Bonds funds vs rolling bond ladders: How much difference is there?

[StillGoing]

Summary
In terms of returns, the short answer to “How much difference is there?” is, if fees and taxes are ignored, not much.

(edited the following summary text on 8 September 2024)
The slightly longer answer is it depends on whether the yield curve is inverted or not, changes in yield and whether new money is being added or withdrawals are being taken with detailed comparisons being given in follow up posts.

However, for the example of a fund or ladder with a maturity range of 0 to 5 years I’ve used, the difference in annualised returns is no more than about 7 bp and often less.

So, based on the limited tests reported here, this means that the choice as to whether a fund or a ladder is used for fixed income is more a choice than one solidly supported by evidence of outperformance one way or another.

Method
Now for some details

This analysis was prompted by some recent discussion about bond fund and rolling ladders in viewtopic.php?t=437867. Further reading can also be found on the wiki at
https://www.bogleheads.org/wiki/Rolling ... bond_funds and reference therein.

I’ve simulated the returns for a rolling bond ladder and bond fund (operational details below) where the highest maturity is 5 years and the lowest maturity is 0 years. In constructing the simulation I’ve made a number of assumptions.

Bonds
1) 5 years bonds are issued every 6 months (or ‘semester’ to borrow longinvest’s terminology in the thread at viewtopic.php?p=2717165) with a coupon equal to the yield to maturity and, therefore, the issue price is at par. The total number of bonds issued is always the same.
2) No other bonds are issued (i.e., there is never more than one bond available at a given maturity).
3) Coupons are paid every 6 months
4) The yield curve is defined as yield to maturity at 6 month intervals
5) The yield curve prior to the construction of the ladder or fund has been fixed for at least 5 years (which means that, initially, coupons are the same at all maturities).
6) Fractional bonds can be held

Rolling ladder
1) New money (if any), maturing bonds, and coupons are always (re)invested in the highest maturity bond (5 years in the example here).
2) Withdrawals (if any) are always taken from the maturing bond and coupons.

Bond fund
1) New money, (if any), maturing bonds, and coupons are distributed across all maturities such the fund always holds an equal number of bonds at each maturity.
2) Withdrawals (if any), are taken such that the fund is left with an equal number of bonds at each maturity.

To avoid overlong individual posts, I’ll post different scenarios separately. We’ll start with a relatively simple case in the next post and compare the outcomes for rolling ladder and fund.

cheers
StillGoing

[StillGoing]

Scenario 1: Constant flat yield curve, no withdrawals or new money

Rolling ladder

In the following table, the initial ladder (Semester 0), and the outcomes for the first two semesters (i.e., after 6 months and 1 year) are shown for the rolling bond ladder (in practice, I’ve run the simulator for 14 semesters – slightly longer than the maximum maturity - but that would be a very long table). For each time (in semesters), the following parameters are shown for each maturity, Yield to maturity (YTM), Coupon, Price, Modified duration (ModDur), and the number of bonds held (initially Nbonds) before reinvesting coupons and the maturing bond (Nbefore) and after (Nafter). In addition, the overall value (NAV), weighted modified duration (Wdur), and the value of coupons issued (TotCpns) are also shown. It should be noted that values are held at full precision in the spreadsheet, but shown to a limited number of decimal places here.

Code: Select all

		Maturity (Semester)													
Time	Param	0	1	2	3	4	5	6	7	8	9	10	NAV	Wdur	TotCpns
0	YTM	3.0	3.0	3.0	3.0	3.0	3.0	3.0	3.0	3.0	3.0	3.0			
	Coupon	3.0	3.0	3.0	3.0	3.0	3.0	3.0	3.0	3.0	3.0	3.0			
	Price	100.00	100.00	100.00	100.00	100.00	100.00	100.00	100.00	100.00	100.00	100.00			
	ModDur	0.00	0.49	0.98	1.44	1.93	2.36	2.85	3.25	3.74	4.12	4.61			
	Nbonds		10.00	10.00	10.00	10.00	10.00	10.00	10.00	10.00	10.00	10.00	10000	2.58	
1	YTM	3.0	3.0	3.0	3.0	3.0	3.0	3.0	3.0	3.0	3.0	3.0			
	Coupon	3.0	3.0	3.0	3.0	3.0	3.0	3.0	3.0	3.0	3.0	3.0			
	Price	100.00	100.00	100.00	100.00	100.00	100.00	100.00	100.00	100.00	100.00	100.00			
	ModDur	0.00	0.49	0.98	1.44	1.93	2.36	2.85	3.25	3.74	4.12	4.61			
	Nbefore	10.00	10.00	10.00	10.00	10.00	10.00	10.00	10.00	10.00	10.00	0			150.00
	Nafter	0.00	10.00	10.00	10.00	10.00	10.00	10.00	10.00	10.00	10.00	11.50	10150	2.61	
2	YTM	3.0	3.0	3.0	3.0	3.0	3.0	3.0	3.0	3.0	3.0	3.0			
	Coupon	3.0	3.0	3.0	3.0	3.0	3.0	3.0	3.0	3.0	3.0	3.0			
	Price	100.00	100.00	100.00	100.00	100.00	100.00	100.00	100.00	100.00	100.00	100.00			
	ModDur	0.00	0.49	0.98	1.44	1.93	2.36	2.85	3.25	3.74	4.12	4.61			
	Nbefore	10.00	10.00	10.00	10.00	10.00	10.00	10.00	10.00	10.00	11.50	0			152.25
	Nafter	0.00	10.00	10.00	10.00	10.00	10.00	10.00	10.00	10.00	11.50	11.52	10302	2.63	

In setting up the initial ladder it has been assumed that the yields have been fixed at 3% for at least 5 years, so the coupon for the bond at each maturity is 3%. It has also been assumed that the initial $10k has been equally distributed across all maturities (i.e., $1k for each of the 10 maturities) and therefore Nbonds is 10 in each each case.

After one semester has elapsed (i.e. 6 months), before reinvestment all the bonds have aged by 6 months (so reduced in maturity by one semester) and each has issued a semi-annual coupon (there are 100 bonds, a semi-annual coupon is $1.50 for each bond, so coupons total $150). Therefore, the maturing bonds (value $1k), plus coupons are now used to buy a total of 11.5 5-year bonds. This means that there are now more bonds at higher maturities and the weighted duration has increased.

After the second semester, the same procedure is followed, the maturing bond and coupons are used to buy 5-year bonds (more bonds are held, so the value of the coupon has increased) and so on for as many semesters as desired.

Bond fund

The results table for the bond fund is of the same format as for the rolling ladder.

Code: Select all

Time	Param	0	1	2	3	4	5	6	7	8	9	10	NAV	Wdur	TotCpns
0	YTM	3.0	3.0	3.0	3.0	3.0	3.0	3.0	3.0	3.0	3.0	3.0			
	Coupon	3.0	3.0	3.0	3.0	3.0	3.0	3.0	3.0	3.0	3.0	3.0			
	Price	100.00	100.00	100.00	100.00	100.00	100.00	100.00	100.00	100.00	100.00	100.00			
	ModDur	0.00	0.49	0.98	1.44	1.93	2.36	2.85	3.25	3.74	4.12	4.61			
	Nbonds		10.00	10.00	10.00	10.00	10.00	10.00	10.00	10.00	10.00	10.00	10000	2.58	
1	YTM	3.0	3.0	3.0	3.0	3.0	3.0	3.0	3.0	3.0	3.0	3.0			
	Coupon	3.0	3.0	3.0	3.0	3.0	3.0	3.0	3.0	3.0	3.0	3.0			
	Price	100.00	100.00	100.00	100.00	100.00	100.00	100.00	100.00	100.00	100.00	100.00			
	ModDur	0.00	0.49	0.98	1.44	1.93	2.36	2.85	3.25	3.74	4.12	4.61			
	Nbefore	10.00	10.00	10.00	10.00	10.00	10.00	10.00	10.00	10.00	10.00	0	10000		150.00
	Nafter	0.00	10.15	10.15	10.15	10.15	10.15	10.15	10.15	10.15	10.15	10.15	10150	2.58	
2	YTM	3.0	3.0	3.0	3.0	3.0	3.0	3.0	3.0	3.0	3.0	3.0			
	Coupon	3.0	3.0	3.0	3.0	3.0	3.0	3.0	3.0	3.0	3.0	3.0			
	Price	100.00	100.00	100.00	100.00	100.00	100.00	100.00	100.00	100.00	100.00	100.00			
	ModDur	0.00	0.49	0.98	1.44	1.93	2.36	2.85	3.25	3.74	4.12	4.61			
	Nbefore	10.15	10.15	10.15	10.15	10.15	10.15	10.15	10.15	10.15	10.15	0	10150		152.25
	Nafter	0.00	10.30	10.30	10.30	10.30	10.30	10.30	10.30	10.30	10.30	10.30	10302	2.58	

A comparison of the results for fund and ladder shows that the NAV after 1 and 2 semesters is the same in each case (and will continue to be the case provided the yields remain unchanged). However, there are two differences between the fund and the ladder. Firstly, by design, the number of bonds at each maturity is always the same for the fund, but this is not the case for the rolling ladder. Secondly, as a consequence, while the weighted duration is a constant 2.58 with time for the fund, for the ladder it gradually increases from 2.58 to 2.63 over the first two semesters. Although not shown, for the ladder the duration peaks at 2.66 after 2.5 years and then falls back to about 2.58 after 5 years.

This is a very simple example to illustrate the method used. In this case, the annualised return is easy enough to calculate from the semi-annual coupon, i.e., =1.015^2=3.022%, which is the same as the result from the simulation. While there are some simplifying assumptions made in the simulation, it should be good enough to compare the outcomes for the more complex scenarios given in the following posts.

cheers
StillGoing

[StillGoing]

Scenario 2: Constant, increasing yield curve, no withdrawals or new money

For this example, I’ve chosen a rising yield curve such that y(s)=0.2s+1, where y(s) is the yield for a maturity of s (in semesters), i.e., a bond with a maturity of one semester (6 months) would have a yield to maturity of 1.2%, 1 year (2 semesters), 1.4%, etc. (the full curve can be seen in the results tables below). The yield curve is the same for all time.

Rolling ladder

Again, the various parameters in the following table are the same as in the first example. Results for the first 3 semesters are shown.

Code: Select all

		Maturity (Semester)													
Time	Param	0	1	2	3	4	5	6	7	8	9	10	NAV	Wdur	TotCpns
0	YTM	1.0	1.2	1.4	1.6	1.8	2.0	2.2	2.4	2.6	2.8	3.0			
	Coupon	3.0	3.0	3.0	3.0	3.0	3.0	3.0	3.0	3.0	3.0	3.0			
	Price	100.00	100.90	101.58	102.07	102.35	102.43	102.31	102.00	101.51	100.84	100.00			
	ModDur	0.00	0.50	0.99	1.45	1.94	2.37	2.86	3.26	3.75	4.13	4.61			
	Nbonds		9.91	9.84	9.80	9.77	9.76	9.77	9.80	9.85	9.92	10.00	10000	2.59	
1	YTM	1.0	1.2	1.4	1.6	1.8	2.0	2.2	2.4	2.6	2.8	3.0			
	Coupon	3.0	3.0	3.0	3.0	3.0	3.0	3.0	3.0	3.0	3.0	3.0			
	Price	100.00	100.90	101.58	102.07	102.35	102.43	102.31	102.00	101.51	100.84	100.00			
	ModDur	0.00	0.50	0.99	1.45	1.94	2.37	2.86	3.26	3.75	4.13	4.61			
	Nbefore	9.91	9.84	9.80	9.77	9.76	9.77	9.80	9.85	9.92	10.00	0			147.65
	Nafter	0.00	9.84	9.80	9.77	9.76	9.77	9.80	9.85	9.92	10.00	11.39	10148	2.62	
2	YTM	1.0	1.2	1.4	1.6	1.8	2.0	2.2	2.4	2.6	2.8	3.0			
	Coupon	3.0	3.0	3.0	3.0	3.0	3.0	3.0	3.0	3.0	3.0	3.0			
	Price	100.00	100.90	101.58	102.07	102.35	102.43	102.31	102.00	101.51	100.84	100.00			
	ModDur	0.00	0.50	0.99	1.45	1.94	2.37	2.86	3.26	3.75	4.13	4.61			
	Nbefore	9.84	9.80	9.77	9.76	9.77	9.80	9.85	9.92	10.00	11.39	0			149.86
	Nafter	0.00	9.80	9.77	9.76	9.77	9.80	9.85	9.92	10.00	11.39	11.34	10299	2.65	

As before, I’ve made the assumption that the initial ladder has been constructed by allocating the same amount of money ($1k) to each of the 10 maturities. With a rising yield curve, the price at each maturity is now different and therefore the number of bonds bought is also different.

As before, after a semester (6 months) has elapsed, the maturing bond (with a price of 100) and coupons from all the bonds are used to purchase a bond with a maturity of 5 years. Since this bond is freshly issued, the price is also 100. After another semester, the same procedure is followed and so on.

The annualised return is just under 3%, so lower than with the flat yield curve used in Scenario 1. The weighted duration is slightly higher because the modified duration of the individual bonds is also slightly higher.

Bond fund

The results for the first 3 semesters are shown in the following table

Code: Select all

		Maturity (Semester)													
Time	Param	0	1	2	3	4	5	6	7	8	9	10	NAV	Wdur	TotCpns
0	YTM	1.0	1.2	1.4	1.6	1.8	2.0	2.2	2.4	2.6	2.8	3.0			
	Coupon	3.0	3.0	3.0	3.0	3.0	3.0	3.0	3.0	3.0	3.0	3.0			
	Price	100.00	100.90	101.58	102.07	102.35	102.43	102.31	102.00	101.51	100.84	100.00			
	ModDur	0.00	0.50	0.99	1.45	1.94	2.37	2.86	3.26	3.75	4.13	4.61			
	Nbonds		9.84	9.84	9.84	9.84	9.84	9.84	9.84	9.84	9.84	9.84	10000	2.59	
1	YTM	1.0	1.2	1.4	1.6	1.8	2.0	2.2	2.4	2.6	2.8	3.0			
	Coupon	3.0	3.0	3.0	3.0	3.0	3.0	3.0	3.0	3.0	3.0	3.0			
	Price	100.00	100.90	101.58	102.07	102.35	102.43	102.31	102.00	101.51	100.84	100.00			
	ModDur	0.00	0.50	0.99	1.45	1.94	2.37	2.86	3.26	3.75	4.13	4.61			
	Nbefore	9.84	9.84	9.84	9.84	9.84	9.84	9.84	9.84	9.84	9.84	0	10000		147.64
	Nafter	0.00	9.99	9.99	9.99	9.99	9.99	9.99	9.99	9.99	9.99	9.99	10148	2.59	
2	YTM	1.0	1.2	1.4	1.6	1.8	2.0	2.2	2.4	2.6	2.8	3.0			
	Coupon	3.0	3.0	3.0	3.0	3.0	3.0	3.0	3.0	3.0	3.0	3.0			
	Price	100.00	100.90	101.58	102.07	102.35	102.43	102.31	102.00	101.51	100.84	100.00			
	ModDur	0.00	0.50	0.99	1.45	1.94	2.37	2.86	3.26	3.75	4.13	4.61			
	Nbefore	9.99	9.99	9.99	9.99	9.99	9.99	9.99	9.99	9.99	9.99	0	10148		149.82
	Nafter	0.00	10.14	10.14	10.14	10.14	10.14	10.14	10.14	10.14	10.14	10.14	10297	2.59	

A comparison between this table and the one for the rolling ladder illustrates that there is a difference in the return after 1 year (2.99% for the ladder and 2.97% for the fund, although this does get bigger with time – see table below) because the ladder is investing coupons and maturing bonds in the highest maturity rather than across all maturities. For the bond fund, since a fixed number of bonds are held at each semester, given the rising yield curve and higher prices for the intermediate maturities, it can be seen that the overall number of bonds bought is lower than for the rolling ladder. Consequently, the value of the coupons after 1 and 2 semesters is also lower and, hence the NAV is lower than for the ladder

Comparison between rolling ladder and bond fund

In the following table, the NAV as a function of time for the rolling ladder and bond fund is shown together with the annualised return and the difference in annualised returns.

Code: Select all

	0	1	2	3	4	5	6	7	8	9	10	11	12	13	14	AR (%)	Diff (bp)
Rolling	10000	10148	10299	10454	10611	10772	10934	11099	11266	11435	11605	11777	11953	12132	12315	3.02	
Fund	10000	10148	10297	10449	10604	10760	10919	11080	11244	11410	11578	11749	11923	12099	12277	2.97	4.4

After 7 years (14 semesters), the annualised returns were (to two decimal places) 3.02% and 2.97% for the rolling ladder and the bond fund, respectively. The annualised return for rolling ladder was 4.4 bp higher than that of the bond fund.

cheers
StillGoing

[StillGoing]

Summary of different yield curves (constant in time) and no withdrawals or new money

The annualised returns (after 7 years), with a constant yield curve and no withdrawals or new money for rolling ladder or bond fund are as follows (Ann Ret is the annualised return, and Diff Ret is the difference in return between the rolling ladder and bond fund)

Code: Select all

				Ann Ret (%)	Diff Ret (bp)
Scenario			Rolling	Fund	Roll/Fund
Flat yield curve (3%)		3.02	3.02	 0.0
Non-inverted (y=0.2s+1)		3.02	2.97	 4.4
Inverted (y=3-0.2s)		1.01	1.02	-1.3

The bond fund and rolling ladder have identical returns when the yield curve is flat while the ladder performs better with a non-inverted yield curve and worse with an inverted yield curve.

Next, I’ll post some results where the yield curve is no longer fixed in time.

cheers
StillGoing

[Florida Orange]

Your analysis seems to show that, since most of the time the yield curve is not inverted, ladders are generally "better". But is the advantage to funds when the curve is inverted so great that it offsets what would otherwise be the advantage of ladders? If not, is your conclusion that, most of the time for most investors, ladders are better?

[StillGoing]

Your analysis seems to show that, since most of the time the yield curve is not inverted, ladders are generally "better". But is the advantage to funds when the curve is inverted so great that it offsets what would otherwise be the advantage of ladders? If not, is your conclusion that, most of the time for most investors, ladders are better?

On the basis of the results presented so far, ladders would be better most of the time. Including the effects of fees, would make them that bit better again. I'm not knowledgeable enough about the US tax system to comment on any difference there.

However, I've still got some more analysis to present (it is not quite ready for posting), but note that there are some further 'it depends' arising from whether you are contributing to or withdrawing funds from the ladder/fund because that can change the outcomes (depending on the yield curve). As it turns out, the problem is more interesting than I thought it would be!

cheers
StillGoing

[Parkinglotracer]

Bond ladders are going to have a constant decreasing duration which in times of interest decreasing will hurt performance a small amount but of course exact same durations should produce same results mathematically.

Over all I use a 5 year treasury bond ladder - a rung matures and I reinvest it, or rebalance it to equities, or spend it. So easy a caveman can do it. No psychological concern about interest rate changes effects on my bond portfolio.

[Tom_T]

Bond ladders are going to have a constant decreasing duration which in times of interest decreasing will hurt performance a small amount but of course exact same durations should produce same results mathematically.

Over all I use a 5 year treasury bond ladder - a rung matures and I reinvest it, or rebalance it to equities, or spend it. So easy a caveman can do it. No psychological concern about interest rate changes effects on my bond portfolio.

Please stop saying this. It's not easy, and requires you (or a spouse) to make a decision every single time. "Should I spend it? No, there's no need for the money right now" (until six months later, there is an unexpected need, and now you have to sell before maturity.)

You know what is so simple that a caveman could do it? Owning a bond fund and never making any decisions on what to do with it.

[sycamore]

Flat yield curve
No change in yield: Fund and ladder are identical
Step decrease in yield: Ladder performs better
Step decrease in yield: Fund performs better

There are two "Step decrease in yield" cases above. I assume the first is supposed to be "Step increase in yield"?

[Blues]

Please stop saying this. It's not easy, and requires you (or a spouse) to make a decision every single time. "Should I spend it? No, there's no need for the money right now" (until six months later, there is an unexpected need, and now you have to sell before maturity.)

You know what is so simple that a caveman could do it? Owning a bond fund and never making any decisions on what to do with it.

Well said, Tom.

Last month, the last of my individual Treasuries matured and am now only using funds (and the TSP) to invest in Treasuries going forward.
Not so much in anticipation of potential cognitive decline in the years to come, but mostly to make things as easy as possible for the missus who has zero interest or desire to manage a portfolio.

Now, with very little effort on her part, (and essentially only changing the registrations as applicable), she can leave the investments as they are set.

And if I'm honest, I enjoy the simplicity myself...especially since we have reached the point where, as the good Dr. Bernstein opines, we no longer have to play the game and can choose to take less risk. (And place most investing decisions, to a greater or lesser degree, on autopilot.)

[Parkinglotracer]

Bond ladders are going to have a constant decreasing duration which in times of interest decreasing will hurt performance a small amount but of course exact same durations should produce same results mathematically.

Over all I use a 5 year treasury bond ladder - a rung matures and I reinvest it, or rebalance it to equities, or spend it. So easy a caveman can do it. No psychological concern about interest rate changes effects on my bond portfolio.

Please stop saying this. It's not easy, and requires you (or a spouse) to make a decision every single time. "Should I spend it? No, there's no need for the money right now" (until six months later, there is an unexpected need, and now you have to sell before maturity.)

You know what is so simple that a caveman could do it? Owning a bond fund and never making any decisions on what to do with it.

Didn’t mean to strike a nerve. I think it’s a funny commercial and cavemen or cavewoman are about the only group it seems to be ok to make fun of these days. No fair telling me I can’t use it then using it!!! lol. In a bond fund the manager of the fund is buying and selling all the time so a decision to sell in a ladder before maturity isn’t the end of the world. We unexpectedly needed some serious money (more than our planned annual spending) and made the decision to sell 4 treasury bonds early to buy a house recently. Luck of the draw rates had gone down since we bought them and they had gained a lot over what we paid for them. Three clicks and the money was in our taxable brokerage mm account.

We keep about a year or two of annual spending in our money market account to prevent having to break the ladder often. For folks that aren’t comfortable buying treasuries ( Kevin had a great tutorial here ) one can do the same even easier process with CDs. I have been burnt with CDs at Marcus with large early termination fees. Maybe I should say understanding what is going on with a bond or CD ladder is so easy a xxxxx can do it vs trying to understand how a bond fund with a 6 or 7 yr duration can go down 13% in a year. I guess it’s pretty simple; rates can go up 2%.

Can’t say I’ll never say it’s so easy a xxxx can do it but I will keep its use to a minimum now that I know how it can be received; peace on this great Labor Day.

[SevenBridgesRoad]

I give you permission to keep using the phrase.

I interpreted Tom T's reaction not so much to the caveman phrase, but to you using it to say rolling bond ladders are so very easy in comparison to a bond fund.

[Tom_T]

I give you permission to keep using the phrase.

I interpreted Tom T's reaction not so much to the caveman phrase, but to you using it to say rolling bond ladders are so very easy in comparison to a bond fund.

I probably needed a smiley face in there. I think bond ladders are fine, but I don't think my wife, if she had to do it, would appreciate it.

The "so easy" remark depends on the person. My electrician actually changed an outlet without turning off the power; he said "I've done this a thousand times." Me, I'm trying to make sure I don't electrocute myself.

[jeffyscott]

Please stop saying this. It's not easy, and requires you (or a spouse) to make a decision every single time. "Should I spend it? No, there's no need for the money right now" (until six months later, there is an unexpected need, and now you have to sell before maturity.)

You know what is so simple that a caveman could do it? Owning a bond fund and never making any decisions on what to do with it.

Well said, Tom.

Last month, the last of my individual Treasuries matured and am now only using funds (and the TSP) to invest in Treasuries going forward.
Not so much in anticipation of potential cognitive decline in the years to come, but mostly to make things as easy as possible for the missus who has zero interest or desire to manage a portfolio.

Now, with very little effort on her part, (and essentially only changing the registrations as applicable), she can leave the investments as they are set.

And if I'm honest, I enjoy the simplicity myself...especially since we have reached the point where, as the good Dr. Bernstein opines, we no longer have to play the game and can choose to take less risk. (And place most investing decisions, to a greater or lesser degree, on autopilot.)

I'm also leaning toward migrating to only using mutual funds for the same reasons. My use of individual fixed income securities has mostly been as a way that I could add a little value, not because of a fear of bond funds. So buying CDs because I could get a better yield than similar treasuries, buying my own T-bills to save the money market ER (at Schwab, where it's 0.34%), etc.

While rolling 5 year T-bills may be relatively simple, buying a short term treasury mutual fund that holds 1-5 year treasuries is even simpler. I'll be curious to see the comparisons incorporating additions/withdrawals. I have said many times that I would guess that there would be little difference between what amounts to selling a small piece of each bond holding each year over a long time period vs. using only the one that matures in a ladder to cover spending.

The differences (about 1-4 bp) in the first batch of comparisons are insignificant, IMO. It's what I would expect, but may also be because it's looking at only up to 5 year maturities.

[50/50]

Please stop saying this. It's not easy, and requires you (or a spouse) to make a decision every single time. "Should I spend it? No, there's no need for the money right now" (until six months later, there is an unexpected need, and now you have to sell before maturity.)

You know what is so simple that a caveman could do it? Owning a bond fund and never making any decisions on what to do with it.

Well said, Tom.

Last month, the last of my individual Treasuries matured and am now only using funds (and the TSP) to invest in Treasuries going forward.
Not so much in anticipation of potential cognitive decline in the years to come, but mostly to make things as easy as possible for the missus who has zero interest or desire to manage a portfolio.

Now, with very little effort on her part, (and essentially only changing the registrations as applicable), she can leave the investments as they are set.

And if I'm honest, I enjoy the simplicity myself...especially since we have reached the point where, as the good Dr. Bernstein opines, we no longer have to play the game and can choose to take less risk. (And place most investing decisions, to a greater or lesser degree, on autopilot.)

I'm basically on the same path as you but I still have some Treasuries and broker CDs to work through. I have been getting my DW more involved with our finances which she always has left to me. She does not like shopping for treasuries or broker CDs. She's okay with local CDs, broker MMFs and treasury bond funds so we are directing our fixed income in that direction. She has been handling all the bill payments for more than a year, so we are good to go on that front. She still has a way to go on taxes, but she is making good progress (manhandling the computer tax program inputs causes wringing of the hands, I'll include myself in that one). Overall, I'm much more confident in her ability to carry on by herself if necessary.

[rkhusky]

Seems like the result is that a bond fund and a rolling bond ladder are about the same in regards to return. Choosing between the two is just a matter of personal preference.

[WoodSpinner]

Bond ladders are going to have a constant decreasing duration which in times of interest decreasing will hurt performance a small amount but of course exact same durations should produce same results mathematically.

Over all I use a 5 year treasury bond ladder - a rung matures and I reinvest it, or rebalance it to equities, or spend it. So easy a caveman can do it. No psychological concern about interest rate changes effects on my bond portfolio.

Please stop saying this. It's not easy, and requires you (or a spouse) to make a decision every single time. "Should I spend it? No, there's no need for the money right now" (until six months later, there is an unexpected need, and now you have to sell before maturity.)

You know what is so simple that a caveman could do it? Owning a bond fund and never making any decisions on what to do with it.

Many brokerages support an auto roll of a Treasury ladder without any fees. It’s pretty simple, nothing to do after the initial setup which is only seconds more difficult than investing in a bond fund.

From my perspective, I would expect a Bond Fund to have a slightl edge over time since their transaction costs are lower plus they can respond to shifts in the Yield curve more effectively .

FWIW, I hold a bond fund as well as a collapsing bond ladder and each serves a very different purpose.

WoodSpinner

[Florida Orange]

No psychological concern about interest rate changes effects on my bond portfolio.

None for me either, and I use funds not ladders. The NAV and yield on my bond funds varies slightly from year to year which I watch with a serene and Zen-like detachment. So easy a bivalve mollusk could do it. I'm glad you found what works for you. Happy Labor Day!

[Svensk Anga]

Bond fund
1) New money, (if any), maturing bonds, and coupons are distributed across all maturities such the fund always holds an equal number of bonds at each maturity.
2) Withdrawals (if any), are taken such that the fund is left with an equal number of bonds at each maturity.

cheers
StillGoing

It appears you assume the bond fund holds its bonds to maturity, but that is not the normal case. I think most sell bonds when they get to the lower end of the fund's maturity range. If there is a normal, positively sloped yield curve, the fund will earn a roll yield that the ladder held to maturity will miss out on. Ladder holders might also set a low maturity limit, but then they incur trading costs.

Is the roll yield enough to compensate for missing out on the last year's coupons on say a 5-year Treasury? What if the curve steepens or flattens over the period under consideration? Then, to make the analysis fair, one has to match durations. A 1-5 year fund is longer duration than a five year ladder held to maturity.

This is complicated enough to make funds attractive. I say that with most of my bonds in a non-rolling TIPS ladder. Non-rolling (with reliable real return) works for my own ever-shortening duration on this planet.

[jeffyscott]

Many brokerages support an auto roll of a Treasury ladder without any fees. It’s pretty simple, nothing to do after the initial setup which is only seconds more difficult than investing in a bond fund.

Many , can you name some? I know of one, Fidelity.

Schwab offers this only for T-bills (I think 6 mo. or shorter) and leaves you in their low yield settlement fund for a week in between each roll. My understanding is that Vanguard does not offer this at all.

[StillGoing]

Flat yield curve
No change in yield: Fund and ladder are identical
Step decrease in yield: Ladder performs better
Step decrease in yield: Fund performs better

There are two "Step decrease in yield" cases above. I assume the first is supposed to be "Step increase in yield"?

Well spotted, thank you. The second is supposed to be 'increase' (I've corrected the original)

[jeffyscott]

The OP inspired me to think of a comparison scenario that I thought was simple enough for me to do without making any errors (hopefully).

I got the yields for 1-10 year treasuries on Dec. 31 2013 from the treasury yield curve with interpolation for 4, 6, 8, and 9 year. I put $10K in each of these and assumed that I could have gotten a zero coupon at the same yield. I compared this to putting $100K in Vanguard Intermediate Treasury Index fund (VSIGX). I got the annual returns for that from Morningstar and assumed that the withdrawals from the fund each year were exactly equal to the individual bond proceeds.

Before doing the spreadsheet, I thought this was a scenario where most factors favored the ladder: each maturing rung was exactly the amount needed, the term of the fund became inappropriately long during the later years, and there was a large increase in interest rates in the final years. The only thing that seemed to favor the fund was it would not start with anything in the low yield 1 & 2 year maturities. While the ladder won, as expected, I was surprised at how little difference there was between the two alternatives.

The end result was that the fund was short by only about $950 in the final year. There was about $13,500 available from the maturing 10 year and abut $12,500 available from the fund. In order for the fund to survive for the full 10 years, I only had to reduce the annual draws by about 0.8% (so to 99.2% of what was available from the individual bonds).

I don't know how to paste legibly from a google sheet, but here's the input data:

yield for 1-10 year
0.13%
0.38%
0.78%
1.27%
1.75%
2.10%
2.45%
2.65%
2.84%
3.04%

VSIGX return (2014-2023)
4.22%
1.65%
1.10%
1.63%
1.30%
6.26%
7.65%
-2.60%
-10.69%
4.39%

[StillGoing]

Bond fund
1) New money, (if any), maturing bonds, and coupons are distributed across all maturities such the fund always holds an equal number of bonds at each maturity.
2) Withdrawals (if any), are taken such that the fund is left with an equal number of bonds at each maturity.

cheers
StillGoing

It appears you assume the bond fund holds its bonds to maturity, but that is not the normal case. I think most sell bonds when they get to the lower end of the fund's maturity range. If there is a normal, positively sloped yield curve, the fund will earn a roll yield that the ladder held to maturity will miss out on. Ladder holders might also set a low maturity limit, but then they incur trading costs.

Is the roll yield enough to compensate for missing out on the last year's coupons on say a 5-year Treasury? What if the curve steepens or flattens over the period under consideration? Then, to make the analysis fair, one has to match durations. A 1-5 year fund is longer duration than a five year ladder held to maturity.

This is complicated enough to make funds attractive. I say that with most of my bonds in a non-rolling TIPS ladder. Non-rolling (with reliable real return) works for my own ever-shortening duration on this planet.

Yes, this is what I've assumed so far. There are some indices that do hold until maturity (e.g., some FTSE ones) and, as you say, many that don't.

For the rising yield curve I've used so far (i.e., yield of 1.2% and 1.4% at semesters 1 and 2), with a 3% coupon, the price is 101.58 at 1 year, while a year's worth of coupons is 3, so having a lower limit of 1 year would reduce the the amount of money per bond. Obviously, a steeper gradient (and higher coupon) would increase the benefit of having a lower limit.

cheers
StillGoing

[WoodSpinner]

Many brokerages support an auto roll of a Treasury ladder without any fees. It’s pretty simple, nothing to do after the initial setup which is only seconds more difficult than investing in a bond fund.

Many , can you name some? I know of one, Fidelity.

Schwab offers this only for T-bills (I think 6 mo. or shorter) and leaves you in their low yield settlement fund for a week in between each roll. My understanding is that Vanguard does not offer this at all.

Fidelity is my main brokerage and I assume (perhaps incorrectly) others have similar capabilities.

[Doc]

While rolling 5 year T-bills may be relatively simple, ...

No rolling 5 year T-Bills is very, very difficult since T-Bills mature in at most 52 WEEKS.

That joke aside, what is often missing in these "rolling" discussions is that there may also be some advantage in taxes since selling the bottom lot will have some of the gain in LTCG not ordinary income assuming your ladder is in a taxable account.

[gavinsiu]

Thanks for the analysis. Just wondering if return includes reinvestment. The article implies that it does. In practice, I end up dumping the interest into mm because it was not enough for another bond.

Autoroll on fidelity also does not reinvest the coupon. The money ends up in settlement.

[1moreyr]

I do both, my bond fund is a 6 year duration and i have a bond ladder going out 1-3 years with I bonds filling the gap of year4/ 5 if necessary. The bond ladder reflects my annual spending budget.

i prefer the bond ladder as changing interest rates will not make my treasury less at maturity (like a bond fund can fluctuate) and I have enough money laid out for 3-4 years that I need. The bond fund is there to put items out 6 y ears on an autopilot. yes i know it's a rolling 6 years vs the ladder but i haven't figured out a better way to cover that.

I am retired and I SWAN knowing i can weather a storm with the appropriate amount of money for about the length of time of a typical down market.

[BirdFood]

Thanks for the analysis. Just wondering if return includes reinvestment. The article implies that it does. In practice, I end up dumping the interest into mm because it was not enough for another bond.

That’s the one place I could see myself buying a bit of a bond fund—as a way to put bits of bond interest in bonds. I’d probably go with some defined maturity fund. So far I haven’t bothered because the MM currently pays so much that I’m fine with bits and pieces sitting there waiting for the next round number to pile up.

[gammalaser]

I am guessing it is implied in the OP posts that the comparisons are relative to Treasuries and not other types of fixed income like corporates, munis, etc. For these other types of bonds, I think the arguments skew strongly in favor of funds in order to achieve diversification across issuers and ratings and to provide better liquidity.

[CloseEnough]

Is this a new issue here? I don't think I've seen it discussed before.

My take, it probably doesn't matter all that much, meaning there is likely not all that much of a difference. Me, being more of a Neanderthal and not even having caveman status, use bond funds, both nominal and TIPS fund (btw, have TIPS been discussed here?). They are so easy a Neanderthal can do it.

Likely what matters more is your overall asset allocation, and other issues.

Happy Labor Day to all!

[Parkinglotracer]

I give you permission to keep using the phrase.

I interpreted Tom T's reaction not so much to the caveman phrase, but to you using it to say rolling bond ladders are so very easy in comparison to a bond fund.

I probably needed a smiley face in there. I think bond ladders are fine, but I don't think my wife, if she had to do it, would appreciate it.

The "so easy" remark depends on the person. My electrician actually changed an outlet without turning off the power; he said "I've done this a thousand times." Me, I'm trying to make sure I don't electrocute myself.

I don’t think my wife will continue my ladder after I am gone either … but she knows how! More likely she will have the Edward Jones guy who has been eyeing her during pickleball manage her portfolio. lol.

[cjcerny]

I think Tom T. makes a good point. A smart investor can definitely get a bond ladder going, but what happens to that ladder when the smart investor passes and their not so interested spouse takes the reins. That spouse is likely going to reach out for help with it, and there are lots of sharks in the water. All my investment decision making is centered on making things as simple as possible for those that get what’s left when I pass so that the odds of it ending up with a shark are as low as possible.

[gavinsiu]

That’s the one place I could see myself buying a bit of a bond fund—as a way to put bits of bond interest in bonds. I’d probably go with some defined maturity fund. So far I haven’t bothered because the MM currently pays so much that I’m fine with bits and pieces sitting there waiting for the next round number to pile up.

This is the part that I find messy with a rolling bond ladders. I setup a reoccurring bond ladder in Fidelity, but despite being zero bonds, only the original amount invested gets autoroll and the rest goes into the settlement. As you pointed out, dumping them inot a bond fund would work.

[gavinsiu]

Here's a question,

If I have a bond fund, can I sell it and repurchase it as a bond ladder of the same duration without incurring a loss?

[sycamore]

Here's a question,

If I have a bond fund, can I sell it and repurchase it as a bond ladder of the same duration without incurring a loss?

You might have a loss, a gain, or basically break even.

But first you need to define "same duration". Do you mean a single bond with same duration as the fund? Or two bonds? Or bond per yearly rung? Or... you get the picture.

And then as a small investor you won't get as good a price as the bond fund managers.

[gavinsiu]

You might have a loss, a gain, or basically break even.

But first you need to define "same duration". Do you mean a single bond with same duration as the fund? Or two bonds? Or bond per yearly rung? Or... you get the picture.

And then as a small investor you won't get as good a price as the bond fund managers.

This comes from the comment that bond fund is basically a rolling bond ladder of the same duration. If that is true, could I sell all of my bond funds, and then use the money to contruct an equivalent duration bond ladder. As you say, I imagine there will be a cost to the transations.

[typical.investor]

Here's a question,

If I have a bond fund, can I sell it and repurchase it as a bond ladder of the same duration without incurring a loss?

Of course you can (minus transaction costs).

Another way to exit would be to buy a treasury zero whose maturity matches the duration of the fund.

If you tried to create a ladder, you’d either have learn to calculate duration or use zeros whose maturity and duration are the same.

If you think about it, you should actually be able to withdraw from a bond fund that has suffered a NAV loss by taking cash and buying treasury zeros (with a maturity LONGER than the fund duration) such that the weighted average of cash and longer bonds is the same as the fund duration WITHOUT realizing a NAV loss. If you kept doing so in a rising rate environment though, eventually you’d run out of bonds long enough to raise your weighted duration.

[ClaycordJCA]

I removed a somewhat contentious exchange. Please remember that if you think a post violates forum policy, report it to the moderators rather than engaging in the thread:

This is a moderated forum. We expect this forum to be a place where people can feel comfortable asking questions and where debates and discussions are conducted in civil tones. Discussions are about issues, not people. If you disagree with an idea, go ahead and marshal all your forces against it. But do not confuse ideas with the person posting them. At all times we must conduct ourselves in a respectful manner to other posters. Attacks on individuals, insults, name calling, trolling, baiting or other attempts to sow dissension are not acceptable. If you feel that someone has attacked you or otherwise violated the policies of this forum, do not respond in kind. Instead, please click the report button on the offending post. This is the quickest method to notify the site moderators.

[typical.investor]

Here's a question,

If I have a bond fund, can I sell it and repurchase it as a bond ladder of the same duration without incurring a loss?

Of course you can (minus transaction costs).

Another way to exit would be to buy a treasury zero whose maturity matches the duration of the fund.

If you tried to create a ladder, you’d either have learn to calculate duration or use zeros whose maturity and duration are the same.

If you think about it, you should actually be able to withdraw from a bond fund that has suffered a NAV loss by taking cash and buying treasury zeros (with a maturity LONGER than the fund duration) such that the weighted average of duration of the cash (duration zero) and longer bonds is the same as the fund duration WITHOUT realizing a NAV loss. If you kept doing so in a rising rate environment though, eventually you’d run out of bonds long enough to raise your weighted duration.

[StillGoing]

Is this a new issue here? I don't think I've seen it discussed before.

My take, it probably doesn't matter all that much, meaning there is likely not all that much of a difference. Me, being more of a Neanderthal and not even having caveman status, use bond funds, both nominal and TIPS fund (btw, have TIPS been discussed here?). They are so easy a Neanderthal can do it.

Likely what matters more is your overall asset allocation, and other issues.

Happy Labor Day to all!

There have been discussions about whether bond funds or DIY rolling ladders are 'best' since at least 2009 (the earliest thread I have seen - it is linked from the wiki page I linked in the first post). As far as I know, the difference has not been quantified before (although given the rather knowledgeable people on these boards I'd be surprised if not - although I couldn't find a thread).

The limited results I've presented support your take - there appears not to be much of a difference, although I have ignored fund fees, transaction costs for the ladder (if any), and taxes.

cheers
StillGoing

[StillGoing]

While rolling 5 year T-bills may be relatively simple, ...

No rolling 5 year T-Bills is very, very difficult since T-Bills mature in at most 52 WEEKS.

That joke aside, what is often missing in these "rolling" discussions is that there may also be some advantage in taxes since selling the bottom lot will have some of the gain in LTCG not ordinary income assuming your ladder is in a taxable account.

Since I am in the UK and have little detailed knowledge of the US tax system, can anyone quantify the effect of any differences in taxation between a rolling ladder and a bond fund. The limit of my understanding is that where it was held (i.e., taxable, tax deferred, etc.) would make a difference.

cheers
StillGoing

[Tom_T]

While rolling 5 year T-bills may be relatively simple, ...

No rolling 5 year T-Bills is very, very difficult since T-Bills mature in at most 52 WEEKS.

That joke aside, what is often missing in these "rolling" discussions is that there may also be some advantage in taxes since selling the bottom lot will have some of the gain in LTCG not ordinary income assuming your ladder is in a taxable account.

Since I am in the UK and have little detailed knowledge of the US tax system, can anyone quantify the effect of any differences in taxation between a rolling ladder and a bond fund. The limit of my understanding is that where it was held (i.e., taxable, tax deferred, etc.) would make a difference.

cheers
StillGoing

In tax deferred, it's simple: nothing is taxed until money is withdrawn, and then it will be taxed as regular income. There are no capital gains or losses.

[sycamore]

You might have a loss, a gain, or basically break even.

But first you need to define "same duration". Do you mean a single bond with same duration as the fund? Or two bonds? Or bond per yearly rung? Or... you get the picture.

And then as a small investor you won't get as good a price as the bond fund managers.

This comes from the comment that bond fund is basically a rolling bond ladder of the same duration. If that is true, could I sell all of my bond funds, and then use the money to contruct an equivalent duration bond ladder. As you say, I imagine there will be a cost to the transations.

I think we're dealing with a case where "in theory" and "in practice" will be different in non-trivial ways.

There's not just the transaction costs aspect. Let's say you have a simple bond fund that holds a set of 10 bonds. When you sell, you effectively sell all of the funds individual holdings instantaneously. But you can't (practically speaking) perform a set of 10 parallel purchases of equivalent bonds. The market will move as you go buy the individual bonds one by one.

Also if practical importance is how much money you buy in each bond. Individual bonds can only be traded in whole units, but with a bond fund you're effectively trading whole and fractional amounts. I think this means you don't end up with same dollar weighted duration, no?

[Tom_T]

You might have a loss, a gain, or basically break even.

But first you need to define "same duration". Do you mean a single bond with same duration as the fund? Or two bonds? Or bond per yearly rung? Or... you get the picture.

And then as a small investor you won't get as good a price as the bond fund managers.

This comes from the comment that bond fund is basically a rolling bond ladder of the same duration. If that is true, could I sell all of my bond funds, and then use the money to contruct an equivalent duration bond ladder. As you say, I imagine there will be a cost to the transations.

I think we're dealing with a case where "in theory" and "in practice" will be different in non-trivial ways.

There's not just the transaction costs aspect. Let's say you have a simple bond fund that holds a set of 10 bonds. When you sell, you effectively sell all of the funds individual holdings instantaneously. But you can't (practically speaking) perform a set of 10 parallel purchases of equivalent bonds. The market will move as you go buy the individual bonds one by one.

Also if practical importance is how much money you buy in each bond. Individual bonds can only be traded in whole units, but with a bond fund you're effectively trading whole and fractional amounts. I think this means you don't end up with same dollar weighted duration, no?

Also, another problem with the theory is that a real bond fund can contain thousands of bonds, with different purchase prices and yields. The theory never applies to real life.

[Parkinglotracer]

I think the difference is mainly psychological for me. In my mind the ladder is more transparent with what is happening with the value of each rung of my ladder as interest rates change and I play the mind game I am not going to sell until maturity so the fluctuations don’t matter. Buying a single bond helped me understand what was happening in my bond fund (when I used to own bond funds) on a grander scale.

[jeffyscott]

While rolling 5 year T-bills may be relatively simple, ...

No rolling 5 year T-Bills is very, very difficult since T-Bills mature in at most 52 WEEKS.

That joke aside, what is often missing in these "rolling" discussions is that there may also be some advantage in taxes since selling the bottom lot will have some of the gain in LTCG not ordinary income assuming your ladder is in a taxable account.

But I think that only applies if you are selling, rather than just allowing them to mature? I know you do (or have done) that, but what had been suggested here was to buy 5 year and then when one matures buy another 5 year.

The Vanguard intermediate treasury index fund would typically be buying 10 year and selling 3 year, so would presumably offer that added benefit. And I see that it did have some LTCG distributions in 2020 and 2021.

[Doc]

That joke aside, what is often missing in these "rolling" discussions is that there may also be some advantage in taxes since selling the bottom lot will have some of the gain in LTCG not ordinary income assuming your ladder is in a taxable account.

But I think that only applies if you are selling, rather than just allowing them to mature? I know you do (or have done) that, but what had been suggested here was to buy 5 year and then when one matures buy another 5 year.

The Vanguard intermediate treasury index fund would typically be buying 10 year and selling 3 year, so would presumably offer that added benefit. And I see that it did have some LTCG distributions in 2020 and 2021.

Re: Selling before maturity. Right, the tax advantage is limited to taxable accounts and selling before maturity as in your 3-10 example case.

To reiterate a previous point. The rolling ladder approach is only beneficial with a normal yield curve not the inverted yield curve we have currently.

[jeffyscott]

To reiterate a previous point. The rolling ladder approach is only beneficial with a normal yield curve not the inverted yield curve we have currently.

So that's presumably the reason that VSIGX did not distribute any LTCG in 2022 or 2023.

[StillGoing]

A step change in the yield curve, no withdrawals or new money

Up until now, the yield curve has been kept constant with time. In the next set of results a step change in yield is applied at the beginning of Semester 2. The single step is either +1% or -1% and is applied equally across all maturities (i.e., a parallel shift in yields) and then held for the remainder of the simulation.

Example case
For example, the following shows the NAV of the fund and the ladder as a function of time for a change in a flat yield curve from 3% to 2%.

Code: Select all

	Semester
	0	1	2	3	4	5	6	7	8	9	10	11	12	13	14	AR (%)	Diff (bp)
Rolling	10000	9950	10114	10015	9916	9815	9713	9610	9506	9401	9295	9188	9080	8971	8860	-1.71	
Fund	10000	9950	10115	10016	9917	9816	9714	9611	9507	9402	9296	9189	9081	8972	8862	-1.71	-0.2

There are two things to note. Firstly, as expected, the NAV rises after the decrease in yield and slightly more for the ladder (which has a slightly longer duration) than for the fund. Secondly, unlike where there was no change in yield, the performance is no longer the same – the fund performs very slightly better.

Summary of results

In the following table, the results for various yield curves and steps are summarised (no new money or withdrawals)

Code: Select all

		Ann ret (%)
Scenario	Rolling	Fund	Diff (bp)
Flat
no change (3%)	3.02	3.02	 0.0
3% to 2%	2.47	2.46	 0.4
3% to 4%	3.58	3.59	-0.4
Non-inverted (x+0.2s)
no change (x=1)	3.02	2.97	 4.4
x=1 to x=0	2.47	2.43	 4.2
x=1 to x=2	3.57	3.53	 4.6
Inverted (x-0.2s)
no change (x=3)	1.01	1.02	-1.3
x=3 to x=2	0.46	0.46	-0.5
x=3 to x=4	1.55	1.58	-2.2

For the flat and inverted yield curve, a drop in yields favours the ladder (i.e., the difference is increased) and vice versa for an increase in yields.

For the non-inverted yield curve, a drop in yields favours the fund (i.e., the difference is decreased) and vice versa for a decrease in yields.

It is worth noting that the difference in annualised returns between the ladder and fund is still relatively small, i.e. no more than 5 bp.

So far I have only dealt with the case where there have been neither withdrawals nor money added. In the next post, I'll present the results where withdrawals are taken.

cheers
StillGoing

[typical.investor]

That joke aside, what is often missing in these "rolling" discussions is that there may also be some advantage in taxes since selling the bottom lot will have some of the gain in LTCG not ordinary income assuming your ladder is in a taxable account.

Since I am in the UK and have little detailed knowledge of the US tax system, can anyone quantify the effect of any differences in taxation between a rolling ladder and a bond fund. The limit of my understanding is that where it was held (i.e., taxable, tax deferred, etc.) would make a difference.

cheers
StillGoing

I could try but ... help me understand why a rolling ladder would be more tax efficient than a fund.

Sure, you could realize LTCG by selling the bottom lot in your ladder, but doesn't a fund do that as well? Effectively speaking, aren't rolling ladders and funds both pretty much relatively constant duration and selling holding to maintain that?

If we look at BND, we see it issues both LTCG and STCG depending on the year.

Code: Select all

BND DISTRIBUTIONS:

YEAR	Ord	LTCG	STGC
2024	100%	0%	0%
2023	100%	0%	0%
2022	97%	3%	0%
2021	90%	10%	0%
2020	93%	5%	2%
2019	100%	0%	0%

Why would a rolling ladder give more opportunities to realize LTCG than BND? The only difference I see is that a ladder could probably avoid the STGC, but that makes no difference as it will be taxed at ordinary rates anyway.

In practice, I would guess a fund might realize more LTCGs than a rolling ladder that is only selling the lowest rung. Surely much of the 2020-2022 LTGCs were realized in longer durations as the fund adjusts its duration and holdings to reflect the index.

I mean as maturity approaches, the value reverts closer and closer to par so selling the bottom rung in a ladder would seem to naturally limit the opportunity to convert to LTCGs. And likely, most simply allow the bottom rung to mature before reinvestment.

In any case, if one simply let the bottom rung mature to reinvest in 2019-2024, then using BND would have had an average of around between 0.25% (22% bracket) and 0.75% (upper brackets) lower tax cost. Note, those in the bottom two brackets where LTCG are 0% would see an average of a 0.55% tax reduction.

So for a ladder to be more efficient, it would have to exceed those amounts.

That's my attempt at quantifying it anyway ...