7 years is 2.3% as it is now, will mean that my original investment will be about 20% less.
    Is my take on this scenario correct or am I off piste.

  • 32 Naeclue April 27, 2017, 7:40 pm

    @grislybear, it would be more precise to say if the yield to maturity of gilts of duration 12.2 years went up by 1%, then the price would drop by about 12% (a first order estimate). The yield to maturity would of VGOV would then be 2.08%, because you have increased it by 1%. How long it would take to back the original investment after that is anyone’s guess as it depends on subsequent movements in the yield curve.

  • 33 grislybear April 27, 2017, 8:13 pm

    @Naeclue Thanks for clarifying that, I was toying with increasing the yield to maturity by 1% but couldn’t get my head around it, however your rewording of it does make it easier to understand. I do appreciate that the length of time it would take you to get your money back is dependent on events in the future but I need a simple narrative to understand the concept.

  • 34 The Accumulator April 30, 2017, 11:02 am

    @ John B, comment 9 – I think you’re asking for something like this:

    3% dividend yield + 5% earnings growth (nominal) – 3% inflation = 5% real return on equities
    From Tim Hale, 3rd edition Smarter Investing

  • 35 The Investor April 30, 2017, 11:38 am

    The Accumulator reminds me that he has written an article in the past on the taxation of bonds versus bond funds! You’ll find it here:

    http://monevator.com/bonds-and-bond-funds-taxed/

  • 36 Nicholas Stone August 2, 2017, 3:56 am

    I have three questions about bond allocations. I’ll post them separately to try to keep them clear!

    My first question regards people who aren’t sure where they are going to retire.

    Case in point: me. Currently I have a portfolio allocation of 25% bonds and 75% equities. I’m 37, and plan to invest until I’m 60. I’ll start to start switching more into bonds when I reach 40.

    My bond ETFS are

    10% IGLS (0-5yr UK Gilts in GBP)
    10% IGIL (global inflation linked government bonds in USD)
    5% ISXF (corporate bonds ex financials in GBP)

    My aim is to reduce currency risk at the point when I wish to retire. As I understand it, the best way to do this is to have ETFS in your home currency.

    Since I don’t know which my home currency will be (either GBP or USD) my plan has been to build a portfolio which has funds in both currencies (currently it’s about 60/40 GBP to USD which more or less represents my gut feeling on where I might end up).

    Does my current bond allocation being in two currencies make sense therefore?

  • 37 Nicholas Stone August 2, 2017, 12:33 pm

    My second question is about my bond choices IGIL and IGLS.

    My understanding that the 0-5 year Gilts are a defence against interest rate rises. Inflation linked bonds will outdo the gilts if inflation is higher than market expectations and the gilts will do better than the linkers if the opposite happens.

    So these two bonds constitute a good pairing then?

  • 38 Ruby August 17, 2017, 4:39 pm

    I hope this question is not too far off topic – I want to ask the question of the community and can’t find exactly the right place for it. (I also don’t want to ride roughshod over the questions above). It relates to bonds (albeit International bonds) and Lars Kroijer’s writing. I am trying to set up my first proper portfolio and need all the help I can get!

    In Lars’s book ‘Investing Demystified’ he talks about adding other government bonds i.e. foreign/International bonds that are sub-AA-rated as an asset class for which one can expect higher returns* than UK Government bonds (i.e. the ‘minimal risk asset’). (*Higher returns with the attendant higher risk :)).

    From the book:

    “While in principle you should buy the bonds rated below AA according to their market values, in reality it is not practical for some investors to find investment products that represent so many different countries. Instead you might buy an emerging market government bond exchange traded fund (ETF) and combine that with a product covering sub-AA eurozone government bonds.”

    Has anyone actually tried to implement this?

    There seems to be several options for emerging market government bond ETFs, one example: SPDR Barclays Emerging Markets Local Bond UCITS ETF (ticker: EMDL)

    But, I cannot find any ETF (or other product) that will fulfil the criteria of covering ‘sub-AA eurozone government bonds’. Does anyone have a suggestion for this?

    I’m not sure about combining something like EMDL with a global government bond ETF, as these ETFs are made up in large part by AAA, AA+ and AA rated bonds, so don’t really fulfil the criteria.
    (Examples of global government bond ETFs I am referring to are:
    ‘IGLO’ iShares Global Govt Bond UCITS ETF &
    ‘XGSG’ db x-trackers II Global Government Bond UCITS ETF)

    Does anyone have any thoughts?

Leave a Comment