Why are bond ETFs typically recommended for diversifying into bonds?

Question

It seems that the conventional wisdom on portfolio allocation suggests that one should include a bond component in one's portfolio. The usual way that people gain exposure to bonds is by buying a bond ETF. The relatively conservative "Bogleheads" index investing community appears to subscribe this thinking; it advocates a "three-fund portfolio" where one of the funds is a total market bond ETF.

The problem is that these bond ETFs never mature. They experience perpetual interest rate risk. A long-term rise in interest rates would lead to a more or less permanent loss of principal. Contrast this to holding a regular bond to maturity. Assuming negligible credit risk, the payout of a regular bond held to maturity is known at the time it is bought. In contrast, the payout of the bond ETF is not known in advance; there is a very real risk of permanent loss of principal when one wishes to sell in X years.

From what I see, by buying a bond ETF, one is essentially betting on interest rates to ensure the preservation of principal. This looks highly speculative to me. I have no business in predicting interest rates.

So my question is: Why does "prudent financial advice" recommend a bond ETF allocation for the purpose of diversifying one's portfolio into bonds? Bond ETFs appear to be a highly speculative "investment" that involves betting on interest rates, with no strong guarantees that the principal would remain intact. Please enlighten me, as I have so far avoided regular bond ETFs in favor of instruments with stronger guarantees of capital preservation: individual investment-grade bonds held to maturity, and investment-grade "target-maturity" bond ETFs held to liquidation.

If the preservation of principal is important, I fail to see any use of bond ETFs. Instead, I would construct a "bond ladder" consisting of individual bonds and "target-maturity" bond ETFs, all held to maturity/liquidation.

Answer 1

You mention that a diversified bond ETF 'has no known maturity date' and thus 'perpetually takes on interest rate risk.' This is partially true.

You then say "This looks highly speculative to me." This is not true, at least, no differently than a single bond.

Consider a simple bond example:

You own a single bond, maturing in 1 year, earning 3% annual interest. Let's call purchase price in January $1000, so you would expect to receive $1,030 at the end of the year. You bought it for exactly its fair price. In June if you were going to sell it to someone, the fair price to do so would be $1,015 - your original price plus the $15 of accrued interest. Then, say global market interest rates double - an equivalent bond would now have interest rates of 6%. This doesn't affect your $1,030 to be received at the end of the year, but someone wouldn't want to pay you $1,015 for your bond, when they could buy a 6 month bond for $1,015 and end up with $1,045 at the end of the year. To earn your maturity amount of $1,030 in 6 months when market interest rates are 6%, in the open market a buyer would want to pay no more than $1,000 - effectively wiping out the accrued interest you've earned, because that accrued interest is perfectly offset by the decreased time value for the months of July-December, when better options are available in the market. Note - you could take that $1,000 offer from a buyer in June, and invest it at market rates of 6%, to still earn your expected $1,030 by the end of the year.

So even when buying a single bond, your market value rises or falls with market interest rates. Your capital is preserved in that holding onto the bond gives you the same nominal dollar value, but since interest rates have risen, that nominal value at the end of the year isn't worth what it once was.

Instead of buying a single bond, you could do a bond ladder - assume you buy 12 bonds, all with 1-year maturities, each month of the year. This would work exactly the same as above, but you would do the calculations individually every single month. What this would do, is basically take on the market rate of interest at each individual month, instead of 'putting all your eggs in one basket' and taking the full rate in January [there are other laddering benefits, like always having liquid cash available for emergencies, so you don't have much tied up]. On one end of the scale, you could buy a 20 year bond today that matures at an incredibly stable interest rate, and at the other end of the scale, you could have a bond ladder that puts, in this case, 1/12th of your capital in the market at this month's interest rate, every month forever.

A blended ETF works the same way, except that there will be maturities continuously occurring, meaning that every day, some bonds mature, dumping value into the ETF, which it then uses to buy new bonds under whatever relevant structure makes up the ETF, with new maturity dates, at current market rates. A diversified bond ETF with rolling maturities is essentially a gigantic bond ladder with maturities happening continuously.

If you look at a bond ETF and think 'hey, that value moves up or down, and I never really get my preserved capital out', you basically do - but like with the bond example about, the true present value of that preserved capital will fluctuate essentially based on the value of money itself - based on market interest rates.

The benefit of all this is that you have higher liquidity, because a diversified fund will be more broadly applicable than individual bonds, so the number of buyers goes up exponentially, whereas a single bond may only trade infrequently, meaning you might need to take a bit of a loss if you needed to liquidate.

Answer 2

Granularity
A single bond has quite a high face value. This may not matter for the institutional investor but makes it problematic for the average Joe who just may not have enough money invested to buy a single bond. This granularity also makes it very hard to diversify properly across multiple bonds and adding your monthly contributions. Buying units of a bond fund is so much more fitting to the needs and abilities of a retail investor.

Required management
When you buy a single bond you get lots of smaller coupon payments which you probably need to reinvest. This again hits the granularity issue pretty hard, how are you going to reinvest 200 €/$/GPB into individual bonds? A bond fund gets rid of all the management issues

Interest rate sensitivity
It is true that long term bonds will lose considerable value when interest rates rise sharply. Combined with low current interest, this gives a pretty low expected returns for long term bonds for the foreseeable future. However, this issue can be mitigated. There is no need to buy a total bond market fund wiht many 20+y bonds. There are plenty of bond funds with shorter and medium maturities which will be a lot less affected by rising interest rates.

Answer 3

I would construct a "bond ladder" consisting of individual bonds

Don't you think this is part of a bond ETFs strategy? Do you think they're holding extremely long-duration bonds indefinitely? What insight do you have that would make a bond ladder perform better in a rising-rate environment?

Holding a bond ETFs is NOT the same as holding a small set of individual bonds indefinitely. They buy and sell bonds just like equity ETFs buy and sell stocks, according to their investment strategy.

Bond ETFs have lower risk (in terms of variance of return) in general, and you can find ETFs with varying levels of risk/return from risk-free govt ETFs to high-yield speculative ETFs, and covering various markets (domestic, international, corporate, municipal, etc.) so there's plenty of options available to diversify an equity portfolio.

Yes, interest rates are relatively low and a rise in interest rates would reduce the value of bonds in general. But Bond ETF managers aren't idiots. They know what to do if rates do rise to reduce that impact and can "rebalance" their portfolio to take advantage of higher rates. Just like equity ETFs can adapt when the equity market falls.

Answer 4

The answer from "Grade 'Eh'" parries your arguments against bond ETFs effectively.

However, there is a major problem with bond ETFs and index funds that you did not mention: They are often cap-weighted, so they automatically have high exposure to the largest bond issuers. Often, the companies and governments that borrow the most become the worst credit risks.