When is a barbell portfolio ever more desirable than a bullet portfolio?

CFA Level III
#1

Is there ever a situation when you would want a barbell portfolio?

#2

i’ve wondered the same…I thought I read somewhere that Bodie has said bullets would always be prefered, but if you were following the “return maximization” FI strategy, you might want to go w/ barbells to add return (…if they increased expected return for acceptable risk)

#3

upward sloping yield curve? I think :slight_smile:

#4

I think - it should rather be downward sloping. long duration would benefit from dec in interest rates and may more than compensate for the loss in shorter side.

#5

declining short and long term rates…steady intermediate rates…looks like a humback yield curve.

#6

ManVsCFA Wrote: ------------------------------------------------------- > declining short and long term rates…steady > intermediate rates…looks like a humback yield > curve. that would make sense

#7

barbell is best whenever rates decline on the long end. the value you gain with long-term bonds will more than compensate your loss on the short-end.

#8

^Not true, if I have a LT Bullet then I get all the upside and no downside on the ST end. A barbell is Best when you need to protect against TWISTS in the Yield Curve, as in hedging MBS’s. If you dont care about prorecting Twists then a Bullet is best.

#9
  1. If you want to take advantage of a yield curve flattening, but need to match the portfolio duration to something more intermediate, then you want a barbell. 2) If the yield curve “bowes out”, so that LT and ST rates drop more than mid-term rates, barbell is better. 3) If you are hedging an MBS, you are essentially using a barbell, though it could be long one part and short the other.
#10

bigwilly Wrote: ------------------------------------------------------- > ^Not true, if I have a LT Bullet then I get all > the upside and no downside on the ST end. A > barbell is Best when you need to protect against > TWISTS in the Yield Curve, as in hedging MBS’s. > If you dont care about prorecting Twists then a > Bullet is best. what about credit risk? where you are in corporates in the short end and treasury in the long end

#11
  1. If I had perfect prediction ability I would still go with a LT Bullet, higher Duration.
#12

I don’t think this is something that can easily be summed up, I think every circumstance is going to be different.

#13

Good point Willy. I guess Barbell is for doing things when you don’t have flexibility to choose your portfolio duration. If you have perfect knowledge, and can change portfolio duration to whatever you want, bullets are always better. If you have to keep duration within some range, and things are happening to the yield curve outside that range, you may want to use a Barbell because bullets aren’t an option.

#14

Lets say you want to begin the liability matching program for your pension plan… You have never implemented a liability approach… So you revamp your whole allo and set aside Fixed Income $: You need to gain duration with that fixed income, but you also need to provide liquidity for the beginning years of your liability schedule… You would match say: 0 to 5 years duration & 20 to 30 years duration of the liability It is better than a bullet portfolio because you gain both the liquidity on the front end and duration exposure on the back end (where most of your risk lies)…

#15

bchadwick Wrote: ------------------------------------------------------- > 2) If the yield curve “bowes out”, so that LT and > ST rates drop more than mid-term rates, barbell is > better. > If the curve bowes out, LT and ST rates are lower than mid-term rates, aren’t you overpaying for the ST/LT bonds in a barbell strategy?

#16

bpl1000 Wrote: ------------------------------------------------------- > bchadwick Wrote: > -------------------------------------------------- > ----- > > > 2) If the yield curve “bowes out”, so that LT > and > > ST rates drop more than mid-term rates, barbell > is > > better. > > > > If the curve bowes out, LT and ST rates are lower > than mid-term rates, aren’t you overpaying for the > ST/LT bonds in a barbell strategy? I think he was talking after the fact: Beginning Curve is flat Ending curve short term rates drop and long term rate drop, but say the mid rates stay the same - The curve bowes out.

#17

It also has more convexity right? so if you expect high volatility and hedge your duration than a barbell could be a good strategy…

#18

a barbell does have higher convexity than a bullet.

#19

^ longer bonds do have more convexity, that is true… Barbells are also a pain in the ass, because once the bond leaves that spectrum, you have to liquidate it, but that is far and beyond the scope of the curriculum (I am talking about actual money management)

#20

quorky Wrote: ------------------------------------------------------- > Lets say you want to begin the liability matching > program for your pension plan… You have never > implemented a liability approach… > > So you revamp your whole allo and set aside Fixed > Income $: > > You need to gain duration with that fixed income, > but you also need to provide liquidity for the > beginning years of your liability schedule… > > You would match say: > > 0 to 5 years duration > & > 20 to 30 years duration of the liability > > It is better than a bullet portfolio because you > gain both the liquidity on the front end and > duration exposure on the back end (where most of > your risk lies)… This is nice… That makes a lot of sense too. I was thinking about how you use barbells to take advantage of yield curve movements while controlling duration - this is about how you use barbells to maintain liquidity while controlling duration. Cool!