Larry Swedroe: No Need For Corp Bonds?

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
Post Reply
Random Walker
Posts: 3193
Joined: Fri Feb 23, 2007 8:21 pm

Larry Swedroe: No Need For Corp Bonds?

Post by Random Walker » Fri May 18, 2018 7:33 am

http://www.etf.com/sections/index-inves ... nopaging=1

Corporate bonds are really hybrid securities: a mix of the risks found in Treasuries and stocks. Even though corporate bonds make up a substantial portion of the world fixed income market, the case for adding them to an individual investor’s portfolio is weak. Larry reviews a paper written by his colleague Jared Kizer. The return on corporate bonds falls short of their yields due to defaults and exercising of call options. It is likely more efficient to construct a portfolio with a somewhat higher equity allocation and only high quality bonds than to use a lower equity exposure and corporate bonds exposed to the credit premium. On top of all this, expenses in the corporate bond market are greater than in the government bond market. And corporates are tax inefficient. The hurdles for corporate bonds are substantial: credit risk, call risk, liquidity risk, tax inefficiency. The correlation between corporate bonds and equities tends to become more positive at just the wrong time, when equities are tanking. Lastly, since corporates entail credit risk, one should diversify by buying through a fund and paying expense ratio. This is not necessary with riskless treasuries.

Dave

MinhN
Posts: 32
Joined: Sun Mar 11, 2018 11:57 pm

Re: Larry Swedroe: No Need For Corp Bonds?

Post by MinhN » Fri May 18, 2018 8:34 am

Despite the credit risk of corporate bonds, I am under the impression that bondholders are paid before shareholders. Is this not a risk premium worth considering in the face of an economic downturn when bankruptcy is rife and equities are tanking?

User avatar
oldcomputerguy
Posts: 3449
Joined: Sun Nov 22, 2015 6:50 am
Location: In the middle of five acres of woods

Re: Larry Swedroe: No Need For Corp Bonds?

Post by oldcomputerguy » Fri May 18, 2018 8:53 am

MinhN wrote:
Fri May 18, 2018 8:34 am
Despite the credit risk of corporate bonds, I am under the impression that bondholders are paid before shareholders. Is this not a risk premium worth considering in the face of an economic downturn when bankruptcy is rife and equities are tanking?
The flip side to that, though, is that, when things are tanking and the bond issuer experiences difficulties and/or declares bankruptcy, the chance of default on a corporate bond increases.
It’s taken me a lot of years, but I’ve come around to this: If you’re dumb, surround yourself with smart people. And if you’re smart, surround yourself with smart people who disagree with you.

dbr
Posts: 27207
Joined: Sun Mar 04, 2007 9:50 am

Re: Larry Swedroe: No Need For Corp Bonds?

Post by dbr » Fri May 18, 2018 9:05 am

The investigation is interesting but it is hard to know what significance to attach to data that essentially consists of the biggest interest rate spike in US history and the thirty year or more decline from that spike. How these results would add up starting today and extending for the next sixty years is an unanswered question. https://www.google.com/search?q=histori ... m40gX6qm1M:

On the other hand, looking at this kind of thing does suggest that it is very difficult to find magic answers to the perennial question on this forum: "What bond fund should I buy?"

AlohaJoe
Posts: 3833
Joined: Mon Nov 26, 2007 2:00 pm
Location: Saigon, Vietnam

Re: Larry Swedroe: No Need For Corp Bonds?

Post by AlohaJoe » Fri May 18, 2018 9:16 am

A bit buried in Kizer's paper is his fundamental claim that Ibbotson's corporate bond return data is wrong for from 1927 to the 1950s or 1960s and overstate returns by ~2% a year.

Considering that that data is used for all kinds of historical backtesting, it would be interesting to see Ibbotson's reply.

I've always felt these kind of articles on corporate bonds a bit unsatisfying because the answer to "why doesn't the credit risk premium exist?" doesn't feel fully formed to me.

Also: Kizer's paper is a rebuttal to an AQR paper from last year in the Journal of Fixed Income saying the exact opposite: that the credit risk premium is real & large. So....I'd hardly call it a definitive end to fight. I have no doubt that AQR will have a reply......

User avatar
nisiprius
Advisory Board
Posts: 36878
Joined: Thu Jul 26, 2007 9:33 am
Location: The terrestrial, globular, planetary hunk of matter, flattened at the poles, is my abode.--O. Henry

Re: Larry Swedroe: No Need For Corp Bonds?

Post by nisiprius » Fri May 18, 2018 9:39 am

I can't deal with half the world saying that the Bloomberg Barclay's Aggregate Index is too heavily weighted in government issues and that everyone should have more corporates than that, half the world saying that everyone should be 100% Treasuries, and the third half saying that everyone should hold some high-yield bonds. (With emerging markets bonds as an interesting special case of high-yield: Burton Malkiel and Charles Ellis suggested replacing Total Bond with a "surrogate" of one-half dividend stocks and one-half emerging markets bonds).

I think that unless you are close to 100% invested in bonds, it's all angels-dancing-on-the-head-of-a-pin stuff.

As for the bond bull market, etc. absolutely everything that has ever happened in the history of US bonds has been unique in history. If you want to make any extrapolations from two-and-a-half data points, be my guest. I'm sticking with the idea that investment-grade bonds are debt that will almost certainly be repaid, and there just can't be a heck of a lot of difference between two bonds that have the same term, coupon, and rating.

Image
Last edited by nisiprius on Fri May 18, 2018 10:39 am, edited 2 times in total.
Annual income twenty pounds, annual expenditure nineteen nineteen and six, result happiness; Annual income twenty pounds, annual expenditure twenty pounds ought and six, result misery.

dbr
Posts: 27207
Joined: Sun Mar 04, 2007 9:50 am

Re: Larry Swedroe: No Need For Corp Bonds?

Post by dbr » Fri May 18, 2018 9:42 am

^well said, and exactly my point as well.

dcabler
Posts: 661
Joined: Wed Feb 19, 2014 11:30 am

Re: Larry Swedroe: No Need For Corp Bonds?

Post by dcabler » Fri May 18, 2018 10:11 am

nisiprius wrote:
Fri May 18, 2018 9:39 am

I think that unless you are close to 100% invested in bonds, it's all angels-dancing-on-the-head-of-a-pin stuff.


Image
This.
Disclosure: I do treasuries where I can. Total Bond where I can't. Treasuries did a little better than total in 2008, but each correction is different. In the limit, probably macht nicht..

User avatar
Munir
Posts: 2466
Joined: Mon Feb 26, 2007 4:39 pm
Location: Oregon

Re: Larry Swedroe: No Need For Corp Bonds?

Post by Munir » Fri May 18, 2018 10:18 am

If nisiprius and dbr agree on something, it would be difficult for anyone to dissuade me otherwise.

Have Vanguard intermediate bond funds ever suffered from bond defaults in a significant fashion?

User avatar
triceratop
Moderator
Posts: 5724
Joined: Tue Aug 04, 2015 8:20 pm
Location: la la land

Re: Larry Swedroe: No Need For Corp Bonds?

Post by triceratop » Fri May 18, 2018 10:28 am

I can't deal with half the world saying that the Bloomberg Barclay's Aggregate Index is too heavily weighted in government issues and that everyone should have more corporates than that, half the world say that everyone should be 100% Treasuries, and the third half saying that everyone should hold some high-yield bonds. (With emerging markets bonds as an interesting special case of high-yield: Burton Malkiel and Charles Ellis suggested replacing Total Bond with a "surrogate" of one-half dividend stocks and one-half emerging markets bonds).

I think that unless you are close to 100% invested in bonds, it's all angels-dancing-on-the-head-of-a-pin stuff.
Sure, but aren't at least two of these saying these things for different and also mostly compatible reasons? It seems to me that you can simultaneously criticize an index for having silly construction rules and also believe that credit risk hasn't been sufficiently rewarded.
"To play the stock market is to play musical chairs under the chord progression of a bid-ask spread."

stuper1
Posts: 67
Joined: Tue Apr 03, 2018 9:30 am

Re: Larry Swedroe: No Need For Corp Bonds?

Post by stuper1 » Fri May 18, 2018 10:42 am

For me, I see more safety in Treasuries and less correlation to stocks, along with about the same return as a total bond index fund, so I prefer to use full Treasuries when possible. Sometimes 401k limited choices make that difficult.

lack_ey
Posts: 6627
Joined: Wed Nov 19, 2014 11:55 pm

Re: Larry Swedroe: No Need For Corp Bonds?

Post by lack_ey » Fri May 18, 2018 10:50 am

MinhN wrote:
Fri May 18, 2018 8:34 am
Despite the credit risk of corporate bonds, I am under the impression that bondholders are paid before shareholders. Is this not a risk premium worth considering in the face of an economic downturn when bankruptcy is rife and equities are tanking?
Sure, but in theory the entire market could be systemically mispriced, so it doesn't result in anything in practice. That's kind of the position you would need to take on this to make the point against.
AlohaJoe wrote:
Fri May 18, 2018 9:16 am
A bit buried in Kizer's paper is his fundamental claim that Ibbotson's corporate bond return data is wrong for from 1927 to the 1950s or 1960s and overstate returns by ~2% a year.
I don't think we should particularly trust the older data anyway that much, as it was a significantly different market then. But that point is kind of interesting.


Jared Kizer doesn't examine corporate bond data outside the U.S. or credit-default swaps, both of which were used in addition to U.S. corporate bond data in Asvanunt and Richardson [2017].

There are differences as well in how the papers attempt to duration match.

Also should be noted that "not significant" according to a statistical test doesn't mean "doesn't exist." There's a failure to reject a null hypothesis at a certain level, maybe from not having enough data, etc., even if we take the model and methodology to be correct.

The methodology is also maybe not the most appropriate or highest power available given all the noise in returns data. We know the causal mechanism by which credit excess returns would exist if in fact they do: higher yields over Treasuries. The question is whether the yield is high enough to overcome effects from defaults and net downgrades. Excess returns data is noisy in the sense that it's significantly driven by changes in spreads (increasing vol and making the mean effect size harder to see), yet we know that over the long run net change in spreads should be small, unless we expect very weird things.

User avatar
Blueskies123
Posts: 439
Joined: Sat Nov 15, 2014 7:18 pm
Location: South Florida

Re: Larry Swedroe: No Need For Corp Bonds?

Post by Blueskies123 » Fri May 18, 2018 10:51 am

Munir wrote:
Fri May 18, 2018 10:18 am
If nisiprius and dbr agree on something, it would be difficult for anyone to dissuade me otherwise.

Have Vanguard intermediate bond funds ever suffered from bond defaults in a significant fashion?
If you go to page 15 in this document:
https://www.nber.org/papers/w15848

You can see there have been period with high defaults. I know they were long ago but if we like to back test stocks as far back as we can we should also be open to back testing bonds a long time.

User avatar
Munir
Posts: 2466
Joined: Mon Feb 26, 2007 4:39 pm
Location: Oregon

Re: Larry Swedroe: No Need For Corp Bonds?

Post by Munir » Fri May 18, 2018 11:03 am

Blueskies123 wrote:
Fri May 18, 2018 10:51 am
Munir wrote:
Fri May 18, 2018 10:18 am
If nisiprius and dbr agree on something, it would be difficult for anyone to dissuade me otherwise.

Have Vanguard intermediate bond funds ever suffered from bond defaults in a significant fashion?
If you go to page 15 in this document:
https://www.nber.org/papers/w15848

You can see there have been period with high defaults. I know they were long ago but if we like to back test stocks as far back as we can we should also be open to back testing bonds a long time.
My question dealt specifically with Vanguard intermediate bond funds that include quality (not junk) corporate issues in their holdings. Have any of them suffered significantly from individual bond defaults in the last 30 years?

garlandwhizzer
Posts: 2021
Joined: Fri Aug 06, 2010 3:42 pm

Re: Larry Swedroe: No Need For Corp Bonds?

Post by garlandwhizzer » Fri May 18, 2018 12:16 pm

nisi wrote:

I think that unless you are close to 100% invested in bonds, it's all angels-dancing-on-the-head-of-a-pin stuff.

As for the bond bull market, etc. absolutely everything that has ever happened in the history of US bonds has been unique in history. If you want to make any extrapolations from two-and-a-half data points, be my guest. I'm sticking with the idea that investment-grade bonds are debt that will almost certainly be repaid, and there just can't be a heck of a lot of difference between two bonds that have the same term, coupon, and rating.
1+

The big picture is that any ST or IT Investment Grade Bond Fund has a very wide divergence from equity volatility. Those which diverge more in equity downturns (corporates) wind up producing higher long term returns due to their increased yields. It's a tradeoff pure and simple. IMO bond holdings are best chosen looking at the portfolio as a whole and at your own circumstances. For example if you have an equity dominated portfolio (75% - 90%) and a small bond allocation (10% - 25%) it makes sense to go with the safest, most liquid, highest quality bonds only (Treasuries) and omit corporates in order to get maximum portfolio stability per dollar of bonds in times of equity crisis. If on the other hand you have a 30/70 or 50/50 equity/bond mix, including corporates as part of the bond allocation seems to me reasonable in some circumstances. A lot of specifics like risk and volatility tolerance, financial goals, and size of portfolio vary widely from one investor to another. It may be overstating the point to say that everyone should always have only Treasuries in all circumstances. My approach is to try to fit bonds into my own specific needs and circumstances which may or may not include corporates, GNMA, or TIPS with Treasuries and may vary in average duration as interest rates and the macroeconomic picture changes. If you're not sure about the details of what to do, I personally believe that TBM is a very good default choice as is ST or IT Treasuries. All of these will fulfill the essential role of bonds in a portfolio.

Garland Whizzer

Dandy
Posts: 5406
Joined: Sun Apr 25, 2010 7:42 pm

Re: Larry Swedroe: No Need For Corp Bonds?

Post by Dandy » Sat May 19, 2018 6:27 am

Interesting discussion. I don't have the technical bond information and analysis skills that many posters have. I think that fixed income allocation often gets people confused e.g. they may mentally target their fixed income for stability/safety but then rue the low returns that safer fixed income products provide. Just like most people understand that the idea of an emergency fund is that the fund is there when you need it but then they just can't stand the low returns that a "safe" E fund will generate.

I also think some bond investors have been enticed to invest in intermediate bonds or longer due to the 30 year or so bond bull market. Are those returns likely to be experienced going forward for the next decade or so?

I have decided to tilt my fixed income allocation toward safety in retirement. Maybe I've been influenced by decades of using a nice Stable Value Fund during my accumulation stage. I have almost 1/3 in no loss of principal (Savings, money market, CDs, legacy EE Bonds), almost 1/3 in short term bond funds and a bit more than 1/3 in intermediate bond funds).
Very little Corporate bond exposure.

I have been influenced by Dr. Bernstein's idea of keeping X years of yearly draw down in "safer" fixed income. In retirement with a higher fixed income allocation the type of fixed income used and its allocation becomes more critical.

columbia
Posts: 987
Joined: Tue Aug 27, 2013 5:30 am

Re: Larry Swedroe: No Need For Corp Bonds?

Post by columbia » Sat May 19, 2018 7:57 am

Can someone point me to a time when owning a broad based corporate bond fund led to losing a tangible percentage of money and that this was *specifically* due to companies defaulting on the bond payments?

User avatar
nisiprius
Advisory Board
Posts: 36878
Joined: Thu Jul 26, 2007 9:33 am
Location: The terrestrial, globular, planetary hunk of matter, flattened at the poles, is my abode.--O. Henry

Re: Larry Swedroe: No Need For Corp Bonds?

Post by nisiprius » Sat May 19, 2018 9:28 am

The Kizer paper is pretty interesting and deserves more discussion. The link to the paper is Re-Examining the Credit Premium and the full paper can be downloaded there at no cost.

I have a problem with the abstract, which states, in part:
The differences in findings appear to be driven by the abnormally high investment-grade credit (IGC) premium associated with the early history of the Ibbotson/Sinquefield data. An examination of that early history calls into question whether the data is truly representative of investment-grade corporate bond returns net of defaults and downgrades. The primary results questioning the early period data are 1) a significantly higher Sharpe Ratio for the IGC premium compared to either equities or interest rate risk during a period encompassing the Great Depression 2) anomalously high returns during the 30s, which was a period of significant credit stress...
I take this to mean that he doesn't have any direct evidence for data quality issues in the earlier years other than an "abnormally" high credit premium and Sharpe ratio. That is to say, his chief reason for doubting the data is that to him it looks too good. I imagine his judgement is sound, but...

Abnormal periods of time do occur. Financial data is bursty. In fact it can be argued that the reason for adopting certain portfolios is not an expectation of a general year-by-year-improvement every year, but a hope that every few decades one will catch one of those one brief shining moments. The small-cap premium, for example. Siegel has observed that all of it is accounted for by the single time period 1975 through 1983. Remove it and the premium goes away. What, exactly, do you say about this? Do you say "it was anomalous, throw it out?" Or do you say "yes, it's there, but you may have to wait an extremely long time for it to show up?" How do you know if the early corporate bond data is wrong, or if it is just the bond equivalent of 1975-1983 in small-cap stocks?

The thing I would take from this is not "the Ibbotson numbers for corporate bonds are too high" but a freshening of my general skepticism about the accuracy of long-term data. I don't believe any of it is accurate enough to support the precision needed to draw the conclusions writers draw. For example, to determine whether or not adding corporates to a stocks/Treasuries portfolio will produce a modest, statistical, long-term benefit in risk-adjusted return.
Last edited by nisiprius on Sat May 19, 2018 11:53 am, edited 1 time in total.
Annual income twenty pounds, annual expenditure nineteen nineteen and six, result happiness; Annual income twenty pounds, annual expenditure twenty pounds ought and six, result misery.

Random Walker
Posts: 3193
Joined: Fri Feb 23, 2007 8:21 pm

Re: Larry Swedroe: No Need For Corp Bonds?

Post by Random Walker » Sat May 19, 2018 11:38 am

Nisiprius makes a great comment about data. In general, longer term data has the potential to be more meaningful because it displays persistence of whatever factor one is looking at. But circumstances change, and shorter more recent data might be more relevant to our current circumstance. To really get a feel for a financial phenomenon, I think we need to look at data over the longest period available and recent sub periods. What really gives me confidence in data is the out of sample tests: different geographic market or for styles even different asset classes.

Dave

User avatar
jeffyscott
Posts: 7328
Joined: Tue Feb 27, 2007 9:12 am
Location: Wisconsin

Re: Larry Swedroe: No Need For Corp Bonds?

Post by jeffyscott » Sat May 19, 2018 5:50 pm

Seems like there's even more of a case there that there is no need for treasuries, that we'd be better off with CDs.
...for individual investors who have access to the CD market. FDIC-insured CDs, which also have no credit risk (as long as you remain within the limits of the insurance), typically carry significantly higher yields (than treasuries)...
Perhaps replacing total bond market fund with a combination of CDs and a corporate bond fund (in tax deferred/Roth accounts) might be a good idea.
...part of the higher yield that investors require on corporate bonds over Treasury bonds is related to the difference in tax treatment...
So that extra yield is a free lunch (or at least a light snack) in TIRA/Roth accounts.

FWIW, over the last 10 years, based on Vanguard funds, long term investment grade returned 7.14%, long term treasury 5.63% and for intermediate it's 4.78% vs. 3.14%. Total stock market was 9.24%. 60% stock and 40% treasuries would be 7.8% or 6.8%, depending on which fund and to get that same return with the corporate bond funds would have required only 31% or 45% stocks rather than 60%.
press on, regardless - John C. Bogle

User avatar
fortyofforty
Posts: 1151
Joined: Wed Mar 31, 2010 12:33 pm

Re: Larry Swedroe: No Need For Corp Bonds?

Post by fortyofforty » Sat May 19, 2018 6:11 pm

I can't find any period using Vanguard data that shows Treasuries outperforming Corporate bonds. There absolutely seems to be a risk premium. Whether it is worth it to an investor is up to her. I don't mind high quality Corporate bonds, which are still much less volatile than stocks, while offering slightly higher returns over any measured period. Not a free lunch, but certainly low-priced.
"In a time of universal deceit, telling the truth becomes a revolutionary act." - George Orwell | There are many roads to doublin'. | Original Vanguard Diehard

User avatar
Robert T
Posts: 2534
Joined: Tue Feb 27, 2007 9:40 pm
Location: 1, 0.2, 0.4, 0.5
Contact:

Re: Larry Swedroe: No Need For Corp Bonds?

Post by Robert T » Sat May 19, 2018 6:51 pm

.
2003-2017: Annualized Return (%) / SD
Long-term corporate bonds = 6.9 / 6.6
Long-term government bonds = 6.0 / 12.3

Some may look at this data and say long-term corporate bonds were ‘better’ because they had 0.9% higher return with almost half the volatility (double the Sharpe ratio). This is true if you were holding a 100% bond portfolio – but if not, it is important not to look at bonds in isolation of your stock holdings.

Added as the bond portfolio of a 75:25 stock:bond portfolio over the same time period

2003-2017: Annualized Return (%) / SD
With long-term corporate bonds = 11.3 / 15.8
With long-term government bonds = 11.4 / 14.1

The portfolio with government bonds had similar/slightly higher returns but with greater downside protection. Same result over longer time periods.

The long-term corporate bonds, long-term government bonds, and stock portfolio (DFA “Balanced” Equity) series are from the DFA Matrix Book.

These results are consistent with Swensen’s views in his 2000 book "Pioneering Portfolio Management". I have not seen a convincing case to add corporate bonds (credit/default risk) to an equity oriented portfolio, and have had a zero default load target in fixed income for the past 15 years. I don’t think I have missed anything (had greater downside protection in 2008, without forgoing long-term portfolio returns).

Obviously, no guarantees.

Robert
.

User avatar
triceratop
Moderator
Posts: 5724
Joined: Tue Aug 04, 2015 8:20 pm
Location: la la land

Re: Larry Swedroe: No Need For Corp Bonds?

Post by triceratop » Sat May 19, 2018 6:57 pm

columbia wrote:
Sat May 19, 2018 7:57 am
Can someone point me to a time when owning a broad based corporate bond fund led to losing a tangible percentage of money and that this was *specifically* due to companies defaulting on the bond payments?
Isn't that a really hard thing to do, simply because of how the funds are set up? Aren't many of these corporate bond funds investment grade corporate bond funds which sell what become fallen angels? Just because bonds you own aren't going into default doesn't mean you aren't losing money.
"To play the stock market is to play musical chairs under the chord progression of a bid-ask spread."

lack_ey
Posts: 6627
Joined: Wed Nov 19, 2014 11:55 pm

Re: Larry Swedroe: No Need For Corp Bonds?

Post by lack_ey » Sat May 19, 2018 7:02 pm

triceratop wrote:
Sat May 19, 2018 6:57 pm
columbia wrote:
Sat May 19, 2018 7:57 am
Can someone point me to a time when owning a broad based corporate bond fund led to losing a tangible percentage of money and that this was *specifically* due to companies defaulting on the bond payments?
Isn't that a really hard thing to do, simply because of how the funds are set up? Aren't many of these corporate bond funds investment grade corporate bond funds which sell what become fallen angels? Just because bonds you own aren't going into default doesn't mean you aren't losing money.
Vanguard's active bond funds hold some small amount of junk, probably some fallen angels but maybe not. I haven't really checked. For the investment-grade index funds, sure. Much more common to see a downgrade than an outright default.

A fund can also lose from net downgrades even if they're not to junk.

columbia
Posts: 987
Joined: Tue Aug 27, 2013 5:30 am

Re: Larry Swedroe: No Need For Corp Bonds?

Post by columbia » Sat May 19, 2018 7:18 pm

triceratop wrote:
Sat May 19, 2018 6:57 pm
columbia wrote:
Sat May 19, 2018 7:57 am
Can someone point me to a time when owning a broad based corporate bond fund led to losing a tangible percentage of money and that this was *specifically* due to companies defaulting on the bond payments?
Isn't that a really hard thing to do, simply because of how the funds are set up? Aren't many of these corporate bond funds investment grade corporate bond funds which sell what become fallen angels? Just because bonds you own aren't going into default doesn't mean you aren't losing money.
It is probably difficult to isolate, which is also probably another reason to avoid getting upset about - seemingly - overblown fear of corporate default: it happens, but you’re unlikely to notice it, amidst a properly diversified fund.

AlohaJoe
Posts: 3833
Joined: Mon Nov 26, 2007 2:00 pm
Location: Saigon, Vietnam

Re: Larry Swedroe: No Need For Corp Bonds?

Post by AlohaJoe » Sat May 19, 2018 8:58 pm

nisiprius wrote:
Sat May 19, 2018 9:28 am
I take this to mean that he doesn't have any direct evidence for data quality issues in the earlier years other than an "abnormally" high credit premium and Sharpe ratio. That is to say, his chief reason for doubting the data is that to him it looks too good. I imagine his judgement is sound, but...
I had the same reaction; his apparent lack of direct evidence was a bit disappointing. I also had the same thoughts about "regime changes" or time-varying premia. We know that the investing world was quite different in the distant past. People talk about how hard it is to invest in corporate bonds today! (outside of mutual funds/ETFs I mean; they are thinly traded, terms & conditions may be poorly understood, etc, etc). I can only imagine how hard it was in 1934 or whatever when only 1/3 of houses even had a phone and I don't even know how bank transfers to your broker worked. Instead of T+2...how long were settlement periods back then?

User avatar
fortyofforty
Posts: 1151
Joined: Wed Mar 31, 2010 12:33 pm

Re: Larry Swedroe: No Need For Corp Bonds?

Post by fortyofforty » Sat May 19, 2018 9:16 pm

Robert T wrote:
Sat May 19, 2018 6:51 pm
.
2003-2017: Annualized Return (%) / SD
Long-term corporate bonds = 6.9 / 6.6
Long-term government bonds = 6.0 / 12.3

Some may look at this data and say long-term corporate bonds were ‘better’ because they had 0.9% higher return with almost half the volatility (double the Sharpe ratio). This is true if you were holding a 100% bond portfolio – but if not, it is important not to look at bonds in isolation of your stock holdings.

Added as the bond portfolio of a 75:25 stock:bond portfolio over the same time period

2003-2017: Annualized Return (%) / SD
With long-term corporate bonds = 11.3 / 15.8
With long-term government bonds = 11.4 / 14.1

The portfolio with government bonds had similar/slightly higher returns but with greater downside protection. Same result over longer time periods.

The long-term corporate bonds, long-term government bonds, and stock portfolio (DFA “Balanced” Equity) series are from the DFA Matrix Book.

These results are consistent with Swensen’s views in his 2000 book "Pioneering Portfolio Management". I have not seen a convincing case to add corporate bonds (credit/default risk) to an equity oriented portfolio, and have had a zero default load target in fixed income for the past 15 years. I don’t think I have missed anything (had greater downside protection in 2008, without forgoing long-term portfolio returns).

Obviously, no guarantees.

Robert
.
Do you have access to data exploring short-term corporate and treasuries, intermediate-term corporate and treasuries, and total corporate and total government bond results? Just curious. I wouldn't recommend restricting yourself to long-term anything, corporate or government.
"In a time of universal deceit, telling the truth becomes a revolutionary act." - George Orwell | There are many roads to doublin'. | Original Vanguard Diehard

greenhill
Posts: 26
Joined: Sun Aug 06, 2017 4:27 am

Re: Larry Swedroe: No Need For Corp Bonds?

Post by greenhill » Sun May 20, 2018 12:57 am

I have an interesting question, though. Does it make sense for non-resident aliens (NRAs) to invest in corporate bonds? NRAs are subject to 30% dividend tax for equities while bonds are tax free. I am an NRA and I only hold the total bond market and TIPS, though.

User avatar
welderwannabe
Posts: 809
Joined: Fri Jun 16, 2017 8:32 am

Re: Larry Swedroe: No Need For Corp Bonds?

Post by welderwannabe » Sun May 20, 2018 6:11 am

greenhill wrote:
Sun May 20, 2018 12:57 am
I have an interesting question, though. Does it make sense for non-resident aliens (NRAs) to invest in corporate bonds? NRAs are subject to 30% dividend tax for equities while bonds are tax free. I am an NRA and I only hold the total bond market and TIPS, though.
Bonds are tax free? I thought that only applied to bond interest from non US companies (although I am only familiar with it in passing)? I am pretty sure bank interest is tax free for aliens, but I hadn't heard that applied to bonds.
I am not an investment professional, but I did stay at a Holiday Inn Express last night.

User avatar
welderwannabe
Posts: 809
Joined: Fri Jun 16, 2017 8:32 am

Re: Larry Swedroe: No Need For Corp Bonds?

Post by welderwannabe » Sun May 20, 2018 6:26 am

In tax advantaged I am roughly 50% corporates and 50% Treasuries/CDs. My time horizon is pretty long, so I am not worried. As long as you didn't sell, people that stayed in corporates through 2009 ended up about even with treasuries by 2011 and better off from that point on. Below is a comparison of intermediate treasury fund with LQD. Who knows how bonds will fare in the next recession.


Image
I am not an investment professional, but I did stay at a Holiday Inn Express last night.

User avatar
Robert T
Posts: 2534
Joined: Tue Feb 27, 2007 9:40 pm
Location: 1, 0.2, 0.4, 0.5
Contact:

Re: Larry Swedroe: No Need For Corp Bonds?

Post by Robert T » Sun May 20, 2018 8:39 am

fortyofforty wrote:
Sat May 19, 2018 9:16 pm
Do you have access to data exploring short-term corporate and treasuries, intermediate-term corporate and treasuries, and total corporate and total government bond results? Just curious. I wouldn't recommend restricting yourself to long-term anything, corporate or government.
Similar result with short and intermediate term bonds using Vanguard funds as an example.
  • • When viewed on their own, the Vanguard short, intermediate, and long-term investment grade [corporate] bonds had higher returns and higher Sharpe ratios than their treasury counterparts.

    • When viewed as part of a 75:25 stock:bond portfolio, the portfolios that included treasuries had similar/same returns and lower downside risk than the portfolios with investment grade corporate bonds.

    • If the stock:bond mix is adjusted to equate volatility (standard deviation), the portfolios with treasuries had between 0.3% (with short-term bonds), and 0.4% (with intermediate or long-term bonds) higher returns than the portfolios with investment grade corporate bonds.

    • If the stock:bond mix is adjusted to equate 2008 downside, the portfolios with treasuries had 0.4% (with short-term bonds), 0.5% (with long-term bonds), and 0.7% (with intermediate-term bonds) higher returns than the portfolios with investment grade corporate bonds.
Corporate bonds have some equity characteristics, as reflected in the OP linked article and Fama-French’s original 1993 article as summarized here (even for the highest grade corporate bonds). Including corporate bonds and stocks in a portfolio typically results in holding the same companies on the stock and bond side of your portfolio. Treasuries provide a ‘purer’ form of diversification (as Swensen describes in detail in his book). An approach of taking risk in equities rather than (default/credit) risk in fixed income (as there is already default risk in equities) has served me well so far. And staying the course with this approach is helped by being underpinned by theory (e.g. Swensen), historical performance, and actual personal portfolio performance over the past 15 years.

Obviously, no guarantees.

Robert

1994-2017 [longest period data is available for all Vanguard funds listed below]
-----------------------------------------------------------------------------------------------------

Annualized return (%)/SD

3.7 / 3.5 = Vanguard Short-term Treasury
4.3 / 4.0 = Vanguard Short-term Investment Grade

5.2 / 6.3 = Vanguard Intermediate-term Treasury
5.8 / 6.2 = Vanguard Intermediate-term Investment Grade

6.9 / 12.1 = Vanguard Long-term Treasury
7.3 / 7.9 = Vanguard Long-term Investment Grade

------------------------------------------------------------------------------------------------------

When included as part of a stock:bond portfolio [stock portfolio = DFA Balance Equity from the DFA Matrix Book]. The Vanguard 'investment grade' funds above a referred to as 'corporate' bonds below. For the last two portfolios in each set, the stock:bond mix is adjusted to match the volatility (standard deviation) and 2008 downside of the portfolio with treasuries. In all cases it resulted in lowering the stock allocation and increasing the corporate bond allocation.

Annualized return (%) / SD / 2008 return

With short-term bonds
9.4 / 13.3 / -29.6 = With Treasuries [75:25 stock:bond]
9.5 / 14.1 / -32.5 = With corporate bonds [75:25 stock:bond]
9.2 / 13.3 / -30.6 = With corporate bonds, matching volatility (SD) to same as portfolio with treasuries
9.0 / 12.9 / -29.6 = With corporate bonds, matching 2008 downside to same as portfolio with treasuries

With intermediate-term bonds
9.9 / 12.9 / -27.9 = With Treasuries [75:25 stock:bond]
9.9 / 14.2 / -32. 8 = With corporate bonds [75:25 stock:bond]
9.5 / 12.9 / -29.6 = With corporate bonds, matching volatility (SD) to same as portfolio with treasuries
9.2 / 12.2 / -27.9 = With corporate bonds, matching 2008 downside to same as portfolio with treasuries

With long-term bonds
10.5 / 12.4 / -25.7 = With Treasuries [75:25 stock:bond]
10.3 / 13.8 / -30.7 = With corporate bonds [75:25 stock:bond]
10.1 / 12.4 / -27.0 = With corporate bonds, matching volatility (SD) to same as portfolio with treasuries
10.0 / 12.0 / -25.7 = With corporate bonds, matching 2008 downside to same as portfolio with treasuries

User avatar
jeffyscott
Posts: 7328
Joined: Tue Feb 27, 2007 9:12 am
Location: Wisconsin

Re: Larry Swedroe: No Need For Corp Bonds?

Post by jeffyscott » Sun May 20, 2018 9:30 am

Why are you choosing such a high equity allocation (75%)?

Why is it assumed to be that you must have only one or the other? As just one example, could have 50% stocks, 25% treasuries, and 25% corporates...or perhaps better for individuals, who have tax advantaged accounts, would be 50% stocks, 25% 5 year CDs, 25% corporate bonds.

Looking at a chart, it appears all this advantage is probably tied to one short period of about 15 months from summer of 2008 to fall 2009, when corporates and treasuries diverged. I do understand that that is the point, that it is hoped that treasuries will go up in stock market crashes.

Image
press on, regardless - John C. Bogle

User avatar
fortyofforty
Posts: 1151
Joined: Wed Mar 31, 2010 12:33 pm

Re: Larry Swedroe: No Need For Corp Bonds?

Post by fortyofforty » Sun May 20, 2018 9:54 am

Also, what are the rebalancing characteristics? Daily, monthly, annually, or based on bands (or a combination of those factors)? Are they assumed perfect rebalancing points, or based on a set schedule? Finally, this is based on a lump sum, measured over time, without regard to periodic investments such as through a retirement plan.
"In a time of universal deceit, telling the truth becomes a revolutionary act." - George Orwell | There are many roads to doublin'. | Original Vanguard Diehard

User avatar
Robert T
Posts: 2534
Joined: Tue Feb 27, 2007 9:40 pm
Location: 1, 0.2, 0.4, 0.5
Contact:

Re: Larry Swedroe: No Need For Corp Bonds?

Post by Robert T » Sun May 20, 2018 11:25 am

jeffyscott wrote:
Sun May 20, 2018 9:30 am
Why are you choosing such a high equity allocation (75%)?
That is my equity allocation.
jeffyscott wrote:
Sun May 20, 2018 9:30 am
Why is it assumed to be that you must have only one or the other? As just one example, could have 50% stocks, 25% treasuries, and 25% corporates...or perhaps better for individuals, who have tax advantaged accounts, would be 50% stocks, 25% 5 year CDs, 25% corporate bonds.
Over the 24 year period from 1994 to 2017 no combination of corporate bonds and treasuries resulted in a higher overall portfolio Sharpe ratio than treasuries alone for either short-term, intermediate term, or long-term bonds separately or for any combination for the 75:25 stock:bond portfolio. The results seem to hold for lower equity allocations. All the data presented in my previous post are publicly available, so anyone can do the analysis for any allocation of equities they like.
jeffyscott wrote:
Sun May 20, 2018 9:30 am
Looking at a chart, it appears all this advantage is probably tied to one short period of about 15 months from summer of 2008 to fall 2009, when corporates and treasuries diverged. I do understand that that is the point, that it is hoped that treasuries will go up in stock market crashes.
From Swensen's book - U.S. treasury bonds "provide a unique form of portfolio diversification, serving as a hedge against financial accidents and unanticipated deflation. No other asset type comes close to matching the diversifying power created by long-term, non-callable, default-free, full-faith-and-credit obligations of the U.S. government.”
fortyofforty wrote:Also, what are the rebalancing characteristics? Daily, monthly, annually, or based on bands (or a combination of those factors)? Are they assumed perfect rebalancing points, or based on a set schedule? Finally, this is based on a lump sum, measured over time, without regard to periodic investments such as through a retirement plan.
Annually rebalanced and assumed lump sum plus in accumulation phase. There are many permutations you can do. As above, all the data used in my previous post are publicly available - so anyone can do any permutations they want.

The are obviously no guarantees that the past 24 year period will be similar to the next 24 years. And I understand the points about the limitations of back-tests - however 15 of the last 24 years covered by actual investment period where treasuries did even better - so not just a theoretical exercise.

This approach works for me. We each have to develop our own approach that we can stick with over the long-term.

Robert
.

User avatar
jeffyscott
Posts: 7328
Joined: Tue Feb 27, 2007 9:12 am
Location: Wisconsin

Re: Larry Swedroe: No Need For Corp Bonds?

Post by jeffyscott » Sun May 20, 2018 12:39 pm

Thanks for all the information, Robert.
press on, regardless - John C. Bogle

Theoretical
Posts: 1425
Joined: Tue Aug 19, 2014 10:09 pm

Re: Larry Swedroe: No Need For Corp Bonds?

Post by Theoretical » Sun May 20, 2018 12:42 pm

I wonder if corporates are a better diversifier for a heavy growth/quality tilt. They're definitely not for value, though I think the structure of Fallen Angel indexes means they could be uniquely suited to harvest value in bonds (a vast array of good bonds get degraded in the crisis and then spring back to life thanks to pension/institutional limits on speculative debt).

azanon
Posts: 1959
Joined: Mon Nov 07, 2011 10:34 am
Location: Little Rock, AR

Re: Larry Swedroe: No Need For Corp Bonds?

Post by azanon » Mon May 21, 2018 9:23 am

Theoretical wrote:
Sun May 20, 2018 12:42 pm
I wonder if corporates are a better diversifier for a heavy growth/quality tilt. They're definitely not for value, though I think the structure of Fallen Angel indexes means they could be uniquely suited to harvest value in bonds (a vast array of good bonds get degraded in the crisis and then spring back to life thanks to pension/institutional limits on speculative debt).
They're definitely not for value? Are you familiar with Vanguard Wellesley and VG Wellington fund and their long term track record?

User avatar
jeffyscott
Posts: 7328
Joined: Tue Feb 27, 2007 9:12 am
Location: Wisconsin

Re: Larry Swedroe: No Need For Corp Bonds?

Post by jeffyscott » Mon May 21, 2018 11:08 am

azanon wrote:
Mon May 21, 2018 9:23 am
Theoretical wrote:
Sun May 20, 2018 12:42 pm
I wonder if corporates are a better diversifier for a heavy growth/quality tilt. They're definitely not for value, though I think the structure of Fallen Angel indexes means they could be uniquely suited to harvest value in bonds (a vast array of good bonds get degraded in the crisis and then spring back to life thanks to pension/institutional limits on speculative debt).
They're definitely not for value? Are you familiar with Vanguard Wellesley and VG Wellington fund and their long term track record?
I don't necessarily agree with the premise, but those funds do have a "quality" orientation.

From M* review of Wellington: "...this almost-90-year-old fund that has long employed a conservative, income-oriented, high-quality-focused process" and for Wellesly: "...this almost-50-year-old fund that employs a disciplined, income-oriented, high-quality-focused process".
press on, regardless - John C. Bogle

azanon
Posts: 1959
Joined: Mon Nov 07, 2011 10:34 am
Location: Little Rock, AR

Re: Larry Swedroe: No Need For Corp Bonds?

Post by azanon » Mon May 21, 2018 3:26 pm

jeffyscott wrote:
Mon May 21, 2018 11:08 am
azanon wrote:
Mon May 21, 2018 9:23 am
Theoretical wrote:
Sun May 20, 2018 12:42 pm
I wonder if corporates are a better diversifier for a heavy growth/quality tilt. They're definitely not for value, though I think the structure of Fallen Angel indexes means they could be uniquely suited to harvest value in bonds (a vast array of good bonds get degraded in the crisis and then spring back to life thanks to pension/institutional limits on speculative debt).
They're definitely not for value? Are you familiar with Vanguard Wellesley and VG Wellington fund and their long term track record?
I don't necessarily agree with the premise, but those funds do have a "quality" orientation.

From M* review of Wellington: "...this almost-90-year-old fund that has long employed a conservative, income-oriented, high-quality-focused process" and for Wellesly: "...this almost-50-year-old fund that employs a disciplined, income-oriented, high-quality-focused process".
I've often found inconsistencies between what is verbally stated in a prospectus, and what style is indicated by popular measures, such as the morningstar style box. Just observing over the years, I've noticed Wellington and Wellesley fall into the large-cap value box. Anyway, when i sometimes find disagreement, I tend to side with the morningstar box. But that also isn't necessarily either/or, they might just focus on "quality" which causes them to also end up buying mostly value stock.

..............

In any event, another point that I think Wellesley/Wellington bring to light is that one sure can get some lights-out performance with the bond side being loaded up with corporates. Maybe they won't be so great going forward, but corporate bonds sure as heck have been very kind to these 2 funds.

Post Reply