Classic Bernstein 2 — Choosing Portfolio Bond Duration

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
Post Reply
User avatar
Topic Author
SimpleGift
Posts: 3869
Joined: Tue Feb 08, 2011 3:45 pm
Location: Central Oregon

Classic Bernstein 2 — Choosing Portfolio Bond Duration

Post by SimpleGift »

PREFACE: This is the second in a series of posts over the next few months highlighting the classic portfolio design and investing insights of William Bernstein from the 1990s and early 2000s. Many new Bogleheads on the Forum today have never been exposed to his early writings — and while the data sets he used may be a bit dated now, his early conceptual work and analyses can still be quite helpful to investors today, new and old alike.

Many thanks to those Bogleheads who sent me PMs with their personal :idea: moments from Mr. Bernstein’s early articles. This Classic Bernstein series is organized with the broadest topics first (asset allocation, diversification, the nature of risk, etc.) and will progress to narrower and more esoteric topics later on (factor investing, credit risk, gold, etc.) — so your patience is appreciated should your suggested personal topic not appear right away!

Previous topics in the series: 1-Asset Allocation & Time Horizon.
----------------------------------------------------------------------------------

In the last post, we saw that the ratio of stocks to bonds that one chooses is the major determinant of long-term portfolio return and risk. But once investors decide on their stock/bond mix, what duration of bonds is best?

To answer this question, Mr. Bernstein computed the historical return/risk profiles of various portfolio mixes of bonds and stocks shown in the two charts below. Both charts reflect the same analysis, but use different time periods and data sets. The chart on the left compares large company stocks combined with 30-day T-bills, 5-year and 20-year Treasuries for the 72-year period from 1926-1998. On the right, the chart shows these same bonds combined with various random stock allocations (including S&P 500 stocks, U.S. small stocks, European stocks, Japanese stocks, Pacific Rim stocks and precious metals equity) for the 27-year period from 1970-1996.

Image
Sources: William Bernstein, The Intelligent Asset Allocator (2000) and Efficient Frontier (10/97)

What did Mr. Bernstein make of these results?
  • a) For long-term investors, if one has a high allocation to stocks (greater than 50%, on the right side of the charts), it doesn’t appear to matter which of the three bond durations one chooses — your returns and risk have historically been similar. Thus, one’s bond duration becomes most important if bonds make up the larger part of your portfolio.

    b) Only with very low volatility portfolios (less than 5% standard deviation) do 30-day T-bills make sense.

    c) With portfolios of moderate volatility (6%-12% standard deviation), 5-year Treasuries have been the superior choice. Over most of its extent in this range, the 5-year curve (in green) lies above the other two curves, indicating that each degree of risk has yielded more return.
Thoughts?
Last edited by SimpleGift on Sat May 28, 2016 8:20 pm, edited 2 times in total.
User avatar
saltycaper
Posts: 2650
Joined: Thu Apr 24, 2014 8:47 pm
Location: The Tower

Re: Classic Bernstein 2 — Choosing Portfolio Bond Duration

Post by saltycaper »

More reinforcement to stick with intermediate Treasuries and avoid long-term issues.
Quod vitae sectabor iter?
lack_ey
Posts: 6701
Joined: Wed Nov 19, 2014 11:55 pm

Re: Classic Bernstein 2 — Choosing Portfolio Bond Duration

Post by lack_ey »

saltycaper wrote:More reinforcement to stick with intermediate Treasuries and avoid long-term issues.
Though of course if you update the data through the last 15-20 years of long-term Treasury outperformance, that changes things at least in magnitude.

That said, it still seems better to me for individual investors to avoid going way out in duration in nominal bonds, especially outside of very high equity allocations.

Though perhaps out of the scope of this discussion, sometimes I wonder if leveraged intermediate term is better for most allocations than just going to long term, given that the yield curve usually is more flattened out when you start going far out there.
User avatar
Topic Author
SimpleGift
Posts: 3869
Joined: Tue Feb 08, 2011 3:45 pm
Location: Central Oregon

Re: Classic Bernstein 2 — Choosing Portfolio Bond Duration

Post by SimpleGift »

lack_ey wrote:
saltycaper wrote:More reinforcement to stick with intermediate Treasuries and avoid long-term issues.
Though of course if you update the data through the last 15-20 years of long-term Treasury outperformance, that changes things at least in magnitude.
Just a note: If Bogleheads with an analytical bent are so inspired, they are to encouraged to replicate any of Mr. Bernstein’s analyses in this series — using current, up-to-date historical data sets — and share their results for discussion in these threads. Updated analysis that includes recent data would be of great value to many of these classic topics!
diyfp
Posts: 51
Joined: Wed Mar 09, 2016 2:22 am

Re: Classic Bernstein 2 — Choosing Portfolio Bond Duration

Post by diyfp »

Simplegift wrote:
lack_ey wrote:
saltycaper wrote:More reinforcement to stick with intermediate Treasuries and avoid long-term issues.
Though of course if you update the data through the last 15-20 years of long-term Treasury outperformance, that changes things at least in magnitude.
Just a note: If Bogleheads with an analytical bent are so inspired, they are to encouraged to replicate any of Mr. Bernstein’s analyses in this series — using current, up-to-date historical data sets — and share their results for discussion in these threads. Updated analysis that includes recent data would be of great value to many of these classic topics!
From Portfolio Visualizer (1972-2015).

100/0 stock/bond with stock being 'US Stock Market':
- US Stock Market : CAGR 10.15 StDev 17.93 Max Drawdown -40.61 Sharpe 0.37

80/20 stock/bond with stock being 'US Stock Market' and bond being:

- Long Term Treasuries: CAGR 10.11 StDev 14.57 Max Drawdown -32.79 Sharpe 0.41
- Intermediate Term Tr: CAGR 09.89 StDev 14.44 Max Drawdown -31.67 Sharpe 0.40
- Short Term Treasuries: CAGR 09.63 StDev 14.48 Max Drawdown -31.65 Sharpe 0.39


60/40 stock/bond with stock being 'US Stock Market' and bond being:

- Long Term Treasuries: CAGR 09.85 StDev 11.84 Max Drawdown -24.56 Sharpe 0.46
- Intermediate Term Tr: CAGR 09.46 StDev 11.14 Max Drawdown -22.14 Sharpe 0.45
- Short Term Treasuries: CAGR 08.96 StDev 11.13 Max Drawdown -22.09 Sharpe 0.41

40/60 stock/bond with stock being 'US Stock Market' and bond being:

- Long Term Treasuries: CAGR 09.41 StDev 10.25 Max Drawdown -15.89 Sharpe 0.46
- Intermediate Term Tr: CAGR 08.88 StDev 08.30 Max Drawdown -12.02 Sharpe 0.51
- Short Term Treasuries: CAGR 08.16 StDev 07.97 Max Drawdown -11.94 Sharpe 0.45

20/80 stock/bond with stock being 'US Stock Market' and bond being:

- Long Term Treasuries: CAGR 08.79 StDev 10.34 Max Drawdown -06.80 Sharpe 0.40
- Intermediate Term Tr: CAGR 08.16 StDev 06.51 Max Drawdown -03.50 Sharpe 0.53
- Short Term Treasuries: CAGR 07.23 StDev 05.37 Max Drawdown -02.06 Sharpe 0.50

0/100 stock/bond with bond being:

- Long Term Treasuries: CAGR 08.00 StDev 12.08 Max Drawdown -13.03 Sharpe 0.30
- Intermediate Term Tr: CAGR 07.30 StDev 06.70 Max Drawdown -04.33 Sharpe 0.40
- Short Term Treasuries: CAGR 06.18 StDev 04.47 Max Drawdown -00.58 Sharpe 0.39


All portfolios rebalanced annually.

Does this mean I should change my AA to 20% stock, 80% Intermediate Term Treasuries? :D
Random Walker
Posts: 4612
Joined: Fri Feb 23, 2007 8:21 pm

Re: Classic Bernstein 2 — Choosing Portfolio Bond Duration

Post by Random Walker »

Your post is consistent with something Larry has been teaching: in equity heavy portfolios, the volatility is dominated by the stocks, so one may as well take the higher yield associated with term risk. In portfolios with low equity allocation, where the point is to avoid volatility, keep duration short.

Dave
User avatar
Topic Author
SimpleGift
Posts: 3869
Joined: Tue Feb 08, 2011 3:45 pm
Location: Central Oregon

Re: Classic Bernstein 2 — Choosing Portfolio Bond Duration

Post by SimpleGift »

diyfp wrote:From Portfolio Visualizer (1972-2015).
All portfolios rebalanced annually.
Thanks so much, diyfp, for generating the data, which I’ve charted below:

Image

These look very consistent with Mr. Bernstein’s curves for the earlier time periods. With high volatility portfolios (>12% standard deviation), any bonds will do, but long bonds have had a slight edge. With low volatility portfolios (<5% standard deviation), short bonds look best. With all moderate volatility portfolios in between (from 6% to 12% standard deviation), intermediate bonds have had the greatest return per degree of risk.

This relative relationship appears to be fairly robust across different historical time periods. Thanks again!
Last edited by SimpleGift on Wed May 04, 2016 9:57 pm, edited 2 times in total.
User avatar
saltycaper
Posts: 2650
Joined: Thu Apr 24, 2014 8:47 pm
Location: The Tower

Re: Classic Bernstein 2 — Choosing Portfolio Bond Duration

Post by saltycaper »

diyfp wrote:
Simplegift wrote:
lack_ey wrote:
saltycaper wrote:More reinforcement to stick with intermediate Treasuries and avoid long-term issues.
Though of course if you update the data through the last 15-20 years of long-term Treasury outperformance, that changes things at least in magnitude.
Just a note: If Bogleheads with an analytical bent are so inspired, they are to encouraged to replicate any of Mr. Bernstein’s analyses in this series — using current, up-to-date historical data sets — and share their results for discussion in these threads. Updated analysis that includes recent data would be of great value to many of these classic topics!
From Portfolio Visualizer (1972-2015).

...

Does this mean I should change my AA to 20% stock, 80% Intermediate Term Treasuries? :D
Nice work. If you believe the next 43 years will look like the last 43 years, and if your goal is to achieve the highest Sharpe ratio, then of course you should. But, besides that not being the goal of many investors, IMHO, the next 43 years will not look like the past 43 years, particularly in the bond market. One reason the 1926-98 data is interesting is not even about the Sharpe ratio. Forget risk-adjusted return, you could have gotten the same return with less volatility by favoring the 5-year note instead of longer-maturity issues over those years, a span of time which I think is a more realistic impression of the future. I love portfolio visualizer; I just wish the data went further back in time.
Quod vitae sectabor iter?
diyfp
Posts: 51
Joined: Wed Mar 09, 2016 2:22 am

Re: Classic Bernstein 2 — Choosing Portfolio Bond Duration

Post by diyfp »

Based on this, why are most Bogleheads in the Total Bond Index instead of the Intermediate Bond Index?
User avatar
saltycaper
Posts: 2650
Joined: Thu Apr 24, 2014 8:47 pm
Location: The Tower

Re: Classic Bernstein 2 — Choosing Portfolio Bond Duration

Post by saltycaper »

diyfp wrote:Based on this, why are most Bogleheads in the Total Bond Index instead of the Intermediate Bond Index?
Many Bogleheads don't believe in backtesting and would prefer to own the entire yield curve. Also, Total Bond is older and, I'm guessing, more widely available in 401(k) plans and the like, so it's more established and has gotten more air time. Some might like the heavier weight to Treasuries. But really, I think it just sounds like it matches up better with the other two funds in the three-fund portfolio, which also have "Total" in their name. Sounds silly, but I wouldn't doubt it. I think Admiral shares of Total Bond are a bit cheaper too. It's "good enough" for most, I suppose.
Quod vitae sectabor iter?
User avatar
Munir
Posts: 2745
Joined: Mon Feb 26, 2007 4:39 pm
Location: Oregon

Re: Classic Bernstein 2 — Choosing Portfolio Bond Duration

Post by Munir »

diyfp wrote:Based on this, why are most Bogleheads in the Total Bond Index instead of the Intermediate Bond Index?
In addition to the reasons listed by others, could another reason be that the Intermediate Bond Index has a longer duration (6.5 vs. 5.7) which may be worrisome if rates go up?
diyfp
Posts: 51
Joined: Wed Mar 09, 2016 2:22 am

Re: Classic Bernstein 2 — Choosing Portfolio Bond Duration

Post by diyfp »

This may be obvious to some but why exactly are long-term treasuries so volatile?

We can see from the graph (and the data) that they're so volatile that adding stocks to a 100% LTT portfolio actually REDUCES volatility!
Valuethinker
Posts: 41173
Joined: Fri May 11, 2007 11:07 am

Re: Classic Bernstein 2 — Choosing Portfolio Bond Duration

Post by Valuethinker »

diyfp wrote:This may be obvious to some but why exactly are long-term treasuries so volatile?

We can see from the graph (and the data) that they're so volatile that adding stocks to a 100% LTT portfolio actually REDUCES volatility!
Because of the macroeconomics we have been through. Ie from very low inflation (1930s) through to high inflation (1970s) and very high interest rates (Paul Volker, 1981) and then systematic disinflation since, with bond yields racing down to catch up.

Simply by the mathematics, LT Bonds are the most sensitive to interest rate changes (the extreme is the 30 year 0 coupon stripped bond). Also (at least in the UK) it's an illiquid market-- institutions like pension funds and insurance companies buy these bonds and just sit on them, so thinly traded and therefore big price jumps.

The LT Treasury has, in fact, been more volatile than US equities. Whether that will continue is an open question-- interest rates could easily fall to 1% (I don't think that is the most likely outcome in the US, but we can no longer totally discount it given what has happened in Japan and the Eurozone, especially).
Valuethinker
Posts: 41173
Joined: Fri May 11, 2007 11:07 am

Re: Classic Bernstein 2 — Choosing Portfolio Bond Duration

Post by Valuethinker »

Munir wrote:
diyfp wrote:Based on this, why are most Bogleheads in the Total Bond Index instead of the Intermediate Bond Index?
In addition to the reasons listed by others, could another reason be that the Intermediate Bond Index has a longer duration (6.5 vs. 5.7) which may be worrisome if rates go up?
I think that's probably true, but it probably has as much to do with simplicity as anything else.

You are taking on all kinds of extra risks on the TBM vs. IT Treasury:

- default risk (corporate bonds)
- call risk (ditto)
- prepayment and extension risk in mortgage backed securities

But therefore you should be rewarded for that. The volatility of TBM is relatively small compared to say TSM let alone an international equity fund.

It's a classic case of a "good enough" solution ("satisficing") and to be honest, in investing, if you sweat the big stuff (bond-equity mix, minimizing costs, etc.) then this probably falls into the small stuff basket.
longinvest
Posts: 4433
Joined: Sat Aug 11, 2012 8:44 am

Re: Classic Bernstein 2 — Choosing Portfolio Bond Duration

Post by longinvest »

Valuethinker wrote:
Munir wrote:
diyfp wrote:Based on this, why are most Bogleheads in the Total Bond Index instead of the Intermediate Bond Index?
In addition to the reasons listed by others, could another reason be that the Intermediate Bond Index has a longer duration (6.5 vs. 5.7) which may be worrisome if rates go up?
I think that's probably true, but it probably has as much to do with simplicity as anything else.

You are taking on all kinds of extra risks on the TBM vs. IT Treasury:

- default risk (corporate bonds)
- call risk (ditto)
- prepayment and extension risk in mortgage backed securities

But therefore you should be rewarded for that. The volatility of TBM is relatively small compared to say TSM let alone an international equity fund.

It's a classic case of a "good enough" solution ("satisficing") and to be honest, in investing, if you sweat the big stuff (bond-equity mix, minimizing costs, etc.) then this probably falls into the small stuff basket.
With IT Treasury bonds, one takes concentration risk. TBM is widely diversified across maturities and types of high-quality bonds (investment-grade corporate and government).

Many Bogleheads simply don't believe any prediction about the alleged future superiority of a concentrated subset of a market and would rather harvest the returns of the entire market.
Bogleheads investment philosophy | One-ETF global balanced index portfolio | VPW
Random Walker
Posts: 4612
Joined: Fri Feb 23, 2007 8:21 pm

Re: Classic Bernstein 2 — Choosing Portfolio Bond Duration

Post by Random Walker »

"Concentration risk" is not relevant with regard to treasuries since they are considered riskless. I guess you could consider intermediate term concentrated with regard to term risk.

Dave
longinvest
Posts: 4433
Joined: Sat Aug 11, 2012 8:44 am

Re: Classic Bernstein 2 — Choosing Portfolio Bond Duration

Post by longinvest »

Random Walker wrote:"Concentration risk" is not relevant with regard to treasuries since they are considered riskless. I guess you could consider intermediate term concentrated with regard to term risk.
Dave,

Yes it is relevant. Concentration in specific maturities makes the investor vulnerable to an unfavorable shape of the yield curve where he would be buying overpriced bonds crossing a maturity threshold, and selling under-priced bonds crossing another maturity threshold.

Seen another way: If a majority of investors preferred 5-to-10 year bonds, 11-year bonds nearing 10-year maturity would spike in price, and 6-year bonds nearing 5-year maturity would drop in price, due to supply and demand.

Yes, the law of supply and demand applies even for default-less government bonds.

It is much safer not to concentrate one's portfolio in specific maturities, and accept overall market returns, whatever they happen to be.
Bogleheads investment philosophy | One-ETF global balanced index portfolio | VPW
Random Walker
Posts: 4612
Joined: Fri Feb 23, 2007 8:21 pm

Re: Classic Bernstein 2 — Choosing Portfolio Bond Duration

Post by Random Walker »

Long invest,
Agree, concentrated regarding term risk
Concentration of quality risk irrelevant since the bond is riskless

Dave
longinvest
Posts: 4433
Joined: Sat Aug 11, 2012 8:44 am

Re: Classic Bernstein 2 — Choosing Portfolio Bond Duration

Post by longinvest »

Random Walker wrote:Long invest,
Agree, concentrated regarding term risk
Concentration of quality risk irrelevant since the bond is riskless
Dave,

A government bond is not risk-less; it is default-less. :)

The law of supply and demand also applies to investment-grade corporate and government bonds. If investors perceive corporate bonds as riskier, they'll demand higher yields/coupons, which will reduce the duration of longer corporate bonds relative to government bonds of similar maturity. This can help reduce volatility, and increase returns.

Also, if a majority of investors preferred government bonds, their yields would get much lower than investment-grade corporate bonds, inflating government bond prices and exposing the investors who buy these pricey bonds to long-term low returns (or even losses, when inflation is taken into account).

Many Bogleheads believe that the safest approach is to widely diversify our investments. So, we invest in three total-market low-cost index funds (total domestic stocks, total international stocks, total bonds), the Three-Fund Portfolio. There will be more rewarding portfolios, but also less rewarding ones. We simply can't know, as of today, which portfolio will win, so we might as well embrace accepting the overall returns of these three markets.
Bogleheads investment philosophy | One-ETF global balanced index portfolio | VPW
sschullo
Posts: 2579
Joined: Sun Apr 01, 2007 8:25 am
Location: Rancho Mirage, CA
Contact:

Re: Classic Bernstein 2 — Choosing Portfolio Bond Duration

Post by sschullo »

Munir wrote:
diyfp wrote:Based on this, why are most Bogleheads in the Total Bond Index instead of the Intermediate Bond Index?
In addition to the reasons listed by others, could another reason be that the Intermediate Bond Index has a longer duration (6.5 vs. 5.7) which may be worrisome if rates go up?
I like the diversification. At my age, I have half my portfolio in this fund. The rest of my bond allocation is in iBonds, international and Wellesley.
"We have seen much more money made and kept by “ordinary people” who were temperamentally well suited for the investment process than by those who lacked this quality." Ben Graham
User avatar
Topic Author
SimpleGift
Posts: 3869
Joined: Tue Feb 08, 2011 3:45 pm
Location: Central Oregon

Re: Classic Bernstein 2 — Choosing Portfolio Bond Duration

Post by SimpleGift »

longinvest wrote:A government bond is not risk-less; it is default-less.

The law of supply and demand also applies to investment-grade corporate and government bonds. If investors perceive corporate bonds as riskier, they'll demand higher yields/coupons, which will reduce the duration of longer corporate bonds relative to government bonds of similar maturity. This can help reduce volatility, and increase returns.
There’s something to be said, I believe, for diversifying among bond types, whatever one’s bond portfolio duration target. In other words, if one’s duration target is 3 years, then something like a Short-Term Bond Index Fund provides some diversity among Treasuries, corporates, etc. The same for an intermediate duration target in the 5-6 year range, where a widely-diversified fund like a Total Bond Index suffices nicely.

The reasoning: The conventional wisdom for bond duration (that a 1% increase in interest rates will realize bond losses equal to average duration) assumes that all maturities (short, intermediate and long) and all issuers (corporate, Treasury, mortgage-backed) will be affected equally. However, in practice, the yield curve doesn't necessarily shift in a parallel manner and various issuers are impacted differently from changes in market rates (chart below).
Some investors are happy to stick with just Treasuries, but there's a case to be made for bond diversity as well.
Bill Bernstein
Posts: 592
Joined: Sat Jun 23, 2007 12:47 am

Re: Classic Bernstein 2 — Choosing Portfolio Bond Duration

Post by Bill Bernstein »

Excellent thread!

Two points:

1) The risk of long bonds isn't well captured by volatility. Consider: the real total return of the long T from mid-1941 to mid-1981 was about -65%. That's what I call risk, far worse than any 40-year real equity return.

2) Yes, adding corporates diversifies bond risk. But it also adds equity risk to the bond portfolio. If you want to add equity risk, do so on the equity side, and keep your powder dry; adding corporates dampens that powder.

Bil
lack_ey
Posts: 6701
Joined: Wed Nov 19, 2014 11:55 pm

Re: Classic Bernstein 2 — Choosing Portfolio Bond Duration

Post by lack_ey »

FWIW I tried recreating it with a different dataset and come up with a little different results than posted earlier.

This is annually rebalanced, using S&P 500 (well, the composite backfilled in the earlier years) with either T-bills or 10-year Treasuries:

Image

In any case you see the effect of the additional years in improving the 10-yr over T-bills because of (1) the excess returns and (2) the negative correlation with stocks over that time period.

Inflation was 2.9% over the full period and 3.2% through 1998.
Random Walker
Posts: 4612
Joined: Fri Feb 23, 2007 8:21 pm

Re: Classic Bernstein 2 — Choosing Portfolio Bond Duration

Post by Random Walker »

Long Invest,
I agree with you on all counts. Where we disagree perhaps is on our philosophy about purpose of bonds in a portfolio. I heavily lean towards taking all my risk on the equity side. I think it's more efficient to take the equity type risk on the equity side of the portfolio. I use bonds purely as portfolio anchor. Seems to me that you look for more return from your bonds. I'm a huge believer in market efficiency and supply/demand, so you won't get many arguments from me. I'm sure the market prices the bond spreads ruthlessly efficiently.

Dave
User avatar
saltycaper
Posts: 2650
Joined: Thu Apr 24, 2014 8:47 pm
Location: The Tower

Re: Classic Bernstein 2 — Choosing Portfolio Bond Duration

Post by saltycaper »

Bill Bernstein wrote: 1) The risk of long bonds isn't well captured by volatility. Consider: the real total return of the long T from mid-1941 to mid-1981 was about -65%. That's what I call risk, far worse than any 40-year real equity return.
A sobering and important reminder.
longinvest wrote: ... Concentration in specific maturities makes the investor vulnerable to an unfavorable shape of the yield curve where he would be buying overpriced bonds crossing a maturity threshold, and selling under-priced bonds crossing another maturity threshold.

Seen another way: If a majority of investors preferred 5-to-10 year bonds, 11-year bonds nearing 10-year maturity would spike in price, and 6-year bonds nearing 5-year maturity would drop in price, due to supply and demand.

Yes, the law of supply and demand applies even for default-less government bonds.

It is much safer not to concentrate one's portfolio in specific maturities, and accept overall market returns, whatever they happen to be.
Price spikes solely for intermediate maturities have not endured over the long term. It sounds like you're describing an inverted yield curve, which happens from time to time--usually before/during a recession--but historically it's been "normal" longer than not. Here's a couple charts where the curve would be inverted when the chart is in the negative.

Image

Image

I wish the 10 minus 2 went back further, as I believe it is more relevant comparison for Total Bond, which I don't think holds much less than 1 or 2 years in maturity. The FFR comparison with the 5-year (all I could find) likely exaggerates what one would experience with the fund.

I like to avoid the inflation-risk and additional volatility of long-term bonds, and I prefer to hold cash and CDs instead of short-term Treasuries. That leaves me with the "meat" of the yield curve to invest in. If someone prefers to own the entire curve, I understand.
Quod vitae sectabor iter?
User avatar
kwan2
Posts: 384
Joined: Thu Jun 14, 2012 9:13 pm

Re: Classic Bernstein 2 — Choosing Portfolio Bond Duration

Post by kwan2 »

so, do many BH, take Mr.Bernstein's advice, and avoid corporate bonds, and not buy TBM or IT Bond Index, because of their containing corporate bonds?

and/or tilt by say buying 50% TBM and 50% UST IT ?

seems to me, would be uncommon in practice for the majority of BH investors, but I've no idea really
“The history of Paris teaches us that beauty is a by-product of danger, that liberty is at best a consequence of neglect, that wisdom is entwined with decay."
lack_ey
Posts: 6701
Joined: Wed Nov 19, 2014 11:55 pm

Re: Classic Bernstein 2 — Choosing Portfolio Bond Duration

Post by lack_ey »

kwan2 wrote:so, do many BH, take Mr.Bernstein's advice, and avoid corporate bonds, and not buy TBM or IT Bond Index, because of their containing corporate bonds?

and/or tilt by say buying 50% TBM and 50% UST IT ?

seems to me, would be uncommon in practice for the majority of BH investors, but I've no idea really
These days if you're going to do anything remotely fancy in fixed income, you should largely be using CDs instead of Treasuries anyway.

I think some nontrivial percentage here really do avoid corporate bonds, though. Some avoid long-term bonds.
Random Walker
Posts: 4612
Joined: Fri Feb 23, 2007 8:21 pm

Re: Classic Bernstein 2 — Choosing Portfolio Bond Duration

Post by Random Walker »

I think it makes great sense to avoid corporates. As one goes down the quality spectrum, the bonds gain equity like characteristics. It's simply more efficient to take one's risk on the equity side.

Dave
User avatar
tludwig23
Posts: 1665
Joined: Thu Dec 30, 2010 3:27 pm
Location: 48deg46"23"N 122deg28'21"W

Re: Classic Bernstein 2 — Choosing Portfolio Bond Duration

Post by tludwig23 »

Random Walker wrote:I think it makes great sense to avoid corporates. As one goes down the quality spectrum, the bonds gain equity like characteristics. It's simply more efficient to take one's risk on the equity side.
Who says one needs to go down the quality spectrum to own corporates? In the decade from 2005-2014, most years (6 of 10) the default rate for investment grade corporate bonds was 0.00%. The highest year 2008 the default rate spiked at a whopping 0.42%. And that's for all investment grade corporates. If one limits to only AAA rated corporates, the default rates are minuscule.
That's what I do: I drink, and I know things. --Tyrion Lannister
Bill Bernstein
Posts: 592
Joined: Sat Jun 23, 2007 12:47 am

Re: Classic Bernstein 2 — Choosing Portfolio Bond Duration

Post by Bill Bernstein »

Of course high-rated corporate bonds never default; they first become lower-rated bonds before they do. Remember, at one time Enron, Wang Labs, GM, and Penn Central bonds were all highly rated.

Those who think that even highly rated corporates will hold up when the sky turns stormy should take another look at how they did during September and October of 2008, particularly at longer durations; they're "money," until you really need it.


Bill
kenner
Posts: 3129
Joined: Sat Mar 01, 2008 8:45 am

Re: Classic Bernstein 2 — Choosing Portfolio Bond Duration

Post by kenner »

Dr. Bernstein,

It is truly an honor when you contribute to our forum.

I doubt that anyone else has the depth of knowledge that you possess regarding the history of the evolution of human economic development.

Thank you.
Norton750
Posts: 119
Joined: Sat Mar 17, 2007 11:25 am

Re: Classic Bernstein 2 — Choosing Portfolio Bond Duration

Post by Norton750 »

Here's one BH who owns only Treasury funds (Short, Intermediate and TIPs) - purposely done for the reasons covered earlier in this thread:
- I use bonds strictly as an an anchor for the equity side of the portfolio
- If I want to take more risk in search of more return, I'll do so on the equity side of the portfolio
- It's only one set of data, but look at the performance of IT Treasuries vs TBM in the depths of the Great Recession. Just as things were looking the darkest, the Treasuries had gained as much as 15% since the start of the Recession and they outperformed TBM by a margin of between 3% and 7% between 9/2008 to 3/2009. Precisely what I'm looking for from my bonds.

And a couple of specific thoughts: I prefer to put my "safe" money in the hands of the organization that controls the printing presses. And I really dislike the behavior of MBS (20% of TBM) in reaction to interest rate changes - their duration increases when rates rise (people hold on to their mortgages) and decreases when rates drop (people refinance their mortgages). Just the opposite of what a bondholder wants.

I will say that I have been exchanging some ST Treasuries for CD's in the last year as CD rates from 2.25% to 2.40% have been offered for 2-year and 3-year CDs. FDIC or NCUA insurance is also fine by me.
User avatar
tludwig23
Posts: 1665
Joined: Thu Dec 30, 2010 3:27 pm
Location: 48deg46"23"N 122deg28'21"W

Re: Classic Bernstein 2 — Choosing Portfolio Bond Duration

Post by tludwig23 »

Maybe my viewpoint is tainted by being in the accumulation phase, but I see the 2008 financial crisis as a big yawn wrt corporate bonds. True, intermediate investment grade corporates fell about 15% between Aug and Oct, 2008, but within a few months they were back up to where they were before the crisis. A small blip.

I personally own VCIT. If it falls 20% I'll rebalance and buy more. As I get closer to retirement I plan to decrease my corporate bond exposure and increase my TIPS exposure. At that point safety will be my primary goal. As long as I have years of accumulation left, it's not.
That's what I do: I drink, and I know things. --Tyrion Lannister
User avatar
Earl Lemongrab
Posts: 7270
Joined: Tue Jun 10, 2014 1:14 am

Re: Classic Bernstein 2 — Choosing Portfolio Bond Duration

Post by Earl Lemongrab »

There are theories, appropriate for this forum, then there are realities and practicalities. My portfolio has about 45% in a 401(k) (deferred) and 17% in Roth IRA. My asset allocation is 40% bonds. My 401(k) does not offer a government bond fund. It has a low-cost Barclay's Aggregate Index fund, a more expensive managed International Bond fund, and a stable value fund. To avoid corporates I would need to use all of my Roth and then keep either taxable funds or muni funds in taxable. That's not a practical decision for something that I really don't think is much of a concern.

Earl
User avatar
Munir
Posts: 2745
Joined: Mon Feb 26, 2007 4:39 pm
Location: Oregon

Re: Classic Bernstein 2 — Choosing Portfolio Bond Duration

Post by Munir »

Random Walker wrote:I think it makes great sense to avoid corporates. As one goes down the quality spectrum, the bonds gain equity like characteristics. It's simply more efficient to take one's risk on the equity side.

Dave
I respectfully disagree. Taking risk with corporate bonds does not equate to taking the far greater risk with equities. Default rates for high quality corporates have been overblown. A transient drop in the NAV of corporate bond funds for a few months in 2008 should not determine the longterm course of holdings in fixed income unless a person panics or wants to immediately rebalance. As a retiree with a 70% position in fixed income that includes a safety cushion of cash, I prefer a generous portion of my bond portfolio to be in corporates. Individual circumstances vary and one rule does not apply to all. For me, a mix of the Total Bond Market Bond Fund and the Intermediate Investment Grade Fund balanced by a comfortable cash position has proven ideal
Random Walker
Posts: 4612
Joined: Fri Feb 23, 2007 8:21 pm

Re: Classic Bernstein 2 — Choosing Portfolio Bond Duration

Post by Random Walker »

I did not say that corporate bonds "equate" to equity risk. I'm not a financial pro, but I'm confident that there is a bit of equity type risk in corporates that increases as one goes down the quality spectrum. I think that specific loadings from the 5 factor model can be attributed to corporates.

Dave
Average
Posts: 57
Joined: Fri Sep 07, 2012 8:05 am

Re: Classic Bernstein 2 — Choosing Portfolio Bond Duration

Post by Average »

Munir wrote:I respectfully disagree. Taking risk with corporate bonds does not equate to taking the far greater risk with equities. Default rates for high quality corporates have been overblown. A transient drop in the NAV of corporate bond funds for a few months in 2008 should not determine the longterm course of holdings in fixed income unless a person panics or wants to immediately rebalance.
FWIW if an investment grade corporate is downgraded and the fund mandate requires selling it, the transient drop for that issue becomes a locked-in capital loss for the investment grade fund. The bond may never default but it was detrimental to the fund NAV.
User avatar
Munir
Posts: 2745
Joined: Mon Feb 26, 2007 4:39 pm
Location: Oregon

Re: Classic Bernstein 2 — Choosing Portfolio Bond Duration

Post by Munir »

Average wrote:
Munir wrote:I respectfully disagree. Taking risk with corporate bonds does not equate to taking the far greater risk with equities. Default rates for high quality corporates have been overblown. A transient drop in the NAV of corporate bond funds for a few months in 2008 should not determine the longterm course of holdings in fixed income unless a person panics or wants to immediately rebalance.
FWIW if an investment grade corporate is downgraded and the fund mandate requires selling it, the transient drop for that issue becomes a locked-in capital loss for the investment grade fund. The bond may never default but it was detrimental to the fund NAV.
How often does that happen for it to have a significant impact on a diversified bond fund? If that situation happens in an equity fund, what would a manager do? Is it any different for a bond fund?

I also trust the Vanguard managers in how they run their funds.
Average
Posts: 57
Joined: Fri Sep 07, 2012 8:05 am

Re: Classic Bernstein 2 — Choosing Portfolio Bond Duration

Post by Average »

Munir wrote:If that situation happens in an equity fund, what would a manager do? Is it any different for a bond fund?

I also trust the Vanguard managers in how they run their funds.
Indexed equity funds will continue to hold what the index holds (avoiding discussion of sampling v full replication - out of scope of conversation) what an active equity fund manager would do is any ones guess.

Indexed bond funds are going to largely track the index so the construction of the index determines quality requirements. For active VG bond funds they contain mandates that tie the managers hands regarding the quality they hold, they only have limited flexibility. Here's a quote from VG regarding the holdings in intermediate term investment grade (bold mine)

"The fund invests in a variety of high-quality and, to a lesser extent, medium-quality fixed income securities, at least 80% of which will be short- and intermediate-term investment-grade securities. High-quality fixed income securities are those rated the equivalent of A3 or better by Moody’s Investors Service, Inc., or another independent rating agency; medium-quality fixed income securities are those rated the equivalent of Baa1, Baa2, or Baa3 by Moody’s, or another independent rating agency. (Investment-grade fixed income securities are those rated the equivalent of Baa3 and above by Moody’s.) The fund is expected to maintain a dollar-weighted average maturity of 5 to 10 years".

As far as magnitude of downgrades for corporates, attached is a paper (which although of a somewhat different topic) happens to contain the corporate credit migrations across the last two recessions. http://www3.prudential.com/fi/pdf/prude ... s-0715.pdf
Average
Posts: 57
Joined: Fri Sep 07, 2012 8:05 am

Re: Classic Bernstein 2 — Choosing Portfolio Bond Duration

Post by Average »

Average wrote:
Munir wrote:If that situation happens in an equity fund, what would a manager do? Is it any different for a bond fund?
Indexed equity funds will continue to hold what the index holds (avoiding discussion of sampling v full replication - out of scope of conversation) what an active equity fund manager would do is any ones guess.
IOW analysts equity downgrades are unimportant; credit rating agencies debt downgrades (especially at the IG threshold) are very important.
User avatar
Munir
Posts: 2745
Joined: Mon Feb 26, 2007 4:39 pm
Location: Oregon

Re: Classic Bernstein 2 — Choosing Portfolio Bond Duration

Post by Munir »

Average wrote:
Average wrote:
Munir wrote:If that situation happens in an equity fund, what would a manager do? Is it any different for a bond fund?
Indexed equity funds will continue to hold what the index holds (avoiding discussion of sampling v full replication - out of scope of conversation) what an active equity fund manager would do is any ones guess.
IOW analysts equity downgrades are unimportant; credit rating agencies debt downgrades (especially at the IG threshold) are very important.
Thank you for the follow-up.

A couple of questions occur to me. What about non-indexed equity funds? Does the analysts' downgrading of their holdings affect their NAV?

The second question is whether this discussion has some real life examples where the analysts' downgrading of the Vanguard Investment Grade Fund holdings resulted in a loss of NAV? I assume the transient drop of VFIDX's NAV in 2008 was not due to downgrading because it only lasted for a few months (the US Treasuries dropped just as much in 2009 for different reasons). So is the issue more theoretical as far as this specific fund is concerned? I don't know, and somehow it seems to me to be one of many risks (which are minor or rare) one takes to balance desired superior returns.
pascalwager
Posts: 1868
Joined: Mon Oct 31, 2011 8:36 pm

Re: Classic Bernstein 2 — Choosing Portfolio Bond Duration

Post by pascalwager »

Norton750 wrote:Here's one BH who owns only Treasury funds (Short, Intermediate and TIPs) - purposely done for the reasons covered earlier in this thread:
- I use bonds strictly as an an anchor for the equity side of the portfolio
- If I want to take more risk in search of more return, I'll do so on the equity side of the portfolio
- It's only one set of data, but look at the performance of IT Treasuries vs TBM in the depths of the Great Recession. Just as things were looking the darkest, the Treasuries had gained as much as 15% since the start of the Recession and they outperformed TBM by a margin of between 3% and 7% between 9/2008 to 3/2009. Precisely what I'm looking for from my bonds.

And a couple of specific thoughts: I prefer to put my "safe" money in the hands of the organization that controls the printing presses. And I really dislike the behavior of MBS (20% of TBM) in reaction to interest rate changes - their duration increases when rates rise (people hold on to their mortgages) and decreases when rates drop (people refinance their mortgages). Just the opposite of what a bondholder wants.

I will say that I have been exchanging some ST Treasuries for CD's in the last year as CD rates from 2.25% to 2.40% have been offered for 2-year and 3-year CDs. FDIC or NCUA insurance is also fine by me.
Me too.

I use two VG Treasury bond funds (short and intermediate) together to strike a balance between inflation risk and reinvestment risk: 4 to 5 years maturity as per Larry Swedroe. (He recommends individual bonds, but I much prefer the convenience/flexibility of funds.)
User avatar
LadyGeek
Site Admin
Posts: 66477
Joined: Sat Dec 20, 2008 5:34 pm
Location: Philadelphia
Contact:

Re: Classic Bernstein 2 — Choosing Portfolio Bond Duration

Post by LadyGeek »

A complete list of Simplegift's series is now in the wiki: Classic Bernstein
Wiki To some, the glass is half full. To others, the glass is half empty. To an engineer, it's twice the size it needs to be.
Post Reply