suit-g30bb5e8eb_1920

Should you invest in corporate bonds?

When I began my investing career, I was pretty ambivalent towards the use of corporate bonds in portfolios. They seemed like the investing equivalent of Stephen Root – nobody really loves him, nobody really hates him, he’s always just kind of… there.

Then, as time went on, I became less and less certain they were adding much of any real value.

This was partly a result of hearing phrases thrown around the office like:

“I can replicate corporate bond returns using just stocks and gilts, so why bother with them?”

“I’m already taking the risk of businesses failing in my stocks, why would I want that in my bonds too?”

“They don’t provide the upside of equities, or the downside protection of gilts. Why do I own these things again?”

And to be honest, these sound like pretty reasonable statements.

But corporate bonds require a bit more deliberation than a few off-the-cuff aphorisms, so I thought it was time I put my thinking down in one place. And that’s what this post is. A hopefully orderly collection of my (current) thoughts on corporate bonds.

Before diving in, it’s worth clarifying that when I’m talking about corporate bonds in this post, I’m referring to your bog standard aggregate corporate bond index – which tends to be mid-high credit quality. I’m not referring to the spicier end of the corporate bond spectrum. There’s no diving into high-yield or emerging market bonds here. Given their risk, they need to be evaluated in an entirely different way, as they’re really more of an equity substitute, so will be the topic of a future post.

With that said, let’s get going.

Contents

 

The basics

 

Let’s get the basics out of the way first.

I know most of my blog’s readers are pretty up to speed on bonds, but in case there are any who aren’t so familiar, I’ll quickly recap what corporate bonds are, why they tend to have higher yields than government bonds, and what their risks are.

We’ll then dive into whether they belong in your portfolio.

 

What is a corporate bond?

 

A corporate bond is debt issued by a company.

An investor who buys a corporate bond is effectively lending money to the company, in return for a series of interest payments. When the bond reaches maturity, the investor receives their original investment back.

Corporate bonds are typically riskier than government bonds, so they usually come with higher interest rates to compensate for this additional risk. The highest quality (and safest, lower yielding) bonds are commonly referred to as “Triple-A” bonds, while the least creditworthy are known as “junk” or “high-yield”.

One important difference between stocks and corporate bonds is that in the event the company goes bankrupt, it must pay its bondholders and other creditors first. Investors who own the company’s shares may be reimbursed for their losses only after all those debts are fully repaid. Corporate bonds are therefore safer than stocks.

 

Corporate bond risks

 

Corporate bonds tend to offer higher yields than government bonds because they come with credit risk, downgrade risk, call risk, and liquidity risk. Here’s what each of those means:

Credit risk: Credit risk is the chance the bond’s issuer will default on its debt obligations. If this happens, the bondholders may not receive their invested principal or interest payments.

Downgrade risk: Downgrade risk refers to the risk of a decline in an issuer’s business, which causes the credit rating agencies to downgrade the issuer’s credit quality. For example, Standard & Poor’s could downgrade a company’s credit rating from an A to a BBB rating, which would likely cause the company’s bonds to fall in price.

Call risk: Many corporate bonds provide the issuer with the ability to redeem the bond prior to maturity. When interest rates drop in the market, bond issuers seek to take advantage of the lower rates by redeeming the outstanding bonds and reissuing at a lower interest rate. Calling a bond puts bondholders at a disadvantage, because once a bond is called, interest payments stop being made on the retired bond, and any new bonds issued by the company will be at a lower interest rate. Call risk has been found to be a negative contributor to corporate bond returns.

Liquidity risk: While there’s almost always people wanting to buy/sell government bonds, corporate bonds aren’t always so easy to transact in. There’s a risk an investor might not be able to sell their corporate bonds quickly due to a lack of willing buyers. Low buying interest in a particular bond can lead to substantial price volatility and adversely impact a bondholder’s total return. In 2008, the market taught investors that significant liquidity risk exists, even in investment-grade bonds. Not only did the spread related to credit risk between Treasuries and investment-grade bonds widen, the liquidity premium also widened. And the weaker the credit, the wider the credit and liquidity premiums became.

It’s worth noting that several of these risks can occur at the same time.

When a company’s in distress, this reduces its ability to pay its debts, which increases its credit risk. Because of this increased risk, ratings agencies may choose to downgrade the company’s credit rating, increasing its downgrade risk. In an effort to improve their health, the company may choose to recall any callable bonds it’s issued to refinance at lower rates, increasing call risk. The bond’s liquidity may also decrease as investors become more reluctant to transact in a riskier company.

How susceptible the bond is to these risks depends mainly on the strength of the bond issuer’s financial health. Those issuers who have strong balance sheets and can withstand the buffeting winds of the market come with lower risk, but lower yields. Those bonds issued by less stable companies come with the benefit of higher yields.

That’s the basics of corporate bonds

 

The evidence: returns, risks and drawdowns

 

Now let’s get to the good bit.

I’ve split this into three sections: the first compares UK corporate bonds to UK government bonds, the second does the same for global corporate vs global government bonds (hedged to sterling), and the third looks at the US.

 

UK corporate bonds vs UK government bonds

 

Going back as far as we have returns for UK corporate bonds, which is since 1998, here’s how they stack up against government bonds from a simple risk/return standpoint:

Corporate bonds vs government bonds risk return - UKSource: Bloomberg. Corporate bonds are the Bloomberg Sterling Corporate TR Value Unhedged – the benchmark for the Vanguard U.K. Investment Grade Bond Index Fund. Government bonds are the FTSE Actuaries UK Gilts All Stocks TR Index – the benchmark for the iShares Core UK Gilts ETF (IGLT). To adjust for duration, I’ve compared the corporate bond index with a duration of 7 years to a weighted mix of the government bond index (12-year duration) and cash (0-duration)

When viewed in isolation, duration matched corporate bonds have outperformed government bonds since 1998, by over 1.5% annually. That’s some pretty useful outperformance. But given the higher risk from the corporates (almost double that of government bonds), the risk-adjusted return win goes to government bonds.

But I’ve never been a fan of looking at assets in isolation. Nobody owns a portfolio of 100% corporate bonds. What matters is how they interact with equities, and the effect they have on an overall portfolio.

When comparing the impact of corporate bonds versus government bonds in a 60/40 portfolio, the annualised 1.5% return advantage we saw for corporate bonds is reduced down to 0.7%:

Comparing 60 40 performance - UK chart

Comparing 60 40 performance - UK tableSource: Bloomberg. Corporate bonds are the Bloomberg Sterling Corporate TR Value Unhedged – the benchmark for the Vanguard U.K. Investment Grade Bond Index Fund. Government bonds are the FTSE Actuaries UK Gilts All Stocks TR Index – the benchmark for IGLT.

In a portfolio context, a 60/40 stock/corporate portfolio returned 0.7% more since 1998, but took on an extra 1% of risk, resulting in almost identical risk-adjusted returns.

So governments vs corporates is pretty much a wash when it comes to risk adjusted-returns.

But what we really rely on bonds for is their crash protection. How do UK corporate bonds perform during a crash when compared to government bonds?

Drawdown chart - UKDrawdown table - UKSource: Bloomberg. Indices are as above.

As a crash protector, government bonds come out ahead. They beat corporates in 5 of the 6 last crashes over 10% – with particularly strong performance in 2008, where corporates fell about half as much as stocks, compared to governments rising 14%.

But again, we’re falling into the trap of looking at things in isolation.

Here’s what the drawdowns look at a portfolio level:

Comparing 60 40 drawdowns - UK chartSource: Bloomberg. Indices are as above.

So although government bonds are the clear winner when viewed in isolation, it’s not quite as clear when you look at a portfolio as a whole.

This is because a 60/40 portfolio’s risk is made up of about 90% from equity, 10% from bonds. So regardless of the type of bonds you have (except maybe high-yield bonds, which act more like equities), because their volatility is so low compared to equities, the exact flavour of bonds doesn’t make a huge amount of difference to overall portfolio risk. They therefore all have pretty similar effects on portfolio drawdowns.

Having said that, the drawdown chart shows most of the time owning government bonds would’ve slightly reduced the drawdowns – with the exception of 2008 when government bonds significantly reduced portfolio losses. It looks like they still have a slight edge over corporates.

So that’s what the corporate vs government bond comparison has looked like for UK investors comparing UK corporate bonds to UK government bonds.

But global bonds are also a good option for UK investors – as long as they’re hedged to sterling. What does the global corporate bond vs global government bond comparison look like?

 

Global corporate bonds vs global government bonds

 

We have slightly less data for our global analysis, only going back to 2000.

Over this time, the risk/return comparison looks to have the same story as UK.

Corporate bonds vs government bonds risk return - globalSource: Bloomberg. Corporate bonds are the Bloomberg Global Agg Corporate TR Index Hedged GBP – the benchmark for the iShares Global Corp Bond ETF (CRHG). Government bonds are the FTSE WGBI Currency hedged (GBP) – although it’s not the benchmark for the iShares Global Government Bond ETF (IGLH), it has almost identical performance, with a longer track record than the actual benchmark (the FTSE G7 Government Bond Index).

Global corporate bonds have stronger returns (an extra 1.5% per year), but have higher risk, so are behind global government bonds on risk-adjusted basis.

In a portfolio context, again it’s almost identical to the results for UK corporates vs government bonds. The 60/40 stock/corporate portfolio returned 0.5% more since 2000 (6.3% vs 5.7%), but took on an extra 0.8% of risk, resulting in an identical risk-adjusted return:

Comparing 60 40 performance - global chart

Comparing 60 40 performance - global tableSource: Bloomberg. Indices are as above.

Again, when looking at crash protection, the results are almost identical for global corporates vs global government bonds. Global corporates did slightly better than the pretty dire UK corporates in 2008, but otherwise the level of crash protection – as well as the relative outperformances of government bonds vs corporates – was similar across all crash periods.

Drawdown chart - globalDrawdown table - globalSource: Bloomberg. Indices are as above.

And again, at a portfolio level, there’s little difference bar the obvious aid from government bonds in 2008. Still, government bonds have tended to come out slightly ahead in a crash:

Comparing 60 40 drawdowns - global chart

Source: Bloomberg. Indices are as above.

The elephant in the room here is the time period over which we’ve been judging corporates versus government bonds. The disadvantage of the datasets for both UK and global corporate bonds is they both start after the monster bull run for bonds began in the 1980s.

Since then, there’s been one trade and one trade only for bond investors, and that’s been “take more risk”. High starting yields in the 1980s combined with large spreads between the yields on offer between corporates versus government bonds has meant the last 40 years has been extremely kind to corporate bond investors Especially so for those sliding down the credit quality ladder to pick up some returns from the riskier end of the bond spectrum. So it’s no surprise that corporate bonds have returned more than government bonds.

(NB: It’s not the falling interest rate environment which has caused the bond bull run – see here for why).

But if we move across the pond to look at some US data, we have monthly returns which include a small sliver of time before the 1980s – which means we can peek at what life was like before the bond bull run took off. And we have some much, much longer-term annual data going back to 1929. This should hopefully provide us with some longer-term conclusions to draw, which can shed some light on the corporate vs government relationship over more than just one bond regime.

 

US corporate bonds vs US government bonds

 

Starting with the monthly data going back to 1972, we’ll apply the same tests as we did to UK and global hedged bonds. The bonds I’ve used here are for an intermediate-term US corporate bond index (9-year duration) against intermediate (10-year) government bond returns. I’ve kept everything unhedged here, so we’re putting our US-investor hats on now, because a) we don’t have a choice, as all the data is unhedged, and b) no UK investor is going to be buying only US corporate bonds anyway – hedged or not. So we’re taking advantage of the longer time period offered to us by looking at unhedged returns in the hopes we can generalise the behaviour of corporates vs governments and apply it to our own UK-centric portfolios.

Here’s how they stack up on simple risk/return stats:

Corporate bonds vs government bonds risk return - USSource: Bloomberg and Shiller data. Corporate bonds are the Bloomberg US Corporate Total Return Value Unhedged USD Index, government bonds are the 10-year government bond return per the Shiller data.

For UK and global hedged bonds since 2000, corporates handily outperformed government bonds. But not so over the longer term in the US.

Here, corporates only have the slight edge over government bonds, adding just under 0.5% of return compared to treasuries. The risk levels of the two were surprisingly almost identical, which shows just how strong US businesses have been since the 70s. I haven’t probed too deeply into the reasons for the lack of volatility of US corporate bonds, but I’d hazard a guess it’s because the default rates for US businesses have been incredibly low (this seems more likely than the alternative – that US government bonds were unusually volatile).

Because of their extremely similar returns, when combining them each with equities, the two resulting 60/40 portfolios performed almost identically:

Comparing 60 40 performance - US chart

Comparing 60 40 performance - US tableSource: Bloomberg. Indices are as above.

The two 60/40 portfolios returned around 9.6%, with the corporate bond portfolio taking slightly more risk, resulting in a slightly lower risk-adjusted return ratio. But still, they were extremely similar.

When looking at crash protection, treasuries come out on top, beating corporate bonds in 8 of the last 10 crashes, by an average of 8%. In the two instances where corporate bonds beat treasuries during a stock market crash, the difference was less than 3% on both occasions:

Drawdown chart - US

Drawdown table - USSource: Bloomberg. Indices are as above.

But at a portfolio level, yet again, the differences in return are diminished.

Comparing 60 40 drawdowns - US chartSource: Bloomberg. Indices are as above.

Although both are similar, it does still look like treasuries have the edge when it comes to drawdown protection.

 

Third party evidence

 

Turning now to some third-party evidence, bond guru (and all-round investment guru, really) Larry Swedroe goes through some evidence put together by his colleague in an article on whether there’s any need to hold corporate bonds.

The research covers the period 1988-2017, and shows that even though investment-grade corporate bonds have tended to have a yield advantage well in excess of 1% compared with comparable-maturity Treasury bonds, the historical return advantage has been only 0.6% per year. A significant amount of the initial yield advantage is lost to downgrades (bonds sold after falling below a BBB credit rating), defaults and companies exercising call options.

This is interesting, as it shows how wary we need to be of assuming the yields shown on corporate bond factsheets. Just because a corporate bond fund says it has a yield of 2%, doesn’t mean you’ll get a return of 2% – you’ll probably end up getting less.

The following table from the paper shows the key results from the paper. For global stocks, the author used a series that is a 60%/40% combination of the S&P 500 and MSCI EAFE indexes; for government bonds, he used the Barclays Treasury and Long-Term Treasury indexes.

Larry Swedroe bond comparison table 1Source: ETF.com

In both comparisons, the analysis shows portfolios only consisting of stocks and government bonds are capable of delivering a similar, if not a slightly better, risk-adjusted return profile when compared with portfolios owning corporate bonds.

The government bond portfolio used for comparison with the 60/40 corporate bond portfolio allocates 68% to global stocks, 26% to the Barclays Treasury Index and 6% to the Barclays Long-Term Treasury Index. (The portfolio compared with the one owning high-yield corporate bonds, which isn’t relevant for this article, allocates 85% to global stocks and 15% to the Barclays Treasury Index.)

This research shows marginally increasing the stock allocation reproduces the basic result that corporate bonds would otherwise provide. The author notes that this shouldn’t be surprising, given credit risk is one of the primary risks driving the returns of stocks as well as corporate bonds. But still, it agrees with what we’ve found that there isn’t much corporate bonds can do which stocks/government bonds can’t.

The second piece of research looks even further back.

The following table covers the 92-year period from 1926 to 2017, and compares the results of two 60/40 portfolios rebalanced annually. Portfolio A’s allocation is 60% to the S&P 500 Index and 40% to long-term treasuries. Portfolio B substitutes long-term corporate bonds for the fixed-income allocation:

Larry Swedroe bond comparison table 2Source: ETF.com

There was just a 0.1% advantage for a portfolio with corporate bonds. Over almost 100 years government bonds returned the same as corporate bonds – which is consistent to what we found with the analysis in the last section.

The author notes the reason most of the higher returns of corporate bonds didn’t show up in the portfolio returns was because long-term treasury bonds mix better with the risks of stocks. Their annual correlation of returns to the S&P 500 was lower than for long-term corporate bonds.

This argument for why government bonds have performed so well when combined with equities is a general one – which is excellent news for us UK investors. It isn’t due to any quirk of the US market, or result of studying a specific time period, but is an idea which is applicable to any investor looking to choose whether to include corporate bonds or not. The similarity of returns between using corporates vs government bonds at a portfolio level is due to the structurally lower correlation of government bonds when combined with equities, which is equal to the higher potential return offered by corporate bonds.

 

Other considerations

 

Costs

 

Before sitting down to write this article, I was pretty confident in what this section would look like. I knew corporate bonds were less liquid than government bonds, and that lack of liquidity tends to manifest in the form of higher trading costs. While I expected OCFs to be broadly similar, transaction costs were likely to be higher for the less-liquid corporates.

But while that might have been true once upon a time, times seem to have changed:

UK bond fund costsSources: Vanguard and Fidelity (for iShares transaction costs)

There’s very little to choose between any of those.

Interestingly, it’s not any more expensive to buy UK corporate bonds than government bonds – all the choices in the table are within 5ish basis points of each other.

And it’s the same story with global bond funds:

Global bond fund costs

Sources: Vanguard and Fidelity (for iShares transaction costs)

It’s not any more expensive to buy global corporate bonds than government bonds – all the choices are again within 5ish basis points of each other.

I’ve ignored the costs of buying US corporates vs government bonds here, as us UK investors aren’t going to be buying US corporate or government bond funds. So who cares which are more expensive.

It’s always satisfying to confirm your prior beliefs, but the data has put me in my place. Contrary to what I thought I’d find going into this post, there’s really no cost advantage to buying government bonds over corporate bonds. As a result, we can’t favour government bonds over corporate bonds simply on account of costs.

Having said that, adding a corporate bond fund into your portfolio is still likely to increase costs.

Adding any extra fund – this isn’t specific to corporate bonds – is going to result in increased trading costs. It’s an extra fund to buy each month if you’re splitting your contributions, and an extra fund to buy/sell when you’re rebalancing. Of course, how much extra you pay in trading costs depends on a) how you make contributions/withdrawals (how often, and whether you split contribution across all your funds), b) how often you rebalance, and c) which platform you’re using. Investors being charged an explicit trading fee by their broker might want to do the maths to figure out how much of an impact adding an extra fund will have.

 

Behavioural risks

 

One final risk worth considering when including another fund into your portfolio is the behavioural one.

The more holdings we hold in our portfolios, the more likely we are to performance chase. I’m using ‘we’ here, because I’m just as guilty of this as anyone else.

Performance chasing is the act of buying more of the stocks/funds/beanie babies which have gone up, and selling the stocks/funds/beanie babies which have gone down. This has the effect of selling at the bottom and buying at the top. It sounds idiotic – what kind of moron would sell at the bottom and buy at the top?

But it’s an incredibly easy trap to fall into. When something’s rocketing in price, it feels like we’re missing out by not owning enough of it (Bitcoin, anyone?) – and when something’s crashing in price, it feels like we’re being reckless by not selling it and stemming the bleeding. But, unsurprisingly, it’s a terrible strategy.

The reason more holdings results in more performance chasing is there’s a higher chance of you seeing a big red number when you log in to view your portfolio. And big red numbers encourage bad behaviour.

If we had a portfolio of a single holding, and logged in to find it 10% up, then we probably wouldn’t care. If we had a portfolio of two holdings, though, and logged in to find one up 60% and one down 40%, then we’d almost certainly feel compelled to do something about this disastrous holding which has crashed 40% – despite the fact the overall portfolio return is the same as the portfolio with the single holding (+10%).

And that’s how performance chasing happens.

This is one of many reasons I favour simplicity over complexity when it comes to portfolio construction: it protects us from our own bad behaviour. The fewer holdings you have, the easier it is to set-and-forget.

So by adding another fund into the mix with a corporate bond fund, it increases the risk we’ll engage in performance chasing if/when corporate bonds have a particularly bad period. This is especially risky when they’re being included as a core part of your portfolio (which is what we’ve been discussing here).

Corporate bonds aren’t the first place you’d imagine performance chasing to impact – bonds don’t have particularly spicy performance, which is why chasing performance is most prevalent in speculative high-growth stocks. Still, I imagine more than a few investors wanted to bail on their UK corporate bond holdings during 2008, as they watched them fall alongside the market while their UK government bond holdings rose 14%. So it’s not an impossibility – I think any increase in the number of holdings increases its likelihood.

Performance chasing isn’t as big a deal if it’s done on a small scale, and unlikely to have a real impact on your portfolio. If it’s done in the small section of your portfolio carved out for random stock punts your mate told you about down the pub, then it’s fine. This is where I engage in my fair share of performance chasing, alongside all the other terrible decisions I make speculating on individual stocks. But when our instinctually bad behaviour is felt on a core part of our portfolio then it becomes more troublesome as it starts dragging on long-term returns.

So that’s a point for keeping things simple and not including an additional fund into the mix.

 

How to replicate corporate bond returns

 

As I mentioned at the start of the post, one of the common criticisms of corporate bonds is that it’s possible to recreate their risk/return characteristics using a combination of stocks and government bonds.

The first question is therefore: “Can you replicate corporate bond returns using only stocks and government bonds?”, and the second question is “What split of equities and government bonds is required to approximate corporate bonds?”

Let’s take those questions in order.

If you were a UK investor investing in a portfolio 60% in global stocks and 40% in UK corporate bonds, you could’ve replicated that exposure and achieved (almost) exactly the same returns by splitting that 40% allocation to corporate bonds in half, and allocating 20% to equities, and 20% to government bonds:

Replicating corporate bonds - UKSource: Bloomberg.

This is obviously specific to the UK, and specific to this time period.

It’s slightly different in the US, although the idea is still the same – we can recreate corporate bond returns using a mix of equities and government bonds.

In the US, as we’ve already seen, we have monthly data going back to 1972, and which shows a 60/40 stock/corporate bond portfolio can be replicated by moving the whole corporate bond element into treasuries of a similar duration.

So corporates and treasuries acted exactly the same:

Replicating corporate bonds - USSource: Bloomberg.

This chart is for intermediate-term bonds (10-years), but the findings are similar for long-term bonds in the US, which you can recreate yourself using the excellent (but US-centric) Portfolio Visualizer:

Replicating corporate bonds - US (long term)Source: Portfolio Visualizer

So it’s clearly possible to recreate corporate bond returns using nothing more than a combination of stocks and government bonds.

The tricky part is that it’s impossible to say in advance what that split of stocks/government bonds should look like in order to replicate corporate bond returns.

If you’re either thinking about removing a corporate bond fund, or considering adding one to your portfolio, you’ll need to make a decision about the relative weights you should be adding/removing from the equity/government bond allocations to keep risk and returns constant.

The long-term data from the US shows corporates can be replaced by 100% government bonds. So if we think that’s the most reliable data for extrapolating into the future then it’s an easy decision – just stick whatever you thought you wanted to allocate into corporates and put it into government bonds.

Shorter-term data for UK/global hedged corporate bonds shows a 60/40 stock corporate bond portfolio could be replicated with an 80/20 stock/government bond portfolio. So in this situation the corporate bonds should be split 50% into stocks, 50% into government bonds.

As a general rule, you can make a strategic guess at the split of corporates based on the credit quality of the bond fund you’re buying. For the examples we’ve been looking at in this post, the credit quality of the corporate bond funds have been medium-high quality. If the corporate bonds you’re buying have a rating of AA/AAA, then the fund is going to behave more like a government bond fund. If it’s further down the credit spectrum and is veering into high-yield territory, then it’s going to behave more like equities.

But an irritating complication is that how corporate bonds behave relative to stocks/government bonds can vary dramatically over time. This is why it’s so hard to create a rule for translating corporate bond exposure at different credit ratings into stock/government bond mixes. When markets are rising, sometimes corporate bonds outperform stocks, sometimes they underperform government bonds.

In fact, when I checked the monthly data for the US market going back to 1972, when looking at the rolling 5-year returns for corporate bonds vs treasuries vs stocks, corporate bond returns were only between those of equities and government bonds in 57% of 5-year periods. That means 43% of the time their performance wasn’t between stocks and treasury bonds, which isn’t what we’d expect at all. In fact, most of the time when they were outside this treasury-stock band, they were underperforming both.

Here’s what their performance looked like during different market environments:

Performance in different market environments tableSource: Bloomberg. n refers to the number of rolling 5-year periods.

When markets were rising (which was the vast majority of the time – 478 out of 532 rolling 5-year periods), corporate bonds underperformed treasury bonds more often than not (62% of the time) – which isn’t what’s supposed to happen. Also weirdly, they managed to outperform stocks 20% of the time.

In a falling market, they fell less than stocks 100% of the time, which we’d expect, but only fell more than treasuries about 50% of the time – which is good, but we’d expect corporates to fall more than treasuries most of the time when markets are falling.

So their behaviour isn’t predictable – especially over the shorter term.

The last thing you want from your diversifier is to provide surprises. You get enough of that from the volatility of equities. So the unpredictability of corporate bonds is another cross against them in my book.

Because they don’t behave in line with expectations all the time, the best investors are able to do is make a strategic guess on how to split their allocation, based on the credit quality of the corporate bonds, and accept that sometimes the corporate bond holding would’ve performed better, and sometimes the equity/government bond split would’ve performed better.

It’s worth remembering that as long as the estimated split is ballpark correct, it’s unlikely to make a big difference to long-term returns anyway. Stocks are still going to be the overwhelming force affecting the portfolio’s risks and returns.

To make some form of informed opinion as to how to recreate corporate bond exposure, our starting point was that in the US we could’ve replicated corporate bond returns with 100% government bonds, and in the UK/global-hedged, it was more like a 50/50 split between equities and government bonds.

Given results were consistent in the US for both periods 1972-date and 1929-date (as seen in the ‘Third party evidence’ section), and were also consistent for both intermediate-duration and long-duration bonds, I’d be more inclined to view these results as being the more convincing. The UK/global data not only covers a much shorter time period, but there are two reasons for why it likely makes corporate bond returns look more similar to equities than might be expected over the longer-term:

  • Firstly, as we’ve seen already, the post-1998 data doesn’t include any returns from before the bond bull market started in 1980s. This boost for corporate bonds will make their returns look more similar to equities, and skew the allocation required to replicate their returns in favour of equities.
  • Secondly, the time period starts with the 2000-2010 ‘lost decade’ for stocks, which included both the dot-com crash and 2008. These lower equity returns for the period will again make the strong corporate bond returns for the period appear more like equity returns, skewing the allocation required to replicate corporate bond returns in favour of equities.

So for the kind of corporate bond funds we’ve been looking at in this article (medium-high credit quality), if I had a gun to my head and had to replicate corporate bond returns – either because I already a corporate bond fund and wanted to sell it, or because I liked the idea of corporate bonds and wanted to include them in my portfolio – then I’d be going for a roughly 80/20 government bonds/equities.

The exact split to replicate corporate bond returns might not be 80/20 going forwards – in fact, I’ll guarantee it won’t be. It’ll go through periods where it significantly underperforms corporate bonds, and periods where is significantly outperforms. But it’s consistent with the idea that over the long run corporate bonds behave more like government bonds than a 50/50 split between government bonds and equities, and the data which indicated a more equal split was biased upwards for bond returns and downwards for equity returns.

If I was considering allocating to corporates, that’s the ratio I’d be taking from my equities/gilts, and if I was considering removing them from my portfolio, that’s how I’d be reallocating the proceeds.

It’s still a bit finger-in-the-air, but as we’ve already seen, the exact makeup of the bond section of your portfolio doesn’t matter hugely anyway. When combined with equities, the risk of your portfolio is going to be dominated by the equity portion, so as long as you have a sensible allocation to higher credit quality bonds to help dampen drawdowns, then your return profile is likely to look similar anyway.

 

Summary

 

Before I get to the conclusion, let me sum up.

  • When comparing the use of corporate bonds with government bonds in a 60/40 portfolio, the results on portfolio risk, return, and drawdown characteristics were all very similar. When comparing the use of UK corporate bonds to UK government bonds in the 60/40, we saw corporates provided slightly higher returns, but slightly higher risk, resulting in equal risk-adjusted returns. For drawdowns, most of the time owning government bonds would’ve slightly reduced the drawdowns – with the exception of in 2008 when government bonds significantly reduced portfolio losses.
  • These findings were exactly the same when looking at global bonds hedged to sterling.
  • In the US, which covered a much longer time period, there was no difference in returns between the 60/40 corporate bond vs government bond portfolios. This meant the slightly higher risk from corporates resulted in slightly higher risk-adjusted returns for the government bonds. When looking at crash protection, government bonds came out on top, beating corporate bonds in 8 of the last 10 crashes, reducing overall portfolio drawdowns by an average of 8%.
  • The academic evidence reviewed showed, firstly, that the higher yields displayed on corporate bond factsheets are far less likely to be delivered to investors, due to downgrades, defaults, and bonds being called.
  • Secondly, it showed that US corporate bonds could’ve been replaced with government bonds in portfolios, which would’ve resulted in identical levels of risk and return.
  • Finally, it showed the reason higher returns on corporate bonds didn’t show up in the portfolio returns was because long-term government bonds mix better with the risks of stocks. Their annual correlation of returns to equities was lower than for long-term corporate bonds. The similarity of returns between using corporates vs government bonds at a portfolio level was due to the structurally lower correlation of government bonds when combined with equities, which equalled the higher potential return offered by corporate bonds.
  • Looking at some additional considerations in the government vs corporate decision, we looked at both costs and behavioural risks.
  • On the cost side of things, the fund expense ratios were almost identical when comparing government bond funds vs corporate bond funds. However, investors should still be cognisant of the increased trading costs associated with adding an extra fund into their portfolio.
  • On the behavioural risks, adding an extra fund into your portfolio increases the chances of performance chasing, which is especially dangerous when it affects a core holding in your portfolio.
  • We then turned to the question of how to replicate a corporate bond holding using a mix of equities and government bonds. We saw that in the UK, global markets, and the US, it was possible to replicate corporate bond returns using a combination of equities and government bonds. But the exact split necessary to mimic their returns is unknowable in advance, and varies over time. To further complicate matters, corporate bond returns were only between equities and government bonds roughly half the time, meaning they didn’t always perform in a predictable way. My finger-in-the-air approximation was an 80/20 split between government bonds/equities, based on the longer-term relative performances of the two types of bonds.

 

Conclusion

 

I’ve directed my work for this post towards the areas I think UK investors are most likely to find relevant to them. As with all things in the world of investing, there’s always more work to do. This isn’t meant to be a fully comprehensive assessment of corporate bonds – just some targeted analysis to help guide your decision-making. I’d love to see some more analysis studying other time periods, other markets, other durations, and other credit qualities. If there are any fellow bond nerds out there who know of any such studies, please let me know.

But based on what I’ve reviewed here, I personally wouldn’t include corporate bonds in my own portfolio. And that’s for the following reasons:

  1. For most portfolios (those with regular equity allocations between 20-80%), the portfolio’s risk will be dominated by stocks anyway, and the exact composition of the bond element will be relatively unimportant. Adding corporate bonds adds clutter and complexity, for not all that much in return.
  2. The last thing you want from your diversifier is to provide surprises. You get enough of that from the volatility of equities. The fact that corporate bonds don’t always behave as you’d expect them to means I’m less inclined to own them.
  3. Government bonds have a good track record of providing better protection during crashes, which is what we need from our bonds, and (in my opinion) the main reason for holding them.
  4. Over the longer-term, government bonds provided returns similar to corporate bonds thanks to their lower correlations to equities. This was true for both intermediate and long-term bonds, but the long-term bond results were especially similar. So if you’re holding longer duration bonds (perhaps as part of a duration matching strategy) then there’s even less benefit to holding corporates.
  5. The yields advertised on corporate bond funds’ factsheets are less likely to live up to expectations, thanks to bond downgrades, defaults, and callable bonds all reducing returns. Government bonds provide more predictable returns, which is useful for retirement planning.
  6. Owning an extra fund brings increased transaction costs, as well as added behavioural risks.
  7. Lastly, I try to avoid making judgements based on current market conditions (as that’s called active management…), but given we’re currently in a low interest rate environment, it seems likely the return premium for corporates over governments isn’t going to be as high as it has been in the past.

Despite all that, I still don’t hate them.

For those investing in an aggregate bond fund (one which owns both government and corporate bonds), I’d prefer to own a pure government bond fund, but the aggregate funds (especially those from Vanguard/iShares) are still a good choice. They tend to own medium-high quality corporate bonds, which, when combined with the significant government bond exposure, still results in a high credit quality bond fund, which is all we really need from our bonds (see Point 1 from the list above).

But for me, simplicity is key. Hence the name of the blog. Ultimately, the best portfolio is the one you can stick with for 60 years, and I find keeping portfolios simple is the best way to ensure there are no nasty surprises along the way.

Are corporate bonds likely to derail your portfolio? Absolutely not. As we’ve seen, the exact makeup of your bond allocation probably matters less than you think.

But do I find them compelling enough to include in my own portfolio?

Thanks, but it’s a no from me.

Share on Facebook
Share on Twitter
Share on LinkedIn

Past performance does not guarantee future performance and the value of investments can fall as well as rise. The information on this site is provided for information only and does not constitute, and should not be construed as, investment advice or a recommendation to buy, sell, or otherwise transact in any investment including any products or services or an invitation, offer or solicitation to engage in any investment activity. Please refer to the full disclaimer on the disclaimer page.

Subscribe
Notify of

4 Comments
Newest
Oldest Most Voted
Inline Feedbacks
View all comments
Charles
July 6, 2022 5:17 am

Really enjoyed reading this. Thanks Occam. I was surprised at the stats in the tables in similarity of returns between corporate bonds and treasury bonds because I was genuinely expecting much more distinct differentiation.

Last edited 1 month ago by Charles
John Williams
April 7, 2022 7:19 pm

I completely understand, and even agree with what you are saying about being able to replicate the past performance of corporate bonds by using a mix of government bonds and equities. However, in the name of diversification, unless there is some penalty in terms of reduced performance, increased risk or increased costs, surely diversification across 3 asset classes rather than just 2 makes most sense in terms of risk spreading. I note that many of the sources quoted compare a 60/40 split between equities / government bonds with a 60/40 split between equities / corporate bonds. A better way of analysing the risks / benefits would be to compare the former with, say, a 50/25/25 split between equities / corporate bonds / government bonds (or some other ratio), accepting that the corporate bonds have a risk profile that is likely to be part way between the two other classes. Then, overall, the risk profile of the 3 way split may be less than the 2 way equity / government bond split because of the diversification (and even if not demonstrably true in the past, it could be in the future – past performance does not predict etc etc). By way of example, if equities in low risk, low return sector A (eg miners & other defensive stocks) and equities high risk, high return sector B (IT and other growth & speculative stocks) had an average mid risk, mid reward total over the past several years, I doubt many people would say one should invest solely in sector A and sector B, and avoid all of the sector C mid-risk mid-reward equities (eg financials, etc etc). Nor would one say to just invest in the mid risk mid reward equities – most would want a spread across all 3 sectors. The only reason I would not diversify if risk and reward can be kept the same would be if it added to costs, and with current tracker funds I am not convinced that adding a corporate bond fund to a portfolio in order to add an entirely new asset class could be considered a poor investment. Whilst I agree that trying to keep a portfolio simple is a good idea, it should not be at the expense of a diversified portfolio.
PS. Love the blog – have learned a lot from it, and always read it. After crypto (which I believe you are covering next), how about some other asset classes eg. property trusts / REITs, infrastructure trusts, and commodities other than gold.

Andrew
April 7, 2022 11:46 am

Super clear and interesting, as ever. Thank you!