I have been wrestling with the question of whether it is better for the retail investor to buy bonds directly or bond exchange-traded funds (ETFs). At this stage, I’m leaning toward bond ETFs as the best way to go for long-term horizons (e.g. retirement funds). Let me throw out some thoughts why. If anyone sees any flaws, let me know before I plunk some money down.
Let’s summarize what the proponents of direct ownership claim:
• individual bonds have maturity dates and offer certainty of returns if held to maturity, unlike bond ETFs
• commissions and management expense ratios (MERs) make bond ETFs more expensive than the commission paid to buy bonds over the counter from dealers.
OK, what to make of this. The first thing to get out of the way is how to do the comparison. Comparing a bond ETF to a single bond doesn’t fly with me. The bond ETF is diversified in terms of default and interest-rate risk, which is good. A single bond is not diversified on those criteria, which is not good. So the comparison, it seems to me, should be between the bond ETF and a diversified ladder of bonds (1).
Now, if you look at bond ladders in use for long-term investing, most seem to be the rolling kind – i.e. maturing bonds are rolled into new bonds. So, the cost of owning bonds directly is not just the commission paid when they are first bought but also the commissions paid on rolling them. The costs of direct ownership would then seem to be more in line with bond ETFs, especially considering over-the-counter commissions are high in percentage terms (particularly for strip bonds).
In addition, a rolling ladder has no fixed maturity date and is subject to the same uncertainty of returns as the bond ETF. It is only when the ladder matures into cash that it offers certainty of returns (2) (3). There is a good discussion of these points, and the equivalency of bond ETFs and rolling bond ladders on the Boglehead Wiki.
Other reasons to prefer bond ETFs:
Another consideration is that it is hard to construct a diversified bond ladder with less than $75,000 or so. Small investors don’t seem to have much of a choice other than bond ETFs. An exception might perhaps be if the retail investor only went with government bonds: since default risk is low, the ladder would not have to be too big.
Yet another consideration: the convenience of owning bond ETFs. From the excellent summary on the HowToInvestOnline blog, we have:
• it’s easier to rebalance a portfolio with bond ETFs
• reinvestment of interest payments is easier with bond funds (not a concern for ladders of strip bonds, though)
• running a bond ladder is more work than owning a bond fund, e.g. got to assess bond issuer’s creditworthiness (except in case of government bonds)
1) A bond ladder spreads a sum of money over bonds of different maturities. An example is $100,000 divided equally over 10 bonds maturing in 1, 2, 3, 4, 5, 6, 7, 8, 9 and 10 years.
2) When the time approaches for taking cash out of a bond ETF, certainty of returns can be had by selling the ETF in parts and putting the proceeds into bonds that mature on the dates needed.
3) Perhaps one exception where a bond ladder might be better than a bond ETF over the long run is for those investors saving for retirement through a non-rolling ladder of inflation-indexed bonds, as Professor Zvi Bodie advocates for conservative investors. In Canada, inflation-indexed bonds are known as real-return bonds and in the U.S., as Treasury Inflation Protected Securities (TIPS).