tl,dr: Individual bonds, particularly US T-Bills, deserve serious consideration as a portion of a retiree’s portfolio, due to predictability/lowered volatility/higher possible floor SWR. More research is needed.
Fair Warning, long post ahead. This is due to the complexity of the topic. tl,dr is given at the top and bottom of the post.
None of the following is new or groundbreaking. I present basic characteristics of individual bonds vs. bond funds/bond ETFs (for brevity, I will now just refer to bond funds), then concentrate on an area where I think more research is warranted. That area is lower volatility of a portfolio containing individual bonds (when held to maturity and assuming no defaults) vs. a bond fund.
Bond fund characteristics:
• Diversification. Not important with USTs, increasing benefit with GSEs/agencies, municipal, and corporate. Also increasing benefit with extended durations and lower-quality bonds. A bond fund typically holds thousands of bonds. A default or two will not materially affect the fund.
• Low Costs. Low cost bond funds are readily available; for instance VWEAX expense ratio is 0.13%, VWESX is 0.21%, and VBIRX is 0.07%. Bond funds make purchases in large quantities, which lowers the bid/ask spread.
• Low correlation with stocks. Lowest for short-term USTs; increases with extended durations and lower quality.
• Liquidity. Easy to buy and sell with low transaction costs.
• Low Volatility. Much lower than stocks. Volatility increases with duration and lower quality.
Bond characteristics:
• Concentration. Individual bond holdings are typically have far fewer issues that a fund. A default can have significant impact.
• Higher transaction costs. Not with USTs and buying at primary auction Bid/ask spreads are relatively high. Frequent trading can significantly reduce returns.
• Lower correlation with stocks than bond funds. If held to maturity and no defaults occur.
• Lower liquidity. Most bonds trade relatively infrequently, especially at the quantities an individual investor buys. Not true for USTs.
• Predictability/Lower volatility than funds. Return of principal and interest payments are known, if held to maturity and no defaults occur. I’m going to stop repeating that caveat for the sake of brevity. Interest-on-interest is not predictable. Not as important if you’re a retiree and ‘eating’ the interest.
Here is the Bogleheads take on this issue: https://www.bogleheads.org/wiki/Individual_bonds_vs_a_bond_fund There are links to four Vanguard white papers at the end of the article. They talk about duration as the “point of indifference” for the owner of a bond fund in dealing with interest rate changes”, a well-known concept, and conclude that “you should always hold bond funds with a duration equal to or shorter than the expected need for your money”. This is valid for an investor in the accumulation phase, but in retirement that equation changes. Assuming a 60/40 portfolio, if stocks return <1.5x bonds then the need for money is immediate. They conclude that rolling bond ladder and funds are similar. If that’s true, what I think they are missing is the lower volatility of bonds. Everything else being equal, lower volatility is better. Of course, everything else is not equal, so the question is, does lower volatility compensate for lower expected returns for a shorter duration/higher quality individual bond portfolio? Also note that these two strategies are not mutually exclusive – a portion of the FI portfolio could be individual bonds (shorter duration/higher quality) and the remaining could be invested in funds. [This is a topic for a different post. When I get my thinking and math straight, I intend to start another thread.]
Here is a link to a free site with 1928-2022 US historical returns for various asset classes (unfortunately it doesn’t include inflation, but that information is readily available). https://pages.stern.nyu.edu/~adamodar/New_Home_Page/datafile/histretSP.html
Which brings us to SORR (Sequence of Returns Risk).
Michael Kitces has a couple articles on “Bond Tent/Rising Equity Glidepath” (two names for the same concept). Kitces is with Buckingham Strategic Wealth, among other things, as is Larry Swedroe. When those two talk personal finance, I listen carefully. I don’t always agree with them 100%, but if I don’t, I want to make darn sure I know *why* I disagree and that my thinking is on solid ground (always understanding that I’m not perfect and being willing to change given new information).
https://www.kitces.com/blog/managing-portfolio-size-effect-with-bond-tent-in-retirement-red-zone/
https://www.kitces.com/blog/managing-portfolio-size-effect-with-bond-tent-in-retirement-red-zone/
His basic point is that SORR is highest the decade before and after retirement, as portfolio returns are much more important than yearly contributions. Therefore, increasing bond exposure/decreasing equity exposure during that period mitigates the SORR when it matters most. He also found that using short-term T-Bills instead of 10-year T-Bonds had varying effects on the SWR - HOWEVER! – the result tended to be positive during the lowest SWR periods and negative during the highest SWR periods. Take a close look at the second graph in the second article. The left and right axes are not labeled, but they are the same as in the first graph. So for the risk-averse retiree who wants the highest possible SWR, this is a good strategy. In plain English, who cares if T-Bills drag down the portfolio’s performance if they increase the SWR?
I would really like to see some of the heavy financial hitters, such as Kitces and Swedroe, explore this idea further.
tl,dr: Individual bonds, particularly US T-Bills, deserve serious consideration as a portion of a retiree’s portfolio, due to their predictability/lowered volatility/higher possible floor SWR. More research is needed.
Fair Warning, long post ahead. This is due to the complexity of the topic. tl,dr is given at the top and bottom of the post.
None of the following is new or groundbreaking. I present basic characteristics of individual bonds vs. bond funds/bond ETFs (for brevity, I will now just refer to bond funds), then concentrate on an area where I think more research is warranted. That area is lower volatility of a portfolio containing individual bonds (when held to maturity and assuming no defaults) vs. a bond fund.
Bond fund characteristics:
• Diversification. Not important with USTs, increasing benefit with GSEs/agencies, municipal, and corporate. Also increasing benefit with extended durations and lower-quality bonds. A bond fund typically holds thousands of bonds. A default or two will not materially affect the fund.
• Low Costs. Low cost bond funds are readily available; for instance VWEAX expense ratio is 0.13%, VWESX is 0.21%, and VBIRX is 0.07%. Bond funds make purchases in large quantities, which lowers the bid/ask spread.
• Low correlation with stocks. Lowest for short-term USTs; increases with extended durations and lower quality.
• Liquidity. Easy to buy and sell with low transaction costs.
• Low Volatility. Much lower than stocks. Volatility increases with duration and lower quality.
Bond characteristics:
• Concentration. Individual bond holdings are typically have far fewer issues that a fund. A default can have significant impact.
• Higher transaction costs. Not with USTs and buying at primary auction Bid/ask spreads are relatively high. Frequent trading can significantly reduce returns.
• Lower correlation with stocks than bond funds. If held to maturity and no defaults occur.
• Lower liquidity. Most bonds trade relatively infrequently, especially at the quantities an individual investor buys. Not true for USTs.
• Predictability/Lower volatility than funds. Return of principal and interest payments are known, if held to maturity and no defaults occur. I’m going to stop repeating that caveat for the sake of brevity. Interest-on-interest is not predictable. Not as important if you’re a retiree and ‘eating’ the interest.
Here is the Bogleheads take on this issue: https://www.bogleheads.org/wiki/Individual_bonds_vs_a_bond_fund There are links to four Vanguard white papers at the end of the article. They talk about duration as the “point of indifference” for the owner of a bond fund in dealing with interest rate changes”, a well-known concept, and conclude that “you should always hold bond funds with a duration equal to or shorter than the expected need for your money”. This is valid for an investor in the accumulation phase, but in retirement that equation changes. Assuming a 60/40 portfolio, if stocks return <1.5x bonds then the need for money is immediate. They conclude that rolling bond ladder and funds are similar. If that’s true, what I think they are missing is the lower volatility of bonds. Everything else being equal, lower volatility is better. Of course, everything else is not equal, so the question is, does lower volatility compensate for lower expected returns for a shorter duration/higher quality individual bond portfolio? Also note that these two strategies are not mutually exclusive – a portion of the FI portfolio could be individual bonds (shorter duration/higher quality) and the remaining could be invested in funds. [This is a topic for a different post. When I get my thinking and math straight, I intend to start another thread.]
Here is a link to a free site with 1928-2022 US historical returns for various asset classes (unfortunately it doesn’t include inflation, but that information is readily available). https://pages.stern.nyu.edu/~adamodar/New_Home_Page/datafile/histretSP.html
Which brings us to SORR (Sequence of Returns Risk).
Michael Kitces has a couple articles on “Bond Tent/Rising Equity Glidepath” (two names for the same concept). Kitces is with Buckingham Strategic Wealth, among other things, as is Larry Swedroe. When those two talk personal finance, I listen carefully. I don’t always agree with them 100%, but if I don’t, I want to make darn sure I know *why* I disagree and that my thinking is on solid ground (always understanding that I’m not perfect and being willing to change given new information).
https://www.kitces.com/blog/managing-portfolio-size-effect-with-bond-tent-in-retirement-red-zone/
https://www.kitces.com/blog/managing-portfolio-size-effect-with-bond-tent-in-retirement-red-zone/
His basic point is that SORR is highest the decade before and after retirement, as portfolio returns are much more important than yearly contributions. Therefore, increasing bond exposure/decreasing equity exposure during that period mitigates the SORR when it matters most. He also found that using short-term T-Bills instead of 10-year T-Bonds had varying effects on the SWR - HOWEVER! – the result tended to be positive during the lowest SWR periods and negative during the highest SWR periods. Take a close look at the second graph in the second article. The left and right axes are not labeled, but they are the same as in the first graph. So for the risk-averse retiree who wants the highest possible SWR, this is a good strategy. In plain English, who cares if T-Bills drag down the portfolio’s performance if they increase the SWR?
I would really like to see some of the heavy financial hitters, such as Kitces and Swedroe, explore this idea further.
tl,dr: Individual bonds, particularly US T-Bills, deserve serious consideration as a portion of a retiree’s portfolio, due to their predictability/lowered volatility/higher possible floor SWR. More research is needed.
post. I think you’ll find many here are in agreement with this conclusion. Kitces does get some mention here, more than Swedroe, but both seem to be viewed favorably.
