Asset allocation within bonds?

[grakster]

As a new Boglehead, I owe many of you thanks since I've been following the forum for several months now. Not only have I been able to find answers to many of my questions, but you've also given me the courage to take charge of my retirement funds and make some important changes. So thank you all, you provide a wonderful service and have helped me enormously.

One of the things I've learned is that the bond allocation for my own retirement investments is too low, and I need to increase it. But with the low rates and inflation risk, it doesn't seem to be a good time to purchase a big chunk of bonds. (I know red flags may be going up at this point, but at worst I think I'm trying to secular market time.) My real question is whether bonds should be viewed fundamentally differently with the extreme low-rate environment we have now.

Asset allocation is based on bonds having certain characteristics and providing specific functions to a portfolio, such as low risk, low volatility, and yield that at least keeps ahead of inflation. With low rates and the possibility of inflation, bonds now have more risk than they used to, and current yields don't seem to provide a sufficient risk premium. Bonds today don't provide the same safety that they used to. Another function of bonds is to provide stability to a portfolio during a stock downturn, often increasing in value as money moves to their relative safety. But if a stock bear market occurs at the same time as inflation increases, then bonds will not necessarily provide the same function in preserving portfolio value.

I've also heard noise reports that many entities have stepped up bond issuance to lock in low rates for long term debt, hoping to earn more return by investing the borrowed money when the economy recovers further. If this is happening, total bond market funds would then be increasing their overall level of inflation risk.

If any of this is true (and I'm sure you'll tell me if I'm wrong), my question is if there's a better way to allocate assets within bonds that reduces some of this risk and gets you more of what you really buy bonds for. For example, rather than a 35% bond allocation in a total market index, could one have a 30% bond allocation divvied up among cash, short and intermediate term bonds that would provide the portfolio with the same level of "ballast" in a low-rate environment as the 35% total market index would in a time of normal rates? What about types of bonds that might have less relative risk in the current low-rate environment - corporates, mortgage, foreign, emerging markets?

Or with at least a 10-year time horizon, is the wisest thing not to worry, and stick with rebalancing into a total market index fund?

[Call_Me_Op]

Welcome to the board!

There is no definitive answer to your question. You will generally find that there are two camps on this board: those who will tell you to "stay the course" with an intermediate bond fund, and those who will tell you to stick with very short durations with very high-quality bonds, short-term brokerage CD's, bank CD's with low withdrawal penalty, and savings bonds (if taxable money). Nobody knows today which approach is better. I like the latter approach because it seems that the reward for extending duration to 5 years is small and the risk is potentially large.

[wesleymouch]

During the 1970s when we had a dramatic increase in interest rates intermediate term bonds did not suffer a nominal loss. They did ,however, (like stocks during that era) deliver a negative real return. The best performing asset was gold; and cash did pretty well with only small real negative per year loss for the decade (better than stocks and bonds). Because of this I use the Permanent Portfolio strategy and continue to hold bonds and figure that my gold will bail me out if we have a repeat of the 1970s. The one change I have made is to hold lots of the bonds I do own in non tax advantaged accounts so I can tax loss harvest if necessary.

[DaveS]

My reaction to the historically low rate environment we are in is to gradually reduce my average duration. In part this is easy because I have some of my bond money in a ladder. So I can just reduce the duration of the replacement bonds when one pays off. I still think bonds have an important role in a portfolio in the form of stability when other things crash. As a result I have not changed my stock/bond allocation. In other posts I have been telling people to try for a average duration of around 4. That can be done by adding a short term bond fund to the usual mix of total bond and TIPS.

For the guy who said he likes the Perm. Port. I lived through the late 70's when long term rates went from about 6% to 12% in a couple of years. That meant a person with 25% in long term bonds got devastated. Do the math. A long term bond with a duration of 20 declines by 20% for each 1% increase in rates. Is that the kind of loss you want to tax harvest? With rates reaching the point of mathematically impossibility to go lower you don't want to have long bonds. Dave

[Call_Me_Op]

For the guy who said he likes the Perm. Port. I lived through the late 70's when long term rates went from about 6% to 12% in a couple of years. That meant a person with 25% in long term bonds got devastated. Do the math. A long term bond with a duration of 20 declines by 20% for each 1% increase in rates. Is that the kind of loss you want to tax harvest? With rates reaching the point of mathematically impossibility to go lower you don't want to have long bonds. Dave

The Permanent Portfolio did very well in the 1970's, despite holding long-term bonds, due to the statospheric rise of gold during the same period. That's the whole point of strong diversification - the basis of that portfolio.

[nisiprius]

I am a lazy, sloppy, "satisficing" investor. I have fairly little interest in the nuances of allocation within bonds, for these reasons:

• Even with only about 1/3 stocks, the volatility of stocks is so much higher than bonds that what you do within the bond allocation just seems unimportant. Who cares whether ones' bonds sailed straight through 2008-2009 (Total Bond) or dropped 10% (Intermediate-Term Investment Grade) when stocks dropped 50%? You can argue until the cows come home about how the 10% drop impaired your rebalancing bonus or whatever, but still.
• The bond market--for investment-grade bonds, and assuming not too many weird things like callability--has got to be pretty efficient, because there just aren't as many variables to look at. If two highly-rated bonds have the same rating, maturity, and coupon, anyone can do the bond math and get the same answer. You don't need to know boo about the business their in, and the ratings agency did the job of looking at the balance sheet. There just can't be a lot of difference in price, and the chances that I'd be able to identify mispricing are slim. You can surf the waves of the yield curve, but everyone else knows what the yield curve is, too. It has to be awfully close to "you get what you pay for."
• Nobody can predict interest rates, and we're not talking about small errors; see the chart below.

This chart is from a Vanguard paper. The thin lines are the market's interest rate predictions--not one guru's vision, the wisdom-of-the-crowds prediction. The thick lines are what actually happened. Notice the utter failure to be even approximately right even as little as one year ahead. People have trouble understanding this chart, and I think the reason is that they can't actually believe what they're seeing. Thin little hairs? Prediction. Thick line? Fact. Any place it looks bushy, the predictions were badly wrong. It looks bushy everywhere.

[YDNAL]

As a new Boglehead, I owe many of you thanks since I've been following the forum for several months now. Not only have I been able to find answers to many of my questions, but you've also given me the courage to take charge of my retirement funds and make some important changes. So thank you all, you provide a wonderful service and have helped me enormously.

One of the things I've learned is that the bond allocation for my own retirement investments is too low, and I need to increase it. But with the low rates and inflation risk, it doesn't seem to be a good time to purchase a big chunk of bonds. (I know red flags may be going up at this point, but at worst I think I'm trying to secular market time.) My real question is whether bonds should be viewed fundamentally differently with the extreme low-rate environment we have now.

Welcome to Bogleland, grakster!

If you invest for the next day or next month or even next year.... one thing.

If you invest for 30, 40 years.... completely different thing.

Please tell us about yourself.

[nydad]

I went through this same process about 2 years ago, when interest rates were historically low and everyone was saying it was a bad time to buy bonds. So I hesitated...

However, after reading various books on asset allocation, and following advice on this forum, I decided to take the plunge and now have ~25% of my portfolio in bonds, mostly split between TIPS, intermed treasuries, and a PIMCO fund in my 401k.

To date, they've done fine - they haven't lost money, and I'd say overall I've made around 5% annually from them over the past 2 years, when people keep on saying bonds are going to crash. I think it was a good decision. No-one knows where interest rates are going. No-one.

If you want to learn more, read Swedroe's book on bonds. Then, set an asset allocation, and just buy. If you're worried rates will rise, then DCA in over a year - this will prevent regret.

Based on Swedroe's book, I stayed away from foreign bonds, and in general try to keep to highest quality bonds. As suggested, there's no harm in keeping your durations short for the time being - a mix of short and intermediate funds for example to get a shorter average duration. Instead, take your risk on the equity side (which I did - overweighting emerging markets equities, which have done really well this year).

[telemark]

Or with at least a 10-year time horizon, is the wisest thing not to worry, and stick with rebalancing into a total market index fund?

That's the right answer. With that time frame a sudden drop in bond prices, even if it does happen, is nothing more than an opportunity to pick up more bonds at lower prices. I wouldn't hold my breath waiting for that to happen, though. In addition to the many reasons frequently given on this board, demographics are strongly in favor of bonds. A whole generation of baby boomers is retiring and moving into safer investments.

EDIT: to add that while I've occasionally argued in favor of dollar cost averaging a lump sum into stocks, I don't recommend doing that with bonds. They're just not volatile enough for it to matter. Just rebalance now and get it over with.

[dratkinson]

I remember a senior investor here once noted that when equities tank, they can lose 50-90% of their value. When bonds tank, they can lose 10-15% of their value; so not much.

Other investors/authors here say intermediate term seems to be the sweet spot on the risk/reward curve.

Bottom line. Chose your favorite flavor (treasury, corporate, muni, TBM,...) in intermediate term and stay the course. If bonds do crash, you can tax-loss harvest among flavors/durations.



Nisi, thanks for the interesting chart, it's going in my "must keep" information as background/justification for my IPS.

[Noobvestor]

Thought experiment - don't take this as advice, necessarily. If I wanted to increase the impact of my bond allocation and thus reduce my bond allocation, I would go for something like the following instead of Total Bond:

1/3 Extended-Duration Treasuries (EDV ETF at Vanguard). Huge interest rate risk, but also something that provides more upside when the market goes down.

1/3 TIPS - returns aren't great, and may not be quite the flight-to-safety asset that Treasuries are, but the inflation protection gives it something over nominals

1/3 Cash - this pulls back your total duration a bit, and helps mitigate the exposure to rising rates faced by the other two options

Note that I didn't include corporates - since part of this thought puzzle involves upping equities, I see no reason to add company-related risks into this side of the equation. Also note that I've done zero back-testing on the above, and used round amounts - it's more an illustration of an idea than a concrete suggestion per se.

Finally, note that I didn't include short bonds as your question leaned toward, because you implied you wanted protection in a low-rate environment - in a low-rate environment, if stocks turn downward, there's only so much protection a short-term fund can provide vs. a long-term fund (if EDV drops a percent in yield, the impact is huge, vs. if BND does, etc...).

[optimpessim]

How about purchasing CDs with taxable funds and considering them a portion of my fixed income allocation? (I am thinking of PenFed 3y) Recently Taylor commented that CDs would be acceptable but I don't know if he was referring only to CDs in an IRA. I have the Vanguard Intermediate Muni Fund but am resistant to the idea of putting more into it at this time thinking that there may be more risk there in the 3 yr period.

[Noobvestor]

How about purchasing CDs with taxable funds and considering them a portion of my fixed income allocation? (I am thinking of PenFed 3y) Recently Taylor commented that CDs would be acceptable but I don't know if he was referring only to CDs in an IRA. I have the Vanguard Intermediate Muni Fund but am resistant to the idea of putting more into it at this time thinking that there may be more risk there in the 3 yr period.

I think CDs as part of fixed income is fine, but definitely not if you want to reduce your FI allocation relative to your equities. Treasuries especially (but high-quality bonds in general) will often go up when equities go down, so unit for unit, they may zig when stocks zag whereas CDs cannot.

FWIW, I own a combination of cash, munis and TIPS.

[grakster]

I am a lazy, sloppy, "satisficing" investor. I have fairly little interest in the nuances of allocation within bonds, for these reasons:

• Even with only about 1/3 stocks, the volatility of stocks is so much higher than bonds that what you do within the bond allocation just seems unimportant. Who cares whether ones' bonds sailed straight through 2008-2009 (Total Bond) or dropped 10% (Intermediate-Term Investment Grade) when stocks dropped 50%? You can argue until the cows come home about how the 10% drop impaired your rebalancing bonus or whatever, but still.
• The bond market--for investment-grade bonds, and assuming not too many weird things like callability--has got to be pretty efficient, because there just aren't as many variables to look at. If two highly-rated bonds have the same rating, maturity, and coupon, anyone can do the bond math and get the same answer. You don't need to know boo about the business their in, and the ratings agency did the job of looking at the balance sheet. There just can't be a lot of difference in price, and the chances that I'd be able to identify mispricing are slim. You can surf the waves of the yield curve, but everyone else knows what the yield curve is, too. It has to be awfully close to "you get what you pay for."
• Nobody can predict interest rates, and we're not talking about small errors; see the chart below.

This chart is from a Vanguard paper. The thin lines are the market's interest rate predictions--not one guru's vision, the wisdom-of-the-crowds prediction. The thick lines are what actually happened. Notice the utter failure to be even approximately right even as little as one year ahead. People have trouble understanding this chart, and I think the reason is that they can't actually believe what they're seeing. Thin little hairs? Prediction. Thick line? Fact. Any place it looks bushy, the predictions were badly wrong. It looks bushy everywhere.

I aspire to be a lazy investor, too - but thanks for providing some good reasons for being so. Don't worry, I'm not trying to predict interest rates, I think my real question is whether these lower rates fundamentally change the risk of bonds, which might suggest changing an overall asset allocation or the types of bonds held. Another argument for being lazy is that if I were to make some decision about how to allocate within bonds based on low rates now, I'd have to make a decision about undoing this at some later point. Which not only starts to sound more like market timing, but as a new Boglehead would undermine the effort to create investment discipline.

One of the things I've found out recently is that one of my retirement plans has access to the signal shares version of Vanguard's Total Bond Market Index (VBTSX), and I'm thinking that maybe the low 0.10% expense ratio might make more difference than any scheme to reduce bond risk.

My reaction to the historically low rate environment we are in is to gradually reduce my average duration. In part this is easy because I have some of my bond money in a ladder. So I can just reduce the duration of the replacement bonds when one pays off. I still think bonds have an important role in a portfolio in the form of stability when other things crash. As a result I have not changed my stock/bond allocation. In other posts I have been telling people to try for a average duration of around 4. That can be done by adding a short term bond fund to the usual mix of total bond and TIPS.

This idea makes good sense, but I'm not sure how I would implement it, or if I could given the limitations of my retirement plans. Are there any mutual funds out there to build a ladder from, and might they be part of the Fidelity No Transaction Fee Network?

If you invest for the next day or next month or even next year.... one thing.

If you invest for 30, 40 years.... completely different thing.

Please tell us about yourself.

Well if it were 30 years I wouldn't worry at all. My investments are all tax-advantaged, and one of the things that prompts this is that I'm turning 50 this year so I'm looking at a 15 year time horizon - but only have 25% bonds right now. No course to stay yet, but I'm trying to come up with a plan to reduce risk over the next few years.

[YDNAL]

If you invest for the next day or next month or even next year.... one thing.

If you invest for 30, 40 years.... completely different thing.

Please tell us about yourself.

Well if it were 30 years I wouldn't worry at all. My investments are all tax-advantaged, and one of the things that prompts this is that I'm turning 50 this year so I'm looking at a 15 year time horizon - but only have 25% bonds right now. No course to stay yet, but I'm trying to come up with a plan to reduce risk over the next few years.

Thanks for the follow-up.

You shouldn't worry over a 15-year time horizon either. By reducing risk by investing in Intermediate Bonds - the apparent sweet spot hystorically - your horizon waaaaay exceeds the Duration of these types of Funds.
http://www.bogleheads.org/wiki/Bonds:_Advanced_Topics

[Beagler]

" I like the (taxable) IT bond index fund because it provides more stability than the LT index fund, and more income than the ST index fund.  The Total Bond Market Index Fund is fine, but I vaguely wonder about a bond fund that has 35% of its portfolio in non-bonds (i.e., GNMA securities, with their risk of being prepaid early,  when interest rates tumble)." John Bogle

[wesleymouch]

Any thoughts on using VTAPX the short term TIPS fund in bond asset allocation. AUM appear to be small.

[Bustoff]

Asset allocation is based on bonds having certain characteristics and providing specific functions to a portfolio, such as low risk, low volatility, and yield that at least keeps ahead of inflation. With low rates and the possibility of inflation, bonds now have more risk than they used to, and current yields don't seem to provide a sufficient risk premium.

Whats wrong with equity and CD's rather than bonds ?

Nobody can predict interest rates

As long rates are at zero, have we not eliminated one of the two possible direction of rates, at least in the long term.

How else can we explain why the most diehard Bogleheads are steering away from long bonds ? If we preach that we cant predict rates, why are we market timing the long end of the curve ?

[Taylor Larimore]

Bustoff:

How else can we explain why the most diehard Bogleheads are steering away from long bonds ? If we preach that we cant predict rates, why are we market timing the long end of the curve ?

Total Bond Market Index Fund is the most often recommended Vanguard and Bogleheads bond fund. It contains a sizable percentage of long bonds as this Morningstar table shows:

Maturity.....% in Total Bond Market
1 to 3 Years---------21.79%
3 to 5 Years-------- 15.55
5 to 7 Years-------- 10.15
7 to 10 Years------ 10.17
10 to 15 Years----- 4.53
15 to 20 Years----- 4.34
20 to 30 Years----- 28.41
Over 30 Years----- 5.07

TBM bond holders do not use "market timing" because they know they will not have all their bonds in the wrong maturity.

Best wishes
Taylor

[grap0013]

I am a lazy, sloppy, "satisficing" investor. I have fairly little interest in the nuances of allocation within bonds, for these reasons:[list][*]Even with only about 1/3 stocks, the volatility of stocks is so much higher than bonds that what you do within the bond allocation just seems unimportant. Who cares whether ones' bonds sailed straight through 2008-2009 (Total Bond) or dropped 10% (Intermediate-Term Investment Grade) when stocks dropped 50%? You can argue until the cows come home about how the 10% drop impaired your rebalancing bonus or whatever, but still.

I'm kinda surprised by this statement from you. For someone as conservative as you, one would think every little bit of downside protection is a must to keep an investor from capitulating and selling low. A couple of percent could be the difference between selling low and staying the course. Couldn't it?

We can debate this all we want, but the facts are with 80%+ equity, long treasuries have worked best in the past. If under 80% equity, one could make a compeling argument for either short or intermed term treasuries. I personally favor the fidelity index 7 year treasury bond fund as a sole bond holding.

[Doc]

Total Bond Market Index Fund is the most often recommended Vanguard and Bogleheads bond fund. It contains a sizable percentage of long bonds as this Morningstar table shows:

Maturity.....% in Total Bond Market
1 to 3 Years---------21.79%
3 to 5 Years-------- 15.55
5 to 7 Years-------- 10.15
7 to 10 Years------ 10.17
10 to 15 Years----- 4.53
15 to 20 Years----- 4.34
20 to 30 Years----- 28.41
Over 30 Years----- 5.07

I don't use TBM as a benchmark so I don't look at the portfolio very often. I thought that TBM breakdown was about 40% short, 40% intermediate and only 20% long in the past. The current breakdown is more like 33/33/33. Has this changed over the past several years? Perhaps I have been thinking of duration and not maturity.

Can anyone bring me up to date?

[bdpb]

This is from Vanguard's Performance and Management page https://personal.vanguard.com/us/funds/snapshot?FundId=0584&FundIntExt=INT#tab=2.
It's different than M*. One list is number of bonds and the other is cap weighted?

Distribution by maturity (% of fund) as of 12/31/2012
Total Bond Mkt Index Adm
Under 1 Year 1.8%
1 - 3 Years 26.4%
3 - 5 Years 29.5%
5 - 10 Years 28.0%
10 - 20 Years 4.1%
20 - 30 Years 9.8%
Over 30 Years 0.4%
Total 100.0%

[Doc]

This is from Vanguard's Performance and Management page https://personal.vanguard.com/us/funds/snapshot?FundId=0584&FundIntExt=INT#tab=2.
It's different than M*. One list is number of bonds and the other is cap weighted?
Distribution by maturity (% of fund) as of 12/31/2012
Total Bond Mkt Index Adm
Under 1 Year 1.8%
1 - 3 Years 26.4%
3 - 5 Years 29.5%
5 - 10 Years 28.0%
10 - 20 Years 4.1%
20 - 30 Years 9.8%
Over 30 Years 0.4%
Total 100.0%

There are a number of differences but I find it hard to believe that # vs. cap weighted is the difference.

1) M*'s data is stale. But I don't think that even a three month lag can make as much a difference as we see.

2) There is the float adjusted difference. But this seem to be in the wrong direction. Long bonds should have a smaller proportional float than short bonds and this would mean that the free float index should have fewer long bonds not more.

for Weighted maturity/fund duration/index duration I found

Ishares AGG 6.40/4.58/4.97 (Effective Duration)
SPDR LAG 6.69/4.90/- (Modified Adjusted Duration)
Vanguard VBTLX 7.1/5.2/5.2 (Average Duration)

(All data from 12/31/12 through yesterday)

These three are certainly closer together than M*'s three month old data but still ???

[am]

Just curious but TBM holds a small percentage in bonds over 30 years. What are these exactly? Think I have heard of 100 year Mexican bonds.

[Peter Foley]

grankster

Another option to consider would be to add TIPs as part of your bond portfolio. I personally use stable value, Total Bond and TIPs. If you have stable value in a deferred account that is worth a look too.

In this interest rate environment, I agree with you that there is more downside risk on the bond side than usual. I also agree there is very little upside potential.

[Clive]

Maturity.....% in Total Bond Market
1 to 3 Years---------21.79%
3 to 5 Years-------- 15.55
5 to 7 Years-------- 10.15
7 to 10 Years------ 10.17
10 to 15 Years----- 4.53
15 to 20 Years----- 4.34
20 to 30 Years----- 28.41
Over 30 Years----- 5.07

TBM bond holders do not use "market timing" because they know they will not have all their bonds in the wrong maturity.

Conventional bonds generally see a rising price as interest rates decline and see a falling price when interest rates rise. Stocks also tend to do well over periods of declining interest rates

Inflation Bonds generally will see a rising price as real yields decline and see a falling price when real yields rise. Gold also tends to do well when real yields move more negative and perform poorly when real yields move more positive

Barbell of 20% each in

Stocks (undated variable coupon conventional bond)
TLT (long dated conventional bond)
T-Bills or SHY (2 year) or VFITX (5 year) or TIP
LTPZ (long dated inflation bond)
Gold (undated inflation bond)

[nisiprius]

As long [as] rates are at zero...

But they are not. Google on "yield curve." Go to Treasury. Read nonzero numbers.

[Clive]

As long [as] rates are at zero...

But they are not. Google on "yield curve." Go to Treasury. Read nonzero numbers.

Buying a 7-year, 1.25% coupon yield bond i.e. at par ($100), and holding for 2 years (5-year bond at that time), and assuming the yield curve is the same at that time i.e. 5-year yield = 0.76%, will see that bond priced at $102.40 (http://www.fixedincomeinvestor.co.uk/x/yieldcalc.html)

So 2 years of 1.25% = 2.5% income and a 2.4% capital gain = near enough 2.5% yearly potential benefit.

why the most diehard Bogleheads are steering away from long bonds ?

Maybe because if the yield curve remains much the same there's not a great deal of difference in reward between a 7-year and a 20-year (longer dated yields taper off much more slowly, so less capital gain if the yield curve remains the same). Generally - if yields decline further the 20-year would gain more. If yields rise the 20-year would lose more. Expectations might be that yields are more likely to rise than fall and as such swapping 20-year for 7-year as being the 'long' end might seem to be a better choice.