Swedroe's guidelines to Bonds

[simplesauce]

I was just reading Larry Swedroe's bonds book. He summarizes some general guidelines to follow. I'm wondering how many Bogleheads follow some of these. Here are a few:

-Avoid any bonds longer than 10-years
-Avoid bonds with ratings under AA
-Favor Treasuries and TIPS
-Municpals are OK if using the higher qualities (AAA, AA). Must diversify.
-Corporates have issues, not recommended unless highest quality. Must diversify
-Say no to Mortgage Backed Securities
-If using high-yield, add to your equities allocation, not bonds. In general, the risk has not been rewarded.

[Tamarind]

My only bond holding is TBM so insofar as it follows the guidelines, I do.

I believe it adheres to most of these guidelines, except maybe the one on mortgage backed securities, as TBM contains a big lump of GNMAs (21%). Also, 15% long bonds, of which I think all are Treasuries.

[Call_Me_Op]

I generally follow these guidelines.

Larry (and I) believe that your portfolio should be dichotomized into risky and safe investments. Bonds belong in the safe category and as such, should be high-quality and short-medium duration.

Some people prefer the messy approach to portfolio construction, with a little of this and a little of that. I don't like this approach because it tends to obscure the risk level.

[jhfenton]

I generally follow those rules. I favor CDs and intermediate treasuries (VSIGX). I avoid anything but short-term corporates (VSCSX). My small muni allocation (VOHIX) is fairly high quality, but Vanguard does include a big chunk of A-rated munis. (With our small muni allocation, it's not worth paying to access one of the Baird funds that Mr. Swedroe likes.)

Since we're 88/12 equity/fixed income, I limit us to high quality bonds. No high yield. if I do ever venture into high-yield, it will be with a fallen angel fund, trying to take advantage of the downgrade factor. Other high yield strategies haven't shown compelling results.

[larryswedroe]

simple
Just to add that should use CDs instead of Treasuries when you have access to them, unless spread in yields very thin

Larry

[Doc]

I follow all of Larry's recommendations as well as I can. One of reasons I don't use TBM is because of the MBS but we do have small positions in mortgages. Some of our investment grade funds hav A and BBB corporates but the M* average is A. I also have one ten plus year TIPS position and I would consider more when/if real rates get into what I consider a reasonable range.

(I keep Larry's bond book on the end table next to my computer. Thanks neighbor. )

[dbr]

I would say I follow those recommendations exactly. I arrived there on my own. I haven't bothered to mess around with brokered CDs in my directed brokerage link in the 401K. I don't think that is even available in that plan.

[glock19]

simple
Just to add that should use CDs instead of Treasuries when you have access to them, unless spread in yields very thin

Larry

Are you saying if yield spread is thin on same maturity/duration then go treasuries? Please elaborate? thanks

[larryswedroe]

glock
Yes, Treasuries should have a premium as they are more liquid and benefit more in flights to quality/liquidity, like in 2008. So CDs need to have a higher yield to "compensate" you for the loss of that benefit. Most of time you can buy CDs at considerable benefit to Treasuries of same maturity. I've seen times when it's been close to 1% on longer term CDs. So as example if I had just a say 10bp premium for a CD I would go with Treasury (also should consider the cost of a fund if not buying Treasury direct). Each investor should decide for him/herself what the premium yield should be. Remember, most of the time you are going to earn that "free lunch" as there is no credit risk. But once in while, like in 08 it can make a signficant difference. So it's the sure thing of the higher yield vs. potential of a benefit in flight to liquidity.
Hope that helps
Larry

[sid hartha]

I don't think the Vanguard BND Eft meets Larry's guidelines and that's the one that I own. Is this concerning?

[Lou354]

^glock19^larryswedroe. Also, interest from treasury securities is exempt from state and local taxes. Interest from CDs isn't. So CDs should pay a higher interest rate to compensate for that.

[Artsdoctor]

Yep. I have learned a lot from Larry over many years and would follow the advice noted above. For me, however, there is an exception because I live in California. California general obligation bonds have an AA- rating and used to have an A rating. So if I were to follow Larry's guidance to the letter, I would have not invested in my state's bonds. However, I felt that I could ameliorate some of that risk by diversification and have used Vanguard's CA intermediate-term tax-exempt fund for many, many years. I realize that there is some risk with that approach but I have made my peace with it and have taken it knowingly.

[MikeMak27]

Larry

Piggybacking on Sid's question above, are etf's such as IUSB, AGG, GVI, and MUB appropriate risk for the reward they give you?

[Taylor Larimore]

I was just reading Larry Swedroe's bonds book. He summarizes some general guidelines to follow. I'm wondering how many Bogleheads follow some of these. Here are a few:

-Avoid any bonds longer than 10-years
-Avoid bonds with ratings under AA
-Favor Treasuries and TIPS
-Municpals are OK if using the higher qualities (AAA, AA). Must diversify.
-Corporates have issues, not recommended unless highest quality. Must diversify
-Say no to Mortgage Backed Securities
-If using high-yield, add to your equities allocation, not bonds. In general, the risk has not been rewarded.

simplesauce:

Larry is very knowledgeable about bonds and I agree with all but one of his "guidelines ("Say no to Mortgage Backed Securities").

The Mortgage Backed Securities (VFIIX) in Total Bond Market (21%) provide additional diversification, lower standard deviation (risk), and enjoyed slightly higher returns than TBM itself during the past 10- and 15 years.

Past performance does not forecast future performance.

Best wishes.
Taylor

[larryswedroe]

Lou, yes that is correct of course on tax exemption for Treasuries but for most should be holding CDs in tax advantaged accounts as preference, unless in low brackets.

Mike
I would not use ETFs for bonds for variety of reasons, and would limit holdings to what I've recommended, and would also avoid MBS as well as corporates so that excludes Total Bond funds.

Larry

[lack_ey]

I don't think the Vanguard BND Eft meets Larry's guidelines and that's the one that I own. Is this concerning?

Sure I guess? What's your threshold for concern? You can find plenty of experts who largely agree and others who disagree*, in any case, and differences between total bond and his recommendations are relatively small in impact in terms of actual returns and risk. Unless you're 80%+ invested in fixed income, you've probably got most of your risk and asset returns dominated by stocks anyway.

*for example, Jack Bogle complains about total bond not owning enough corporates, Burton Malkiel seems to be going off the deep end these days on dividend stock substitution and tilting corporates, Rick Ferri likes total bond itself and deliberately adds junk bonds to it (and TIPS), a large number of asset managers running balanced portfolios don't have an issue with longer dated bonds/corporates/MBS/etc., and so on. Some research is more positive on corporate bonds (and some of the other things), for example "The Credit Risk Premium" from Asvanunt and Richardson.

Larry

Piggybacking on Sid's question above, are etf's such as IUSB, AGG, GVI, and MUB appropriate risk for the reward they give you?

My name's not Larry and he already responded but

IUSB - iShares Core Total USD Bond Market ETF - basically total bond+ that includes junk too
AGG - iShares Core U.S. Aggregate Bond ETF - total bond
GVI - iShares Intermediate Government/Credit Bond ETF - 1-10 yr maturity corporate/govt bonds
MUB - iShares National Muni Bond ETF - broad market investment grade muni bonds

Based on the criteria listed, would not like the long-term bonds in IUSB or AGG, the MBS in both, most corporate bonds in both (especially the longer-term ones, including those with call options, and the high amount that are BBB and A), the junk bonds in IUSB. Could use CDs instead of the Treasuries.

With GVI you're basically getting the same thing as 2/3 Vanguard Short-Term Bond Index Fund (BSV), 1/3 Vanguard Intermediate Term Bond Index Fund (BSV) but paying a higher ER and spreads. Those Vanguard funds cover 1-5 and 5-10 respectively. In both cases again you're getting corporates, a lot in the lower investment grade, especially in BIV in perhaps longer than would like, etc. And losing out on CD return above Treasuries.

With MUB you're paying for a higher ER over Vanguard Tax-Exempt Bond Index Fund ETF (VTEB) but it's more liquid so if you're trading frequently or making frequent small transactions on a platform with commission free trades for it then I guess MUB is okay. The credit quality is lower than he would prefer (over 20% below AA) and this is a mid-to-long-term-ish index with perhaps more term risk than would be comfortable. Many of the bonds are also long maturity but callable, and thus have negative convexity. In ETF form, for munis I guess SPDR Nuveen Barclays Short Term Municipal Bond ETF (SHM) would be less bad from that perspective. In mutual fund land there's also Baird Quality Intermediate Muni Bond (BMBIX).

The separate concern about bond ETFs is covered elsewhere. IMHO a bit overblown for the ETFs sticking to high quality.

[sid hartha]

I don't think the Vanguard BND Eft meets Larry's guidelines and that's the one that I own. Is this concerning?

Sure I guess? What's your threshold for concern? You can find plenty of experts who largely agree and others who disagree*, in any case, and differences between total bond and his recommendations are relatively small in impact in terms of actual returns and risk. Unless you're 80%+ invested in fixed income, you've probably got most of your risk and asset returns dominated by stocks anyway.

*for example, Jack Bogle complains about total bond not owning enough corporates, Burton Malkiel seems to be going off the deep end these days on dividend stock substitution and tilting corporates, Rick Ferri likes total bond itself and deliberately adds junk bonds to it (and TIPS), a large number of asset managers running balanced portfolios don't have an issue with longer dated bonds/corporates/MBS/etc., and so on. Some research is more positive on corporate bonds (and some of the other things), for example "The Credit Risk Premium" from Asvanunt and Richardson.

I guess not a huge concern to me and no need to make any changes. You know what they say when experts disagree! Carry on sorry to derail.

[PaulF]

Well, I chose my bond holdings after reading Larry's book almost exactly 10 years ago. So I followed it to a tee until recently, all in funds of TIPS, ST Treasury, and IT Treasury in a 403b. (This year, I bought a big dose of a TBM fund due to its being the best choice in my 457.)

[MikeMak27]

Larry and lack_ey,

Thank you for the informative responses. I currently have my bond portion of my portfolio in taxable since I plan to buy a home in 5 years or so. Because of it being in taxable, I bought a bond etf instead of an index fund since bogleheads guide to investing says etf's are more tax efficient. Both of you suggest bond etf's are bad. Should I buy the equivalent index fund instead and why?

[aj76er]

would also avoid MBS as well as corporates so that excludes Total Bond funds.
Larry

Theory aside, it is interesting to note that TBM held up quite nicely in the last, historic, MBS-related crisis (e.g. 2008). I'm curious if you feel the extra diversification of TBM is "good enough" for most investors?

[Doc]

With GVI you're basically getting the same thing as 2/3 Vanguard Short-Term Bond Index Fund (BSV), 1/3 Vanguard Intermediate Term Bond Index Fund (BSV) but paying a higher ER and spreads. Those Vanguard funds cover 1-5 and 5-10 respectively. In both cases again you're getting corporates, a lot in the lower investment grade, especially in BIV in perhaps longer than would like, etc. And losing out on CD return above Treasuries.

This is basically what I do. It comes very close to Larry's recommendations. Because of tax considerations I break up the two Vg Treasury/Corporate index funds into their components but the overall portfolio is driven by the the BBgBarc 1-10 Gov/Credit index that GVI uses.

[larryswedroe]

ajer
Simple answer is NO. And TBM performance should be adjusted for the higher risks taken, both with calls and credit risks, along with the equity like risks of corporate debt. So don't just look at raw returns.
Also MBS didn't do as poorly as they normally might have when rates fall because people COULD NOT PREPAY and refi again because there wasn't equity. So you didn't get the usual earlier repayments.

With corporate debt, as I have pointed out, one should adjust for the beta exposure you really are holding. And most investors fail to account for it.

Larry

[lack_ey]

Larry and lack_ey,

Thank you for the informative responses. I currently have my bond portion of my portfolio in taxable since I plan to buy a home in 5 years or so. Because of it being in taxable, I bought a bond etf instead of an index fund since bogleheads guide to investing says etf's are more tax efficient. Both of you suggest bond etf's are bad. Should I buy the equivalent index fund instead and why?

The bit about tax efficiency for ETFs has to do with helping avoid capital gains distributions from turnover and investor redemptions in stock funds. Not really applicable to bonds.

The issues Larry Swedroe mentioned are addressed previously here:
http://www.etf.com/sections/index-inves ... nopaging=1
viewtopic.php?t=207918

It's more of a potential issue in times of market stress, and probably shouldn't materially impact ETFs holding high quality bonds, rather than some kind of constant disadvantage.

would also avoid MBS as well as corporates so that excludes Total Bond funds.
Larry

Theory aside, it is interesting to note that TBM held up quite nicely in the last, historic, MBS-related crisis (e.g. 2008). I'm curious if you feel the extra diversification of TBM is "good enough" for most investors?

To elaborate on the above, most MBS, like in Vanguard total bond particularly, are government-backed (GNMA) or pseudo-maybe-implicitly-government backed (Fannies, Freddies). These sailed right through the financial crisis.

The argument against these MBS is that the optionality is against the investor no matter which way rates go, not about credit risk. If rates go up, people keep their mortgages and don't prepay, and you're stuck with lower-yielding, longer-term bonds. If rates go down, people prepay their mortgages more and the higher rate you thought you were getting starts to evaporate. The convexity on these is thus negative, unlike say a Treasury bond. That's why the yield is higher on these than other bonds of similar credit risk.

Some like let's say Swedroe, David Swensen, etc. are saying that this risk is not worth it, that the discount is not high enough. That's a somewhat active judgment against the market, which of course sets the price to what it is now. Other people have other thoughts about this risk.

[tetractys]

I've been following Larry's guidelines, for the most part, since his bond book came out.

In taxable I use Vanguard Inter-Term Tax-Exempt Adm. In tax deferred I hold intermediate term Inflation-indexed securities along with intermediate term bonds. I try to avoid mortgage backed securities, which is easy at Vanguard by using Inter-Term Bond Index instead of Total Bond Mkt Index. One thing I've noticed is that Inter-Term diversifies better than Total Bond in most market swings. I also like the higher returns that I think have a lot to do with not having the negative convexity drag of mortgage backed securities.

Code: Select all

FROM VANGUARD'S WEBSITE:

            Total Bond Mkt Adm     Inter-Term Bond Adm	 	 	 
1-year      2.60%                  2.83%
3-year      2.94%                  3.66%
5-year      2.14%                  2.85%
10-year     4.29%                  5.37%

- Tet

[lack_ey]

I've been following Larry's guidelines, for the most part, since his bond book came out.

In taxable I use Vanguard Inter-Term Tax-Exempt Adm. In tax deferred I hold intermediate term Inflation-indexed securities along with intermediate term bonds. I try to avoid mortgage backed securities, which is easy at Vanguard by using Inter-Term Bond Index instead of Total Bond Mkt Index. One thing I've noticed is that Inter-Term diversifies better than Total Bond in most market swings. I also like the higher returns that I think have a lot to do with not having the reverse convexity drag of mortgage backed securities.

Code: Select all

FROM VANGUARD'S WEBSITE:

            Total Bond Mkt Adm     Inter-Term Bond Adm	 	 	 
1-year      2.60%                  2.83%
3-year      2.94%                  3.66%
5-year      2.14%                  2.85%
10-year     4.29%                  5.37%

- Tet

No, the intermediate-term bond index fund had higher returns than total bond over those periods because it took more term and credit risk. It's what you get for owning more corporates and sticking to longer duration over periods in which both were rewarded (credit risk not around 2008-2009 but largely ever since, including a rally through 2016, which gets picked up in the 1-year return especially and saves the intermediate-term fund by compensating for the underperformance of term risk as Treasury rates rose a bit).

[Doc]

I also like the higher returns that I think have a lot to do with not having the reverse convexity drag of mortgage backed securities.

The intermediate index also has a slightly higher duration 6.55 vs 5.94 for TBM. (M* data.) Whether MBS gives you higher or lower returns depends on what is happening in the mortgage market. (See Larry's response to ajer.)

Added based on lack-ey's post: A vs AA

[tetractys]

Code: Select all

FROM VANGUARD'S WEBSITE:

            Total Bond Mkt Adm     Inter-Term Bond Adm            
1-year      2.60%                  2.83%
3-year      2.94%                  3.66%
5-year      2.14%                  2.85%
10-year     4.29%                  5.37%
Av. mat.    8.3 years              7.2 years
Av. dur.    6.0 years              6.5 years
Av.coup.    3.1%                   2.8%

The inclusion or non-inclusion of MBS's is the largest factor that differentiates the two funds and the risks thereof. -- Tet

[Doc]

The inclusion or non-inclusion of MBS's is the largest factor that differentiates the two funds and the risks thereof. -- Tet

Code: Select all

Credit Quality		
	     VBILX	VBTLX
Type	% Bonds	Benchmark
		
AAA	     54.38	68.8
		
AA	     4.41	4.62
		
A	    16.61	11.92
		
BBB	    24.6	14.66
		
BB	     0		0
		
B	     0		0
		
Below B	0	0
		
Not Rated	0	0

OK so the addition of AAA MBS dilutes the amount of BBB so the lessor performance of TBM is because of the MBS. That may be true but it's a convoluted way of thinking about it.

But based on the Treasury yield the duration is not the difference.

[aj76er]

would also avoid MBS as well as corporates so that excludes Total Bond funds.
Larry

Theory aside, it is interesting to note that TBM held up quite nicely in the last, historic, MBS-related crisis (e.g. 2008). I'm curious if you feel the extra diversification of TBM is "good enough" for most investors?

To elaborate on the above, most MBS, like in Vanguard total bond particularly, are government-backed (GNMA) or pseudo-maybe-implicitly-government backed (Fannies, Freddies). These sailed right through the financial crisis.

The argument against these MBS is that the optionality is against the investor no matter which way rates go, not about credit risk. If rates go up, people keep their mortgages and don't prepay, and you're stuck with lower-yielding, longer-term bonds. If rates go down, people prepay their mortgages more and the higher rate you thought you were getting starts to evaporate. The convexity on these is thus negative, unlike say a Treasury bond. That's why the yield is higher on these than other bonds of similar credit risk.

Some like let's say Swedroe, David Swensen, etc. are saying that this risk is not worth it, that the discount is not high enough. That's a somewhat active judgment against the market, which of course sets the price to what it is now. Other people have other thoughts about this risk.

To both Larry and lack_ey: thank you for all the good insight!

So, in using the standard 3-fund, it appears I may not be accounting for the equity-like risk of TBM. My preferred AA is 60/35/5 (stocks/bonds/cash), with all 35% in TBM in my 401k (I'm fortunate to have access to institutional shares of Vanguard Total Bond Market @ 4bps). With this in mind, perhaps I can "correct" my equity exposure in a couple of different ways:

1. Increase % of TBM so that its treasury holdings account for my desired bond allocation. If assuming TBM holds ~70% treasuries, then I should increase my percentage by ~42%
-or-
2. Use the "brokerage link" option in my 401k to buy a pure treasury fund for the bond portion of my AA. Fidelity is my 401k custodian, so I could buy FIBAX (Fidelity Intermediate Treasury Bond Index Fund - Premium Class) @ 9bps

Do you agree with this assessment?
Thanks!

[larryswedroe]

You can adjust to account for the equity like holdings that corporates provide. However, the problem is if you look at the say average exposure you miss that the loadings on equity change dramatically in bear markets. So say a high yield fund might look like 25% equity risk, basically small value, but in bear market it can jump to 50%. That's one of the problems --the risks don't mix well.

And also the TBM composition shifts as more or less MBS issued.

To me I cannot really see any good reason to own TBM, but that's me
Larry

[matto]

You can adjust to account for the equity like holdings that corporates provide. However, the problem is if you look at the say average exposure you miss that the loadings on equity change dramatically in bear markets. So say a high yield fund might look like 25% equity risk, basically small value, but in bear market it can jump to 50%. That's one of the problems --the risks don't mix well.

This is also true of any vanilla stock/bond portfolio. An 50/50 stock/bond portfolio is for example 30/70 in volatility terms normally, but during a bear market becomes 10/90.

So if corporates are bad for this reason, then to be consistent you should be volatility balancing your equities during a bear market (i.e. selling stocks to keep volatility balanced), which I don't think you do.

[lack_ey]

You can adjust to account for the equity like holdings that corporates provide. However, the problem is if you look at the say average exposure you miss that the loadings on equity change dramatically in bear markets. So say a high yield fund might look like 25% equity risk, basically small value, but in bear market it can jump to 50%. That's one of the problems --the risks don't mix well.

This is also true of any vanilla stock/bond portfolio. An 50/50 stock/bond portfolio is for example 30/70 in volatility terms normally, but during a bear market becomes 10/90.

So if corporates are bad for this reason, then to be consistent you should be volatility balancing your equities during a bear market (i.e. selling stocks to keep volatility balanced), which I don't think you do.

The point there is not that junk bonds get more volatile during bear markets but that the correlation with equities increases then on the downside, and you get drawdowns at the worst possible time. Thus an analysis of how junk bonds fit with other assets might miss or understate the effect. They can look diversifying to some extent under a variety of environments but aren't likely to do so when it counts.

On the other hand, equities have a correlation of 1 with equities at all times.

[matto]

You can adjust to account for the equity like holdings that corporates provide. However, the problem is if you look at the say average exposure you miss that the loadings on equity change dramatically in bear markets. So say a high yield fund might look like 25% equity risk, basically small value, but in bear market it can jump to 50%. That's one of the problems --the risks don't mix well.

This is also true of any vanilla stock/bond portfolio. An 50/50 stock/bond portfolio is for example 30/70 in volatility terms normally, but during a bear market becomes 10/90.

So if corporates are bad for this reason, then to be consistent you should be volatility balancing your equities during a bear market (i.e. selling stocks to keep volatility balanced), which I don't think you do.

The point there is not that junk bonds get more volatile during bear markets but that the correlation with equities increases then on the downside, and you get drawdowns at the worst possible time. Thus an analysis of how junk bonds fit with other assets might miss or understate the effect. They can look diversifying to some extent under a variety of environments but aren't likely to do so when it counts.

On the other hand, equities have a correlation of 1 with equities at all times.

Everything you wrote also applies to an equity/treasury portfolio.

Here, let's construct a synthetic corporate bond. Let's pretend an AAPL bond is equal to 50% treasury + 50% AAPL stock.

cor(AAPL stock, AAPL bond) > 0.5 because the stock's volatility is greater than the treasury's volatility. Let's say it's 0.6.

No let's imagine a bear market. Treasury volatility goes up 50% and stock volatility goes up 100%.

Now the cor(AAPL stock, AAPL bond) has increased during the bear market (i.e. it is > 0.6) because the stock's volatility has increased more than the bond's volatility.

So I have just shown that a stock/bond portfolio's correlation to pure stocks goes UP during a bear market. This is exactly the same claim as corporate bonds.

So everyone agrees that a corporate bond is equity like and bond like. Swedroe is making the active management/market timing call that the combination is worse than the sum of its parts. The opposite point of view is that corporate bonds are *better* than the sum of their parts. I am making neither of those calls.

If you want to go further, then corporate bonds are essentially short bond volatility because they have call risk. Okay, sure. But they are also long equity volatility via the put option of liquidation preference (i.e. a corp bond is more like a treasury + stock + put call, or equivalently, treasury + call). So they are short some volatility, long other volatility.

So again, why do people here think that some factors (small, value) are real and worthwhile, while other factors (yield volatility) are unrewarded? Purely via backtesting. It is active management/market timing, let's be honest about it. Nothing wrong with that, and nothing wrong with choosing TBM for simplicity, but let's not use faulty logic.

[Fundhunter]

I was just reading Larry Swedroe's bonds book. He summarizes some general guidelines to follow. I'm wondering how many Bogleheads follow some of these. Here are a few:

-Avoid any bonds longer than 10-years
-Avoid bonds with ratings under AA
-Favor Treasuries and TIPS
-Municpals are OK if using the higher qualities (AAA, AA). Must diversify.
-Corporates have issues, not recommended unless highest quality. Must diversify
-Say no to Mortgage Backed Securities
-If using high-yield, add to your equities allocation, not bonds. In general, the risk has not been rewarded.

I follow this, except I own Vanguard Short Term Investment Grade Admiral (VFSUX), which has a small percentage of corporate bonds lower than AA. I hold VG Intermediate Bond Index also (VBILX). Except for TIPS, I avoid all long bonds like leprosy, so do not own any Total Bond index funds. Zero high yield junk. Of course, that means I own more than 3 funds, which is fine with me.

[lack_ey]

You can adjust to account for the equity like holdings that corporates provide. However, the problem is if you look at the say average exposure you miss that the loadings on equity change dramatically in bear markets. So say a high yield fund might look like 25% equity risk, basically small value, but in bear market it can jump to 50%. That's one of the problems --the risks don't mix well.

This is also true of any vanilla stock/bond portfolio. An 50/50 stock/bond portfolio is for example 30/70 in volatility terms normally, but during a bear market becomes 10/90.

So if corporates are bad for this reason, then to be consistent you should be volatility balancing your equities during a bear market (i.e. selling stocks to keep volatility balanced), which I don't think you do.

The point there is not that junk bonds get more volatile during bear markets but that the correlation with equities increases then on the downside, and you get drawdowns at the worst possible time. Thus an analysis of how junk bonds fit with other assets might miss or understate the effect. They can look diversifying to some extent under a variety of environments but aren't likely to do so when it counts.

On the other hand, equities have a correlation of 1 with equities at all times.

Everything you wrote also applies to an equity/treasury portfolio.

Here, let's construct a synthetic corporate bond. Let's pretend an AAPL bond is equal to 50% treasury + 50% AAPL stock.

cor(AAPL stock, AAPL bond) > 0.5 because the stock's volatility is greater than the treasury's volatility. Let's say it's 0.6.

No let's imagine a bear market. Treasury volatility goes up 50% and stock volatility goes up 100%.

Now the cor(AAPL stock, AAPL bond) has increased during the bear market (i.e. it is > 0.6) because the stock's volatility has increased more than the bond's volatility.

So I have just shown that a stock/bond portfolio's correlation to pure stocks goes UP during a bear market. This is exactly the same claim as corporate bonds.

So everyone agrees that a corporate bond is equity like and bond like. Swedroe is making the active management/market timing call that the combination is worse than the sum of its parts. The opposite point of view is that corporate bonds are *better* than the sum of their parts. I am making neither of those calls.

If you want to go further, then corporate bonds are essentially short bond volatility because they have call risk. Okay, sure. But they are also long equity volatility via the put option of liquidation preference (i.e. a corp bond is more like a treasury + stock + put call, or equivalently, treasury + call). So they are short some volatility, long other volatility.

So again, why do people here think that some factors (small, value) are real and worthwhile, while other factors (yield volatility) are unrewarded? Purely via backtesting. It is active management/market timing, let's be honest about it. Nothing wrong with that, and nothing wrong with choosing TBM for simplicity, but let's not use faulty logic.

You're right, under reasonable assumptions, stock/Treasury allocations become more correlated with stocks during bear markets. In of itself the property of becoming more correlated with stocks during bear markets is not necessarily disqualifying. It's a combination of all characteristics that are considered. And that broader case is made elsewhere. You're responding to a narrow part of the whole.

I agree that it's an active call, and I was just trying to state the argument without personally really agreeing with it.

Also, it's not obvious that you should decompose corporate bonds into equity and bond components, as credit risk is not really the same thing as equity risk and is only partially explained by it (or you can break it into further calls if applicable, vol, etc. as you did). So I don't think it's trivially true and goes without saying that you should expect higher correlation in bear markets based on this understanding.

[larryswedroe]

Matto
Equities correlation to equities are always 1 but not the case for corporate bonds, which become more equity like in bear markets for stocks, so your example doesn't hold. The point is that investors in corporate bonds have more tail risks than with Treasuries, and the lower the credit rating the more the equity like risks
Larry

[matto]

...

You're right, under reasonable assumptions, stock/Treasury allocations become more correlated with stocks during bear markets. In of itself the property of becoming more correlated with stocks during bear markets is not necessarily disqualifying. It's a combination of all characteristics that are considered. And that broader case is made elsewhere. You're responding to a narrow part of the whole.

I agree that it's an active call, and I was just trying to state the argument without personally really agreeing with it.

Also, it's not obvious that you should decompose corporate bonds into equity and bond components, as credit risk is not really the same thing as equity risk and is only partially explained by it (or you can break it into further calls if applicable, vol, etc. as you did). So I don't think it's trivially true and goes without saying that you should expect higher correlation in bear markets based on this understanding.

The decomposition was a simplified argument. Stock + bond is a first order approximation, stock + band + treasury volatility is a second order. And so forth.

I have seen the broader case and I disagree with it. I addressed an argument which was demonstrably false (the faulty correlation argument). I also addressed how call risk is repeatedly called an unrewarded risk, which is backwards looking and simply market timing/active management. If there are more arguments why corporate bonds are bad, then consider that the $T corp bond market thinks differently.

I'm not saying it's required, I'm saying people who say it is universally bad or 'unrewarded' (as if that implies it will be unrewarded) are just falling prey to backtesting. Again, if simplicity makes you want to avoid them, that's fine, but that's more of a "stocks/treasuries can accomplish anything corporate bonds can do, and so there is no downside (or upside) to avoiding them, however there is gained simplicity."

[matto]

Matto
Equities correlation to equities are always 1 but not the case for corporate bonds, which become more equity like in bear markets for stocks, so your example doesn't hold. The point is that investors in corporate bonds have more tail risks than with Treasuries, and the lower the credit rating the more the equity like risks
Larry

You again misunderstand the logic for why this is the case. Read my above post. *Any* stock/bond portfolio becomes more equity like in bear markets.

Here, let me change your quote to prove my point:

Equities correlation to equities are always 1 but not the case for a balanced portfolio, which become more equity like in bear markets for stocks, so your example doesn't hold. The point is that investors in a balanced portfolio have more tail risks than with Treasuries, and the more equity heavy the portfolio the more the equity like risks

[larryswedroe]

Matto
No that isn't correct, and I'm not misunderstanding anything. Equities are equities. Only thing that changes is that the risks of equities show up, increase. But corporate bonds not only become more risky but their correlation with equities goes up because they contain equity risks, risks that are correlated, while Treasuries tend to become negatively correlated in equity bear markets due to flight to quality.

[matto]

Matto
No that isn't correct, and I'm not misunderstanding anything. Equities are equities. Only thing that changes is that the risks of equities show up, increase. But corporate bonds not only become more risky but their correlation with equities goes up because they contain equity risks, risks that are correlated, while Treasuries tend to become negatively correlated in equity bear markets due to flight to quality.

I hate to be persistent, but yes you are misunderstanding. However, we can agree to disagree and I won't bother further explanation unless someone else wants it.

[larryswedroe]

Matto
Sorry don't agree but don't see any point in continuing as IMO you are simply twisting things---equities cannot become more equity like---they are equities, their risks are always there, it's just that sometimes they show up and other times they don't. The simplest example is that the market beta of a TSM portfolio is 1 and is always one. It doesn't change. But sometimes beta is more volatile and sometimes less. That's it. I'm done

[lack_ey]

...

You're right, under reasonable assumptions, stock/Treasury allocations become more correlated with stocks during bear markets. In of itself the property of becoming more correlated with stocks during bear markets is not necessarily disqualifying. It's a combination of all characteristics that are considered. And that broader case is made elsewhere. You're responding to a narrow part of the whole.

I agree that it's an active call, and I was just trying to state the argument without personally really agreeing with it.

Also, it's not obvious that you should decompose corporate bonds into equity and bond components, as credit risk is not really the same thing as equity risk and is only partially explained by it (or you can break it into further calls if applicable, vol, etc. as you did). So I don't think it's trivially true and goes without saying that you should expect higher correlation in bear markets based on this understanding.

The decomposition was a simplified argument. Stock + bond is a first order approximation, stock + band + treasury volatility is a second order. And so forth.

I have seen the broader case and I disagree with it. I addressed an argument which was demonstrably false (the faulty correlation argument). I also addressed how call risk is repeatedly called an unrewarded risk, which is backwards looking and simply market timing/active management. If there are more arguments why corporate bonds are bad, then consider that the $T corp bond market thinks differently.

I'm not saying it's required, I'm saying people who say it is universally bad or 'unrewarded' (as if that implies it will be unrewarded) are just falling prey to backtesting. Again, if simplicity makes you want to avoid them, that's fine, but that's more of a "stocks/treasuries can accomplish anything corporate bonds can do, and so there is no downside (or upside) to avoiding them, however there is gained simplicity."

I guess the question I would have is this:

If you take corporate bonds or high yield over a full business cycle, determine beta to equities for the full period, and subtract that out, does the residual [corp bond - average equity beta component] underperform during equity bear markets? Does the correlation or drawdowns during bear markets beyond the level expected based on the effective equity component they contain? Or alternatively, is the beta significantly higher or lower during certain periods? Or is the higher correlation with equities during bear markets already explained by underlying, roughly constant effective equity exposure?

The way I hear people talk about them, the implication I think is that there is residual underperformance and the risks are thus not worth it, but I haven't checked it for myself.

[Kevin M]

I follow all of Larry's recommendations as well as I can.

Really? What about this one:

Just to add that should use CDs instead of Treasuries when you have access to them, unless spread in yields very thin

(I keep Larry's bond book on the end table next to my computer. Thanks neighbor. )

Does the book include Larry's recommendations about CDs?

Kevin

[Taylor Larimore]

Does the book include Larry's recommendations about CDs?

Kevin:

Larry's book, "The Only Guide to a Winning Bond strategy You'll Ever Need" has four pages about Certificates of Deposit (CDs). This is his "Summary":

CDs (at least those that carry FDIC insurance) are generally safe investments and worth considering for a portion of a fixed-income portfolio, especially for tax-deferred or tax-exempt accounts."

Best wishes.
Taylor

[Cardio]

...

You're right, under reasonable assumptions, stock/Treasury allocations become more correlated with stocks during bear markets. In of itself the property of becoming more correlated with stocks during bear markets is not necessarily disqualifying. It's a combination of all characteristics that are considered. And that broader case is made elsewhere. You're responding to a narrow part of the whole.

I agree that it's an active call, and I was just trying to state the argument without personally really agreeing with it.

Also, it's not obvious that you should decompose corporate bonds into equity and bond components, as credit risk is not really the same thing as equity risk and is only partially explained by it (or you can break it into further calls if applicable, vol, etc. as you did). So I don't think it's trivially true and goes without saying that you should expect higher correlation in bear markets based on this understanding.

The decomposition was a simplified argument. Stock + bond is a first order approximation, stock + band + treasury volatility is a second order. And so forth.

I have seen the broader case and I disagree with it. I addressed an argument which was demonstrably false (the faulty correlation argument). I also addressed how call risk is repeatedly called an unrewarded risk, which is backwards looking and simply market timing/active management. If there are more arguments why corporate bonds are bad, then consider that the $T corp bond market thinks differently.

I'm not saying it's required, I'm saying people who say it is universally bad or 'unrewarded' (as if that implies it will be unrewarded) are just falling prey to backtesting. Again, if simplicity makes you want to avoid them, that's fine, but that's more of a "stocks/treasuries can accomplish anything corporate bonds can do, and so there is no downside (or upside) to avoiding them, however there is gained simplicity."

I guess the question I would have is this:

If you take corporate bonds or high yield over a full business cycle, determine beta to equities for the full period, and subtract that out, does the residual [corp bond - average equity beta component] underperform during equity bear markets? Does the correlation or drawdowns during bear markets beyond the level expected based on the effective equity component they contain? Or alternatively, is the beta significantly higher or lower during certain periods? Or is the higher correlation with equities during bear markets already explained by underlying, roughly constant effective equity exposure?

The way I hear people talk about them, the implication I think is that there is residual underperformance and the risks are thus not worth it, but I haven't checked it for myself.

Not Larry, but I recall he did a blog post in which he analyzed ST corporate bonds vs Treasuries as the FI portion of a balanced portfolio of equities and FI. Over the period analyzed (which was a non-cherry picked, reasonably long interval), the corporate FI fund gave no net return benefit and was associated with higher volatility. If memory serves, corporate bonds had a slightly higher return than Treasuries, but when placed in the portfolio, no net benefit due to the risks showing up at the wrong time, flight to quality,etc. as Larry has explained.

[Fundhunter]

I follow all of Larry's recommendations as well as I can.

Really? What about this one:

Just to add that should use CDs instead of Treasuries when you have access to them, unless spread in yields very thin

(I keep Larry's bond book on the end table next to my computer. Thanks neighbor. )

Does the book include Larry's recommendations about CDs?

Kevin

Yes, at page 113, he says FDIC insured CDs are worth considering for a portion of fixed income, particularly in tax-deferred or tax-exempt accounts.

I dusted my book off since this thread started and am rereading it.

[Kevin M]

Does the book include Larry's recommendations about CDs?

Yes, at page 113, he says FDIC insured CDs are worth considering for a portion of fixed income, particularly in tax-deferred or tax-exempt accounts.

Thanks for sharing, but "worth considering for a portion of fixed income" isn't nearly as strong an endorsement as what Larry has been saying more recently, including in this thread, which is that CDs should be used in preference to Treasuries (as long as the yield premium is reasonable). Larry also has stated that CDs sometimes make more sense even in taxable accounts for his firm's high net-worth clients, especially shorter maturities.

Kevin

[aj76er]

Does the book include Larry's recommendations about CDs?

Yes, at page 113, he says FDIC insured CDs are worth considering for a portion of fixed income, particularly in tax-deferred or tax-exempt accounts.

Thanks for sharing, but "worth considering for a portion of fixed income" isn't nearly as strong an endorsement as what Larry has been saying more recently, including in this thread, which is that CDs should be used in preference to Treasuries (as long as the yield premium is reasonable). Larry also has stated that CDs sometimes make more sense even in taxable accounts for his firm's high net-worth clients, especially shorter maturities.

Kevin

Hi Kevin,

Do you feel this recommendation applies to Brokered CDs as well? Relative to 5yr Treasuries, I don't think the spreads are that different (although haven't checked recently). I believe this is the only type of CD available to me in my 401k. Even if there were some premium to CDs, it seems it would be difficult to rebalance with equities if/when tolerance bands are hit.

Thanks!

[robertmcd]

What is larry's opinion on using vanguard inter term government bond index vs inter term treasury? the admiral shares government index is currently yielding 1.93% vs the admiral shares treasury 1.81%. Isn't the government index protected against default risk?

[larryswedroe]

robert, that is fine, unless have access to CDs.

ajer
I have owned brokered CDs and if need be can sell them prior to maturity, though not as liquid as Treasuries, so you may want to keep some liquidity in safer assets to rebalance in bear markets

Kevin, see comment to ajer above. It's one reason one should consider holding some Treasuries or government/agencies vs. all CDs. Trade off is the higher yield in CDs with the lower benefit they offer in bear markets with liquidity flights. One is certain, the other tends to show up rarely, though does happen. And of course in most 401k plans don't have access to CDs.

Larry