Current Relevance of Short Term Bonds - Bill Bernstein Q

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
Post Reply
Topic Author
ajc3388
Posts: 24
Joined: Wed Oct 20, 2010 2:53 am

Current Relevance of Short Term Bonds - Bill Bernstein Q

Post by ajc3388 »

I'm a big fan of Bill Bernstein's works and try to read through most of what he puts out. He has explicitly stated that risk is to be taken with the risky assets, and the bond portion recommended to be a mix of short term bonds + cash. But how relevant is a short term bond allocation in the current environment given low interest rates, especially if they must be placed in a taxable account?

Vanguard's short term bond index fund (admiral) VBIRX has an SEC yield of 0.92% as of yesterday 8/5/14. After tax returns, assuming taxed at 35% ordinary income, is then approximately 0.92%*(100%-35%) = 0.60%. The expense ratio on this is 10 bps, so your after tax, after fee return becomes 0.60%-0.10% = 0.50%. So at the end of the day, you're receiving 0.50% return in order to take the 2.7 years of duration risk. In other words, you lose money for all short term rate rises of 0.18% or greater. This seems like a fairly misaligned risk/return profile for something that's supposed to protect your principal. To be clear, I'm not advocating timing the short term bond market for maximizing the 1-2% return. I'm coming from a standpoint of preserving cash for when it's needed most (rebalancing when risk markets fall). It seems like an X% cash rather than X% short term bonds makes more sense (or at the very least, there seems to be almost no difference).
dbr
Posts: 46181
Joined: Sun Mar 04, 2007 8:50 am

Re: Current Relevance of Short Term Bonds - Bill Bernstein Q

Post by dbr »

Holding bonds dilutes the impact of volatility in the stocks you hold. In general that dilution is a little better if the bonds have less volatility than if they have more, hence short term bonds. Return does not come into it except when you look at the cost to the portfolio in return. In that case you are also diluting the return. Using lower return bonds dilutes return a little more than using higher return bonds. To evaluate the question overall one would have to calculate the risk and return of the portfolio as a whole as a function first of stock/bond ratio and then secondly as a function of volatility of the bonds and the stock/bond correlation of returns. Then you have to find out if using somewhat higher return, higher risk bonds is better or worse than using safer bonds but a little more allocation to stocks. I bet bond duration within reasonable bounds does not matter very much on average.

I actually don't know how that answer comes out and how much it changes as average returns, volatility of returns, and correlation of returns moves around. Mr. B. didn't give the details behind the recommendation, so it isn't clear how robust that advice really is and the rationale behind it. My guess is that as long as one takes that as a recommendation not to stretch for yield in bonds by going to long bonds or to high yield bonds, then it makes sense. I doubt that it matters if the choice is between high quality short bonds and high quality intermediate bonds in allocations that are about 50/50.

There can be a different analysis in the case of a 90% stock holder perhaps thinking of holding long bonds, or a different analysis talking about things like the "Larry" portfolio with high risk stocks and increased bond allocations, and so on.
User avatar
nisiprius
Advisory Board
Posts: 52216
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: Current Relevance of Short Term Bonds - Bill Bernstein Q

Post by nisiprius »

(Shrug) Why say more than "at the moment, the difference between 'short term bonds' and 'cash' is less than usual," particularly if you are holding your cash in a well-chosen FDIC-insured retail bank account, rather than a money market mutual fund.

How much risk to take within the low-risk part of the portfolio is a matter of taste, and it is particularly UNimportant if you have any meaningful stock allocation.

Just a back of the envelope calculation. Let's say you are 50% stocks, 50% bonds and that stocks and bonds have zero correlation. Well, then, using Morningstar's 15 year standard deviation numbers for Total Stock and Total Bond, then I THINK (mavens please correct if necessary) that the standard deviation of the mix is the weighted Cartesian sum of the parts...

15-year standard deviation
Total Stock, 15.87
Total Bond, 3.50
Prime Money Market, 0.62

50/50 mix of Total Stock and Total Bond, sqrt(50% x 15.87^2 + 50% x 3.5^2) = 11.41
50/50 mix of Total Stock and Prime Money Market, sqrt(50% x 15.87^2 + 50% x 0.62^2) = 11.23

I say the difference in "risk" between 11.41 and 11.23 is not even going to be noticeable, so you can pick whatever you like anywhere from pure cash to Total Bond and it won't matter much. Follow your instincts and preferences.
Last edited by nisiprius on Wed Aug 06, 2014 2:18 pm, 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.
User avatar
CyberBob
Posts: 3387
Joined: Tue Feb 20, 2007 1:53 pm

Re: Current Relevance of Short Term Bonds - Bill Bernstein Q

Post by CyberBob »

Morningstar interview; Bill Bernstein and fixed income:
Bernstein: Well, first of all, there's a lot of concern about fixed income as you've already alluded to. People are worried that the returns are going to be low, and I think that's almost a mathematical certainty. If yields stay where they are you're going to get a very low yield; that's the best-case scenario. If yields rise from here, then you are going to achieve probably negative returns with any duration at all. Yields on Treasuries have fallen over the past 30 years from the midteens down to 0%, 1%, or 2%; they can't fall another 14% from here. And I think unfortunately people are expecting that to happen. So you have to back up and ask yourself, what is the purpose of your fixed-income assets. Well, they're for emergency needs. They're to buy stocks when they are cheap, so you can sleep at night. And they're to buy that corner lot from your impecunious neighbor who suddenly has a need of cash. It's not to achieve a return.

And so with that in mind you want your safe assets to be as safe as you can, and you should be investing in four things, which are Treasuries at the short end, and certificates of deposit and money markets, and then a clothes pin for your nose to be able to deal with those very low yields that you're going to be getting.
http://www.morningstar.com/cover/videoc ... ?id=557820

Bob
rj49
Posts: 630
Joined: Tue Feb 22, 2011 11:22 pm

Re: Current Relevance of Short Term Bonds - Bill Bernstein Q

Post by rj49 »

A safe alternative that would satisfy Dr. BIll's criteria would have been investing in 7-year Penfed CDs starting in 2007. It would have brought a 6.25% compounded yield from 2007-20014, and it could have been reinvested in January for another 7 years at 3%. No clothespin needed (except at tax time). Zero volatility, with a put option in case of rising interest rates.

Notice that Bill didn't include dividend-paying stocks, REITs, high-yield bonds, unconstrained bond funds, or any other risky source of yield currently being flocked to. That low-risk approach also makes it difficult to recommend Target Date and other balanced funds for those in retirement, because of the longer duration of most of the bond components and and the possibility of bond losses being locked in as you're forced to withdraw from both stocks and bonds at the same time for income needs.
dbr
Posts: 46181
Joined: Sun Mar 04, 2007 8:50 am

Re: Current Relevance of Short Term Bonds - Bill Bernstein Q

Post by dbr »

Bill B.:

"So you have to back up and ask yourself, what is the purpose of your fixed-income assets. Well, they're for emergency needs. They're to buy stocks when they are cheap, so you can sleep at night. And they're to buy that corner lot from your impecunious neighbor who suddenly has a need of cash. It's not to achieve a return."

Possibly excepting the last sentence, this statement is beyond bizarre. It's no wonder people get so confused trying to invest according to this, that, or another snippet from Mr. Benrstein, or Mr. Bogle, or somebody else that we know is a thoughtful adviser to investors until we run across things like the above. Well, ok, the "sleep at night" part is right if it means to refer to the essence of the question, which is to control the risk of the portfolio. But the statement is associated as a dependent on buying stocks when they are cheap, which is not what helps you sleep at night. Then again, soundbytes are not usually well formulated grammatically and therefore end up being very confusing.
longview
Posts: 385
Joined: Wed Mar 19, 2008 5:26 am

Re: Current Relevance of Short Term Bonds - Bill Bernstein Q

Post by longview »

dbr wrote: Well, ok, the "sleep at night" part is right if it means to refer to the essence of the question, which is to control the risk of the portfolio. But the statement is associated as a dependent on buying stocks when they are cheap, which is not what helps you sleep at night..
It is a bit, for me anyway. The knowledge that you have some dry powder for the next stock market implosion helps me sleep at night -- because being able to invest "cheaply" at that time theoretically allows me to recover faster. There is a feeling of control in hoping the market crashes so you can invest cheaply (better for the long run), and hoping that the market grows so that you get capital appreciation (better for the short run).
(To color my comments: my situation is ER trying to make a large portfolio that is 99% taxable last 45 years)
dbr
Posts: 46181
Joined: Sun Mar 04, 2007 8:50 am

Re: Current Relevance of Short Term Bonds - Bill Bernstein Q

Post by dbr »

longview wrote:
dbr wrote: Well, ok, the "sleep at night" part is right if it means to refer to the essence of the question, which is to control the risk of the portfolio. But the statement is associated as a dependent on buying stocks when they are cheap, which is not what helps you sleep at night..
It is a bit, for me anyway. The knowledge that you have some dry powder for the next stock market implosion helps me sleep at night -- because being able to invest "cheaply" at that time theoretically allows me to recover faster. There is a feeling of control in hoping the market crashes so you can invest cheaply (better for the long run), and hoping that the market grows so that you get capital appreciation (better for the short run).
The data would show that the expected return from holding more stocks is greater than the expected return of holding less in stocks, no matter what rebalancing scheme you use. In short you would be better off holding your "dry powder" in stocks rather than in bonds.

Now if you are advocating that you can change your asset allocation to hold a higher allocation to stocks when they have gone down relative to some arbitrary point and a smaller allocation to stocks when they have gone up relative to some arbitrary point, then it is going to take some proving as to exactly how to do that.
AviN
Posts: 479
Joined: Mon Mar 10, 2014 8:14 am

Re: Current Relevance of Short Term Bonds - Bill Bernstein Q

Post by AviN »

nisiprius,

I'm not a maven, but I believe the asset weights need to be squared too. So the numbers should be:

50/50 mix of Total Stock and Total Bond, sqrt(0.5^2 * 15.87^2 + 0.5^2 * 3.50^2) = 8.13
50/50 mix of Total Stock and Prime Money Market, sqrt(0.5^2 * 15.87^2 + 0.5^2 * 0.62^2) = 7.94
nisiprius wrote:(Shrug) Why say more than "at the moment, the difference between 'short term bonds' and 'cash' is less than usual," particularly if you are holding your cash in a well-chosen FDIC-insured retail bank account, rather than a money market mutual fund.

How much risk to take within the low-risk part of the portfolio is a matter of taste, and it is particularly UNimportant if you have any meaningful stock allocation.

Just a back of the envelope calculation. Let's say you are 50% stocks, 50% bonds and that stocks and bonds have zero correlation. Well, then, using Morningstar's 15 year standard deviation numbers for Total Stock and Total Bond, then I THINK (mavens please correct if necessary) that the standard deviation of the mix is the weighted Cartesian sum of the parts...

15-year standard deviation
Total Stock, 15.87
Total Bond, 3.50
Prime Money Market, 0.62

50/50 mix of Total Stock and Total Bond, sqrt(50% x 15.87^2 + 50% x 3.5^2) = 11.41
50/50 mix of Total Stock and Prime Money Market, sqrt(50% x 15.87^2 + 50% x 0.62^2) = 11.23

I say the difference in "risk" between 11.41 and 11.23 is not even going to be noticeable, so you can pick whatever you like anywhere from pure cash to Total Bond and it won't matter much. Follow your instincts and preferences.
pascalwager
Posts: 2327
Joined: Mon Oct 31, 2011 8:36 pm

Re: Current Relevance of Short Term Bonds - Bill Bernstein Q

Post by pascalwager »

rj49 wrote:A safe alternative that would satisfy Dr. BIll's criteria would have been investing in 7-year Penfed CDs starting in 2007. It would have brought a 6.25% compounded yield from 2007-20014, and it could have been reinvested in January for another 7 years at 3%. No clothespin needed (except at tax time). Zero volatility, with a put option in case of rising interest rates.

Notice that Bill didn't include dividend-paying stocks, REITs, high-yield bonds, unconstrained bond funds, or any other risky source of yield currently being flocked to. That low-risk approach also makes it difficult to recommend Target Date and other balanced funds for those in retirement, because of the longer duration of most of the bond components and and the possibility of bond losses being locked in as you're forced to withdraw from both stocks and bonds at the same time for income needs.
I don't know about 7 years. He does allow for shopping around at internet banks for maximum yield rather than just using VBS, but he specifies a 5-year CD ladder to achieve an overall maturity of about 2.5 years (new book).
VT 60% / VFSUX 20% / TIPS 20%
AviN
Posts: 479
Joined: Mon Mar 10, 2014 8:14 am

Re: Current Relevance of Short Term Bonds - Bill Bernstein Q

Post by AviN »

Even a 7 year CD has low duration risk if you can break it for a small fee. You need to watch out for changes your CD's terms though and be ready to break the CD if your bank makes unfavorable changes regarding withdrawals.

I just opened a few 5-year Roth IRA CD's with Synchrony Bank.
pascalwager wrote:
rj49 wrote:A safe alternative that would satisfy Dr. BIll's criteria would have been investing in 7-year Penfed CDs starting in 2007. It would have brought a 6.25% compounded yield from 2007-20014, and it could have been reinvested in January for another 7 years at 3%. No clothespin needed (except at tax time). Zero volatility, with a put option in case of rising interest rates.

Notice that Bill didn't include dividend-paying stocks, REITs, high-yield bonds, unconstrained bond funds, or any other risky source of yield currently being flocked to. That low-risk approach also makes it difficult to recommend Target Date and other balanced funds for those in retirement, because of the longer duration of most of the bond components and and the possibility of bond losses being locked in as you're forced to withdraw from both stocks and bonds at the same time for income needs.
I don't know about 7 years. He does allow for shopping around at internet banks for maximum yield rather than just using VBS, but he specifies a 5-year CD ladder to achieve an overall maturity of about 2.5 years (new book).
pascalwager
Posts: 2327
Joined: Mon Oct 31, 2011 8:36 pm

Re: Current Relevance of Short Term Bonds - Bill Bernstein Q

Post by pascalwager »

AviN wrote:Even a 7 year CD has low duration risk if you can break it for a small fee. You need to watch out for changes your CD's terms though and be ready to break the CD if your bank makes unfavorable changes regarding withdrawals.

I just opened a few 5-year Roth IRA CD's with Synchrony Bank.
pascalwager wrote:
rj49 wrote:A safe alternative that would satisfy Dr. BIll's criteria would have been investing in 7-year Penfed CDs starting in 2007. It would have brought a 6.25% compounded yield from 2007-20014, and it could have been reinvested in January for another 7 years at 3%. No clothespin needed (except at tax time). Zero volatility, with a put option in case of rising interest rates.

Notice that Bill didn't include dividend-paying stocks, REITs, high-yield bonds, unconstrained bond funds, or any other risky source of yield currently being flocked to. That low-risk approach also makes it difficult to recommend Target Date and other balanced funds for those in retirement, because of the longer duration of most of the bond components and and the possibility of bond losses being locked in as you're forced to withdraw from both stocks and bonds at the same time for income needs.
I don't know about 7 years. He does allow for shopping around at internet banks for maximum yield rather than just using VBS, but he specifies a 5-year CD ladder to achieve an overall maturity of about 2.5 years (new book).
Then I guess Bill would only require a ladder if you were purchasing through a brokerage as opposed to direct purchase from a bank.
VT 60% / VFSUX 20% / TIPS 20%
Post Reply