- Monevator - https://monevator.com -

How would you fare in a bond market crash?

For most of the past 30 years, nobody has worried about lower prices for government bonds.

In the UK and the US, bond yields – which go down as bond prices go up – have fallen as steadily as TV channels have multiplied and footballers’ salaries have gathered zeroes.

The grind has been so remorseless that for most of the time we didn’t even notice it, like a frog being boiled alive on a beach in the Costa del Sol because he forgot his sun hat.

This wasn’t the sort of bull market that excited private investors. Like most, I didn’t really see a reason to prefer bonds to cash when I started investing – let alone favour them over equities.

Yet bonds went on to beat shares over ever-lengthier historical periods, after two stock market crashes [1] in a decade beat equities up something rotten.

The long gilt yield was very nearly 15% in 1980, and as recently as 1990 it was at 12%. It continued to fall pretty steadily for 20 years – barring a wobble in 1994 that was called a ‘massacre [2]‘ at the time – to approach 5% by the turn of the century. Then, in the nervous wake of the financial crisis, ‘safe’ UK and US bond yields headed unthinkably lower.

In 2012, ten-year gilts sneaked briefly below 2%, which would have seemed about as plausible to a red braces wearing Yuppie in the City in the 1980s as a mobile phone that didn’t do double-duty as a doorstop.

[3]

Bond yields have fallen for 30 years. What goes down must go up?

Yields are unlikely to fall much further on this graph [4] – if only because at 2% there’s not much further for them to go. Theoretically, yields could halve again to 1% – or less –which is the sort of thing that happened in Japan. But at current levels I’m reminded more of Zeno’s paradoxes [5].

Indeed, scars hard-earned on the investment battlefield make me much more worried about falling bond prices than missing out on further gains or forgoing their 2% income a year. That is why I personally hold no government bonds, and prefer cash for ballast.

But how fearful should those holding UK government bonds (aka gilts) really be?

Italian, Spanish, and even Greek bondholders were once pretty confident about their ‘safe’ investments, too, and look where it got them.

How far would your own bonds fall if yields rose in a bond market crash?

A word of warning to would-be traders

Before I show you how to estimate how your bond holdings could fare in a sell-off, a word of warning.

My co-blogger, The Accumulator, has already asked whether pure passive investors should sell their bond funds [6] ahead of a correction, and concluded they should not.

And despite my bearish stance on bonds, I agree with him.

Firstly, a bond crash is by no means inevitable – do a Google and you’ll find people have been expecting one in the UK and the US since at least 2008 (which was when I first got worried). Yields could bump along the bottom like this for many more years if we continue to stumble in and out of recession.

Secondly, even if you do successfully move out of government bonds, are you really ready to become a more active investor – with all the hassle and likelihood of inferior returns it entails? Will you know when to get back into bonds, or how much to allocate to equities instead, or what to do if bond prices keep rising instead of falling like you expected?

The evidence is pretty clear that most people would do best to invest passively and rebalance between their asset classes periodically. Like this, you’ll automatically top-up any declining bond allocation from your other assets that are hopefully doing better. And you’ll benefit from the secret wisdom of passive investing [7], which is that it works when you follow the plan, rather than vacillating with the headlines.

Read The Accumulator’s article [6] on whether to sell your bond fund if you’re on the passive camp. Then by all means read the rest of this article for your erudition – but not, I’d suggest, ahead of taking evasive action.

How bond prices will fall when yields rise

Most people hold their bonds via a bond fund or ETF. However it’s easier to first explain how bond prices – and hence the value of a portfolio of bonds – can fall by looking at what happens with individual bonds.

We need to establish a few things:

Please read my articles on UK government bonds [9], how to calculate yields [10], and what drives yields [11] and prices higher and lower if you don’t already understand the vital differences between bonds, cash [12], and equities.

The key point is that bonds deliver a fixed return at whatever price you buy them for. This return is made up of all the income you’ll receive for the life of the bond, plus the face value of the bond that you’ll be repaid when it matures.

By combining these two income flows (income plus return of capital) and looking at the years left to run on the bond, you can calculate the redemption yield, which is the estimated annual return you’ll get from the bond if you hold to maturity, assuming you can reinvest the income at the same rate.

Redemption yields are nearly always positive. This is vitally important to appreciate when anyone talks to you about a bond market crash. If you hold a UK government bond (gilt) to maturity, you’ll get your money back via the total return.

For instance five-year gilts are currently priced at £140. When they mature, the holders will be repaid the face value of £100 per bond.

By the looks of it, then, anyone buying them today faces a certain loss of just over 28%.

However this five-year gilt began life with many years to run, and was initially issued when rates were high. The coupon on the bond is 8.75%, entitling the holder to £8.75 in income a year for the life of the bond.

That adds up to £43.75 over the five years remaining, which when added to the £100 face value means a buyer today can expect to receive £143.75 back from their investment, or a small profit of £3.75. Do the maths (or look it up [13]) and you’ll find this is equivalent to an annual redemption yield of 0.69%.

In other words you’ll expect a positive total return from this bond, even though you face a certain capital loss.

This is very different to equities, where neither capital returns or dividends are ever certain in advance. That’s the first huge thing to realise if you’re fearful about a bond market collapse.

(I am not going to go into real returns in this article, which take into account inflation. Of course 0.69% is a terrible real return with inflation over 2% and hideously unattractive in my view, but it is the going rate as I write and whether it is sensible or not is not the subject of this piece).

Duration, and why it matters

So far, so arcane – such is the world of bonds – and now for another wrinkle.

Remember I said that redemption yield calculations assume that you can reinvest your coupon at the same rate – but also that yields rise and fall over time?

Clearly there’s a conflict here. Redemption yield is a useful approximation, but that’s all it is, as higher or lower yields may be available over a bond’s lifetime of reinvesting its coupon.

If the bond has only a year or two to run, you can be pretty confident of the rate you’ll get when you reinvest. But what about when you’re reinvesting in 20 years? All bets are off, as interest rates are likely to fluctuate all over the place during that time.

This means that bonds become riskier the longer they have left to run – which is why long bonds (those with ten or typically many more years to run) pay a higher rate then short-dated bonds.

Risk and reward go together, remember? If your friendly neighborhood bond peddler wants you to buy a 30-year bond when one-year bonds are available sporting very certain returns, you’re going to want a higher rate of return to compensate you for tying your money up for 30 years of sleepless nights.

Bond traders wised up to this years ago. While the rest of us drank, danced, and looked forward to the advent of cheap air travel, they struck upon a measure called duration, which is properly defined as how long (in years) it takes an investor in a bond to get her money back through interest and capital payments, but is commonly referenced as the sensitivity of a bond to interest rates.

The bond wonks then went a step further, and created modified duration, which is a slightly more accurate measure of the sensitivity of a particular bond to interest rates. I’ll (incorrectly) use the terms pretty interchangeably from here.

You can work out the duration [14] of a bond for yourself using a formula, but I wouldn’t bother. It’s available via various proprietary data sources [15], and I’ve just discovered you can also download it via the government’s own data on daily gilt prices [16] via the DMO website. You can also try plugging numbers into a duration calculator [17].

If you do so you’ll see a long list of gilts in issuance, and you’ll notice that duration increases the longer a bond has left to run.

This makes perfect sense, since as we’ve discussed you’re more uncertain about interest rates the further out you go.

You’ll also see that duration is always less than the time left to maturity.

Again, pretty logical; you’ll benefit from the cash flows from the annual interest coupon, as well as the principle repayment at the end, so you’ll get your money back before the bond matures1 [18].

Finally, duration is affected by yields [19] and the coupon rate on the bond, as well as by how long it has left to run. Duration falls as yields increase, and vice versa.

This makes perfect sense, too – if you’re getting 8% a year on a 20-year bond, it’s going to take you a lot less time to get your money back than if you are being paid 2% a year.

This final factor has especially big practical ramifications when rates are very low, like today, as I’ll get onto later.

How bond prices will fall when yields rise

So we have two vital pieces of knowledge:

Together this means that if you hold a portfolio of gilts and intend to run them all to maturity, you can assume that you are going to get your money back, with a bit of interest.

However you also know that their value will go up and down until then as interest rates fluctuate, according to their duration.

Let’s put all this together with some examples.

To work out how much a particular bond will rise or fall if its yield to maturity rises or falls, you multiply its duration by the hypothetical change in interest rates. (Remember that as yields of bonds rise, prices fall, and vice versa).

For instance, consider a ten-year gilt maturing in 2022 (named 4% Treasury Gilt 2022), which I see from the DMO website has a modified duration of 8:

If the yield rises by 0.5%, the price will fall by 4% (because 0.5 x 8 = 4)

If the yield rises by 1%, the price will fall by 8%

If the yield rises by 2%, the price will fall by 16%

On the other hand:

If the yield fell by 0.5%, the price would rise by 4%

Another way to work out how your bonds will change in price is to use a bond calculator [20], and to try out various different scenarios. You should find the numbers returned are the same as those you’ll calculate using duration.

How do you work out how bond funds might fall?

Exactly the same technique can be used to calculate how bond funds will fall as interest rates rise [21].

The trick is to use average yield and duration data that factors in all the bonds in a fund’s portfolio.

Such data should be available in your fund’s latest fact sheet. For example, here’s bond data [22] for the iShares short-dated gilt ETF (Ticker: IGLS).

How much should you worry about a bond market crash?

Now you know how vulnerable your bond holdings are to a sell-off and rising yields, what if anything should you do about it?

As I said at the top, there’s a strong answer for saying you should leave well alone. While it may seem a no-brainer that rates will rise and bond prices fall, it’s seemed that way before:

Here’s a fund manager writing in the FT [23] in August 2010:

“All told, benchmark ten-year bond yields may normalise at rates between 3.25% to 3.5% in the eurozone and 3% to 4% in the US in the course of next year.”

Actually, rates on US ten-years are now less than 2%, and nobody talks about ‘eurozone rates’ any more, given all the divergences.

By December 2010, other FT pundits [24] were growing more confident a crash was coming:

“A great 28-year bull market in bonds in its dying throes, and inflationary pressures building, unless leverage and herding behaviour have suddenly become a thing of the past, no investor should be surprised to find that bond markets are vulnerable to sharp and painful adjustments, of which last week’s movements are a foretaste.”

Actually, UK gilt yields have approximately halved since then to their low point this summer.

Even more amusingly, we can go back to 2006, and this FT piece [25] entitled A dangerous bubble in the gilt market:

“Real interest rates on long-dated government gilts have fallen still further, in some cases to below half the level in the US. This is a bubble on top of what may be a global bond market bubble.”

I could go on and on, back to the 1990s or even the 1980s.

The point is not that these writers are idiots – far from it. I was worried about a bond bubble [26] in 2008. It’s that the future path of rates is very uncertain, however certain it looks at any time.

For most people, bonds are in a portfolio to protect against stock market crashes and to reduce the volatility of returns. Private investors are not advised to trade bonds for the best outcome (and I repeat again that held to maturity, all UK gilts will currently deliver a positive nominal return). Given how gilts have confounded the experts it is probably foolish to try.

That said, I hold no gilts, and it’s partly because they give me the willies. Call me foolish, but I prefer to hold cash, for its higher yield and optionality2 [27].

What makes bonds particularly risky at the moment is the low yields you get for holding them. As we’ve seen, this increases duration, and so makes them much more vulnerable to interest rate shocks.

You don’t need to be a bond market nerd to understand this. It’s simple maths.

Total return in a year from a bond = Change in price + income received

In normal times, a decent yield will shore up your total return. For example, if a bond yielding 5% falls 10% in a year, the total loss for the year would be -5%.

In the great bond collapse [28] of 1994, the Vanguard fund that tracks the US bond market saw a negative return [29] of roughly -3% for the year! (It did fall further intra-year). Hardly the stuff of nightmares, is it? All those interest payments ameliorated the capital loss (which proved temporary anyway, as yields soon resumed their march down, and hence prices rose).

When yields are very low, this safety cushion is not available.

We’ve had a few discussions about the wisdom of holding bonds on Monevator in the current climate, and I suspect some readers think I’m complacent about the risks. I’m really not complacent, as I hope this article demonstrates.

However you’ve got to realise that the average person would have sold gilts years ago if they were looking to trade them, and may well have bought them back at a higher price as yields continued to fall. The bond market is inconceivably deep and liquid, and is currently being buffeted about by abnormal factors like QE, too. You think you know better than it does at your peril.

A more pragmatic response if you’re concerned about today’s low yields but want to keep owning bonds would be to hold bonds (or bond funds) with a lower duration. As we’ve seen, this reduces their sensitivity to interest rates, so you’ll not be hit as hard by a crash should rates rise.

I don’t hold any gilts, as I say. Cash suits me fine as a buffer.

You pays your money and takes your choice.

  1. For short-dated bonds, and given the low yields of today, we might better say you will theoretically get your money back before you are repaid the principle. In practice, the lump sum principle repayment will be required for short and medium bonds and even most long bonds at low yields to return your money, plus a bit more. [ [34]]
  2. That is, the ability to quickly deploy it into another asset class such as shares or bonds. [ [35]]