My rethink of bond fund choice

[Lsbcal]

I just want to share this data in case it might help others think through their allocations. Also go ahead and critique this if you have objections.

Over the years I have moved all around in the bond fund category. This represents 10% of our portfolio with the rest of the fixed income in inflation indexed bonds of various maturities (iBonds and TIPS). Recently I looked at what amount of added equities I would add to a bond fund holding to make it beat just straight bond funds.

Since we only hold 10% bond funds this means just increasing equities by a small amount. The result indicates just increasing equities by 1% in our portfolio outperforms straight bond funds. In other words, if we have a 60/10/30 portfolio (60 equity, 20 bond funds, 30 inflation indexed and cash) then going to a 61/9/30 portfolio with the bonds in an intermediate Treasury fund work best Perhaps this is intuitive but the results over several decades are worth mentioning here.

Here are the results from July 1987 through October 2024 for various bond funds. These results are expressed in compound growth rates (CAGR). The holding are mostly for holding using Vanguard funds during this 37 year stretch.

Short term Treasury = 3.9%
Short term Investment grade = 4.8%
Intermediate investment grade = 5.5%
Total Bond Market = 5.0%
Intermediate Treasury = 5.1%
Intermediate Treasury 90% / Equities 10% = 5.7%

So not an earth shaking discovery but gives me peace of mind knowing that the bond fund part is in pretty safe US Treasuries.

 

[pb4uski]

You nailed it. BND is 68% U.S. government bonds.

 

[Lsbcal]

My choice is VGIT with 99.9% government. By adding the equity of 10% of the VGIT I have explicitly added equity risk. With BND it appears some equity like risk is in the BBB and bonds below AAA. Not that that is a bad choice but I don't know off hand how to calculate the equity like risk. I just want to be explicit about the equity risk.

 

[pb4uski]

It's not equity risk at all. It's credit default risk but the spread between investment grade corporate bonds and treasuries have been pretty narrow lately.

 

[audreyh1]

I think Lsbcal means that corporate bonds tend to suffer more when equities tank. And this is true - the spreads tend to widen, and funds holding lower credit quality tend to get hit harder. They may not have increased defaults but the bond market concerns over defaults increases.

 

[target2019]

My choice is VGIT with 99.9% government. By adding the equity of 10% of the VGIT I have explicitly added equity risk. With BND it appears some equity like risk is in the BBB and bonds below AAA. Not that that is a bad choice but I don't know off hand how to calculate the equity like risk. I just want to be explicit about the equity risk.

If I interpret this correctly. if I hold a percentage of BND total bond market, for the period you studied, the performance would be equivalent to 90% VGIT plus 10% equity fund (such as VTI).

How will you use VGIT in the future, meaning when will the funds be drawn upon? That context will help me, I think.

Thanks.

 

[VanWinkle]

It is true that if you are looking for more returns, I increase the stock exposure vs looking at more risky bonds. That is why junk bonds have never been on my menu as I can just raise equity exposure slightly to gain the same increased return.

 

[audreyh1]

It is true that if you are looking for more returns, I increase the stock exposure vs looking at more risky bonds. That is why junk bonds have never been on my menu as I can just raise equity exposure slightly to gain the same increased return.

Yep, I learned to avoid junk bonds long ago and that I might as well own equity instead.

I haven’t moved to treasury only fixed income index funds yet, but I can certainly see doing that as I get older.

 

[Lsbcal]

If I interpret this correctly. if I hold a percentage of BND total bond market, for the period you studied, the performance would be equivalent to 90% VGIT plus 10% equity fund (such as VTI).

How will you use VGIT in the future, meaning when will the funds be drawn upon? That context will help me, I think.

Thanks.

These are not quite equivalent since VGIT + 10% equity outperforms BND by 0.7% annually. Note that just the VGIT beat BND for the period studied.

BTW, VGIT is what I now use but the data for intermediate Treasury came from the fund VFIUX and instead of BND is used the fund VBTLX (equivalent I think).

The way I would draw upon the bond funds is unclear to me as these are all held in retirement accounts and I can move assets around without any penalty. I only have RMD's to deal with.

 

[Lsbcal]

It is true that if you are looking for more returns, I increase the stock exposure vs looking at more risky bonds. That is why junk bonds have never been on my menu as I can just raise equity exposure slightly to gain the same increased return.

Some years back on Bogleheads there was a big debate among two professionals (Swedroe and Ferri) about having a small exposure to junk bonds. It got heated and out of control as I remember. Egos ... we all have one


Not to be an alarmist but one should take into account worst case markets. I believe that during the Depression of the 1930's corporate bonds did not do well, especially long dated ones. This is another reason to keep the equity like risk separate from fixed income. BTW, I am not saying anything new as this has been noted by some respected investors.

 

[target2019]

These are not quite equivalent since VGIT + 10% equity outperforms BND by 0.7% annually. Note that just the VGIT beat BND for the period studied.

BTW, VGIT is what I now use but the data for intermediate Treasury came from the fund VFIUX and instead of BND is used the fund VBTLX (equivalent I think).

The way I would draw upon the bond funds is unclear to me as these are all held in retirement accounts and I can move assets around without any penalty. I only have RMD's to deal with.

Thanks. I'll have to rerun my understanding then.

As I reread everything I noticed a math error in your allocation statement. "(60 equity, 20 bond funds, 30 inflation indexed and cash)."

 

[Lsbcal]

Thanks. I'll have to rerun my understanding then.

As I reread everything I noticed a math error in your allocation statement. "(60 equity, 20 bond funds, 30 inflation indexed and cash)."

Oops yes, the sentence should read:

In other words, if we have a 60/10/30 portfolio (60 equity, 10 bond funds, 30 inflation indexed and cash) then going to a 61/9/30 portfolio with the bonds in an intermediate Treasury fund work best

 

[ERhoosier]

I am not a fan of bond funds (ETFs or mutual funds) for fixed income AA when capital preservation is a goal. Individual bonds are very easy to buy these days at any discount broker. Fund expenses erode returns, and should interest rates rise (a real risks as Federal deficit goes ever higher) the NAV of funds can decline significantly (~1% for each 1yr of fund's effective duration). My observation is these NAV declines are rarely fully recouped as rates later decline (even when including interest payments). Lesson I learned the hard way. I currently buying only short/intermediate IG individual bonds (mostly gov't/agencies now) with the intent to hold to maturity (predictable capital and income).
I would make an exception for few target date" bond funds that hold their individual issues to maturity, but again fund expenses tend to erode returns (for same portfolio of bonds).

 

[Lsbcal]

I looked at the Jun 2002 period to Nov 2024 since the 5 year Treasury rate was variable during this period but ended at the same current rate of about 4.05%. During this period rates rose as high as 5.1% and declined as low as 0.2%. The resulting compound rates of return were:

Intermediate Treasuries (VFIUX) = 3.5%
Intermediate Treasury 90& / Equities 10% = 4.3%

FWIW, I would probably not hold bonds should we get to super low rates as in 2020 - 2022. I sold off my bond funds at the beginning of 2022 and avoided the route. So I am not a set and forget it type.

 

[audreyh1]

During that period, does the model rebalance to maintain the 90/10 allocation?

 

[Lsbcal]

During that period, does the model rebalance to maintain the 90/10 allocation?

Yes, for computational sake I rebalance monthly.

P.S. I corrected my last post as the equities to keep it simple were just the SP500 here.

 

[audreyh1]

Do you see a big difference if you rebalance annually?

 

[Lsbcal]

Do you see a big difference if you rebalance annually?

Not easy for me to run that in my spreadsheet since all my data is monthly. However, I'd guess rebalancing annually would increase the return since equities have risen over long periods of time.

Personally nowadays I sell excess equities when they hit my AA limit. I'm up 1% above my limit and will probably sell some on Monday.

 

[audreyh1]

Rebalancing seems to benefit most when the two set classes have time to diverge.

However you were using a very short time period in your 2022 - 2024 analysis. The annual rebalancing would be for much longer periods.

 

[Montecfo]

These are not quite equivalent since VGIT + 10% equity outperforms BND by 0.7% annually. Note that just the VGIT beat BND for the period studied.

BTW, VGIT is what I now use but the data for intermediate Treasury came from the fund VFIUX and instead of BND is used the fund VBTLX (equivalent I think).

The way I would draw upon the bond funds is unclear to me as these are all held in retirement accounts and I can move assets around without any penalty. I only have RMD's to deal with.

Sure but aren't you liquidating something to do RMDs?

 

[Lsbcal]

Rebalancing seems to benefit most when the two set classes have time to diverge.

However you were using a very short time period in your 2022 - 2024 analysis. The annual rebalancing would be for much longer periods.

The analysis I just mentioned today was for 2002 - 2024. So 22 years.

 

[Lsbcal]

Sure but aren't you liquidating something to do RMDs?

I am not quite following this comment. The mechanics of RMD's don't really come into how I hold fixed income, For RMD's I just move some equity/fixed_income money to short term money in January for taking RMD's in December.

 

[audreyh1]

The analysis I just mentioned today was for 2002 - 2024. So 22 years.

Oops sorry- misread.

 

[Montecfo]

I am not quite following this comment. The mechanics of RMD's don't really come into how I hold fixed income, For RMD's I just move some equity/fixed_income money to short term money in January for taking RMD's in December.

That was my point. You aren't consuming it but you need annual liquidity to fund your RMDs.

Perhaps you are saying you liquidate in a way that preserves your allocation. If so that does answer the question. But in a down equity market would your approach be the same?

 

[Lsbcal]

In my case I have a ladder of TIPS and those will provide most of the RMD money. The Treasury bond fund is only about 10% of the portfolio.