Bonds: Simple vs Complicated

[galeno]

My wife and I will be 60 in 2017. We like to hold 60% world stocks and 5% cash.

Which bond portfolio do you like best for our retirement?

1. 100% TBM: ER = 0.25%. YTM = 1.84%. Duration = 5.6 yr.

2. 50% USA treasuries + 50% corps: ER = 0.12% YTM = 2.3%. Duration = 6.5 yr.

3. 30% USA treasuries + 30% corps + 20% TIPS + 20% Junk: ER = 0.21% YTM = 2.9%. Duration = 6.0 yr.

 

[Big_Hitter]

macaulay duration or modified duration?

 

[galeno]

Neither. Average effective duration.

 

[samclem]

I would favor option #2 (50% US Treasuries, 50% corporates). While corporate bonds can be expected to have higher returns than Treasuries over most years, the Treasuries can appreciate well when there is a panic/crash/market failure, etc and resultant flight to quality. Dark times like that are when my stocks will be taking a hit, so the diversification value of Treasuries can be important. OTOH, having the corporates will provide solid returns in years when stock prices take a dip due to more "normal" dips in valuation.
IIRC, there has been research that indicates some value in shifting between Treasuries and corporates based on stock valuations: When stock prices get bid up a lot, holding Treasuries can be better (because the eventual steep decline in stock prices and the market's reaction to that is more likely to disproportionately benefit government bonds). During periods of more "normal" stock valuations, the higher interest rates paid by corporate bonds is more beneficial.

Disclaimer: There are many people who know a lot more about bonds than I do.

 

[Big_Hitter]

my bond portfolio is all individual munis

I'm not sure I'd be going long right now in bond funds, ymmv

 

[galeno]

The average effective duration of the FI allocation can be decreased by holding (more) cash.


E.g. by adding 5% cash to the 35% bonds lowers 5.6 to 4.9, 6.5 to 5.7. and 6.0 to 5.3.

 

[galeno]

In 2008-9, intermediate USA treasuries went up about 13%; TBM up 8%. Intermediate corps bonds stayed flat at 0%.


USA treasuries have the best "correlation effect" when stocks crash.

 

[pb4uski]

You and I have the same overall AA... 60/35/5.

I divide equities into 70% US, 27% developed international and 3% emerging markets. I divide fixed income into 64% US investment grade, 16% US high yield, 17% developed international and 3% international emerging markets. I concede that I probably slice and dice more than necessary.

For US investment grade, I hold PenFed 3%/5 year CDs and target maturity corporate bond ETFs from Guggenheim and Blackrock that mature in 2020 (and a whole life policy that I bought in my 20s that pays ~4%). The ERs for the ETFs are .24% and .10%, respectively. For US high yield, I invest in Guggenheim Bulletshare ETFs maturing in 2017 to 2019 that have a .40% ER. The last time I checked, the weighted average yield for the above was 2.8% and the weighted average duration was 3.1. I concede that I am giving up some yield in my efforts to mitigate interest rate risk.... which so far has been a bit of a fool's errand.

If I had a lot of cash to invest in fixed income, I might take a hard look at the Andrews CU 3%, 7 year CDs recently offered. In fact, I'm considering swapping it out for my Guggenheim Bulletshares and IBonds.

I hold my cash in a Discover Bank online savings account that pays 0.95%.

I'm not keen on US treasuries... I think they are still as Warren Buffet said, return-free risk.... in the near term. I guess of the three alternatives you posted that #3 would be my preference.

 

[galeno]

So far one vote for #2 and one for #3.


I'm NOT a fan of slice and dice. I LOVE simplicity. but I will S/D if it gives clear advantages.


Advantages of #2. Lowest ER. Better yield. Disadvantages: complication (2 ETFs) and highest duration.


Advantages of #3. Lower ER. Best yield. Disadvantages: complication (4 ETFs) and higher duration.

 

[DEC-1982]

The average effective duration of the FI allocation can be decreased by holding (more) cash.


E.g. by adding 5% cash to the 35% bonds lowers 5.6 to 4.9, 6.5 to 5.7. and 6.0 to 5.3.

Is this calculation for 35 % bonds, 5% cash basically (35*5.6+5*0)/40=4.9?

I do consider CD as cash, and they are 1-year, 2-year and 5-year CDs. Are they zero duration for calculation purposes?

 

[Gotadimple]

#3
You have an equal balance of Treasuries and Corporate, including some short-term bonds. You get inflation protection with TIPS, and Junk Bonds act inverse to bonds in pricing. The only thing missing for a bit of growth are international emerging market bonds.

Within my 41% Bonds/4% Cash allocation by asset mix is:
Treasuries 34%
Corporates 38%
TIPS 17%
Junk 5%
Emerging 5%

 

[galeno]

Cash has a duration = 0. Xyr CD has a duration of X yr. If interest rates rise by 1% the NAV of the CD will fall by X%.


Regarding the EM bonds I notice you hold 55% equities. We hold 60%. I'd rather have 5% more TWM vs the 5% in EM bonds.

 

[Lsbcal]

Worst for non-treasury bonds would be a 1930s like depression. Not likely but one should have a plan just in case. With 60% stocks could be sleepless nights.

 

[galeno]

I'm OK with 60% stocks if using 100% TBM or 100% USA treasuries, or even 50/50 treasuries/corps.


If using 30% Treasuries + 30% corps + 20% TIPS + 20% junk I should probably decrease equities to 40-45%.

 

[friar1610]

I know this isn't a direct answer to your question, but...

My wife and I have IRAs that used to contain only VG Total Bond index. I recently added a small slice (about 1/3 of total) of VG Wellesley (35% value stocks/65% corporate bonds) to that mix. This gives us about 11% equity and ups the corporate bond amount. I have been thinking of buying a bit more more Wellesley and then swapping the TB index for an intermediate treasury fund. That would give us a "risky" segment (Wellesley) and a "safe" segment (Treasuries). I am at the age where I have to take RMDs (and my wife will be next year), so the plan would be to take the RMD from the segment that's doing the best in any given year.

EDIT: Of course, I include the equities within Wellesley in computing my overall AA which is somewhat more conservative than yours.

 

[brewer12345]

The one thing to be very careful of in total bond index funds is the negative convexity/extension risk embedded in the index. A lot of indicies have a substantial portion of the underlying bonds in agency mortgage-backed paper, typically based on 30 year fixed mortgage paper. These bonds have an effective duration quoted on them, but this is really just a guess. In a rising rate environment, duration balloons and you take it in the shorts. Caveat emptor.

 

[audreyh1]

The one thing to be very careful of in total bond index funds is the negative convexity/extension risk embedded in the index. A lot of indicies have a substantial portion of the underlying bonds in agency mortgage-backed paper, typically based on 30 year fixed mortgage paper. These bonds have an effective duration quoted on them, but this is really just a guess. In a rising rate environment, duration balloons and you take it in the shorts. Caveat emptor.

A lot of diversified bond funds hold these too.

 

[brewer12345]

A lot of diversified bond funds hold these too.

Absolutely. All fine if rates do not move much and you get extra yield. Rates start spiking and you get killed.

 

[pb4uski]

Cash has a duration = 0. Xyr CD has a duration of X yr. If interest rates rise by 1% the NAV of the CD will fall by X%.


Regarding the EM bonds I notice you hold 55% equities. We hold 60%. I'd rather have 5% more TWM vs the 5% in EM bonds.

Perhaps for brokered CDs but not for bank CDs... for bank CDs you could cash out for a EWP of in most cases a half a year of interest.

 

[bmcgonig]

A lot of diversified bond funds hold these too.

If I remember correctly Ishares AGG has quite a bit of these, and Vanguard BND has little or none.

 

[galeno]

Would you say 50% USA treasuries + 50% corps is better than 100% TBM because of the MBS?

I like the combo better because we get an expected performance bump of 0.74%.

Absolutely. All fine if rates do not move much and you get extra yield. Rates start spiking and you get killed.

 

[brewer12345]

Would you say 50% USA treasuries + 50% corps is better than 100% TBM because of the MBS?

I like the combo better because we get an expected performance bump of 0.74%.

MBS offers compensation for the embedded options you have sold to the mortgage borrowers: you get a noticeably higher yield and little or no credit risk. So it is not completely black and white.

I would say that I don't personally like holding MBS because I don't like the potential tail in big interest rate move scenarios (including lower rates: in a rate downdraft you get prepaid as everyone refis and you have to reinvest at lower rates). But if you are trying to spread your risks, I suppose including them is potentially a good diversification move. In the current environment, I am happy to keep my durations short, minimize MBS exposure, and stay away from junk.

 

[galeno]

What is your FI duration? Ours (35% TBM + 5% CASH) = 4.9 yr.

 

[eytonxav]

Right now, I would stick with short duration MF and/or an individual bond ladder.

 

[audreyh1]

If I remember correctly Ishares AGG has quite a bit of these, and Vanguard BND has little or none.

Huh? Vanguard BND (Total Bond Market ETF) has 23.36% in securitized (mostly mortgage backed) bonds plus 1.73% US Agency. In fact - exactly mirroring the Total Bond Market Index mutual funds it tracks.

Just because something is low in US Agency bonds doesn't mean it isn't holding a large slug of mortgage backed paper.

In comparison - AGG holds 0.39% US Agency and 17.96% in securitized loans (mostly Agency MBS Pass-Through). Actually less than BND and the Vanguard Total Bond Market Index mutual funds.