Which maturity?

[Justin]

For my next bond purchase, should I be purchasing bonds with 22 year maturity, or those with 24 year maturity?

I purchase my bonds on the secondary market. Until now I had been stocking up on bonds that mature in 2037 (GSBG37 at http://www.asx.com.au/asx/markets/inter ... NMENT_BOND). However recently the Australian Government has come out with GSBK39, which mature in about 24 years time.

So do I continue purchasing the 2037 bonds until they have just 20 years remaining on them, and roll them all over at once? Or should I start buying bonds that mature in 2039?

Note that my income is in a state of flux, and so it may be a long time (1-2 years) before I can next rebalance.

(by the way, I'm completely ignoring the yield on the two bonds, simply want to know which maturity to get)

Thanks in advance.

[Justin]

Actually I just thought of a third option. So it's:

1) Continue purchasing 2037-maturity bonds until it's time to roll over the entire lot,
2) Purchase bonds with the longest maturity no matter what (in this case 2039-maturity bonds), or
3) Roll over your entire 2037-maturity holding into the bonds that mature in 2039, and also purchase 2039-maturity bonds.

[Pet Hog]

Actually I just thought of a third option. So it's:

1) Continue purchasing 2037-maturity bonds until it's time to roll over the entire lot,
2) Purchase bonds with the longest maturity no matter what (in this case 2039-maturity bonds), or
3) Roll over your entire 2037-maturity holding into the bonds that mature in 2039, and also purchase 2039-maturity bonds.

If there are no fees for buying and selling, I would pick option 3.  That way you will have four years until you reach 20-years-to-maturity, and you will get slightly higher volatility and slightly higher yield with the 2039 bonds.  If there are fees, then I would go for option 2 and let the 2037 bonds do their thing for the next two years before making your next decision, possibly coinciding with your next rebalance.  I can't see any reason to follow option 1, unless you want to minimize the diversity of your bond holdings to minimize the number of trades (i.e., rebalance in 2017, then 2021, but not in 2019) to minimize fees.

[Justin]

Hi Pet Hog - thanks for the response.

There are fees ($15/trade, plus a spread), but you made me realise that the fee/spread would be eclipsed by the capital appreciation/depreciation, which is where I should be focusing my energy.

Still, less trades is better. I'm leaning towards 2).

[barrett]

I think with new money to invest, I would buy the 2039 bonds because they have the longest duration. Bonds with longer duration make more powerful moves when it's time for them to do their thing. They also get whacked the hardest when rates rise, but I am guessing you already know that.

A minor psychological consideration is that it also looks to me (I only have the chart you posted to go by) as if the 2039 bonds will kick off interest payments in a different month than the 2037 bonds do. Maybe it's just me, but I like to see those interest payments land in my account regularly. The way our 30-year issues are set up here in the US, that happens twice a year. I am loaded up on bonds that mature 5/15/43 so I see nice interest payments show up on 5/15 and 11/15 each year... and nothing in between except for price appreciation or depreciation.

This is just a mental thing that, at least for me, makes holding long bonds a bit easier. What month the bonds pay interest doesn't make you richer or poorer. But I do think that we humans like to see our investments do something.

Question for you, Justin... It looks to me that yields on 90-day bills there in Australia are about 2%. Is that true?

[ochotona]

No interest in a bond maturity ladder?

[Justin]

Hi Barrett/Ochotona - thanks, yes I'm favouring just buying the 2039 bonds and leaving my 2037 bonds alone for another couple of years.

Question for you, Justin... It looks to me that yields on 90-day bills there in Australia are about 2%. Is that true?

Not sure exactly where you are looking but Cash Rate in Australia is currently 2%.

[barrett]

No interest in a bond maturity ladder?

I think it was Craig who wrote or said somewhere that if an investor runs a PP for several years, a rough bond ladder is more or less created through rebalancing and adding new money

[Justin]

No interest in a bond maturity ladder?

I think it was Craig who wrote or said somewhere that if an investor runs a PP for several years, a rough bond ladder is more or less created through rebalancing and adding new money

Hi Barrett - if that's the case then you may have inadvertently answered my question. I wasn't aware of this.

[barrett]

No interest in a bond maturity ladder?

I think it was Craig who wrote or said somewhere that if an investor runs a PP for several years, a rough bond ladder is more or less created through rebalancing and adding new money

Hi Barrett - if that's the case then you may have inadvertently answered my question. I wasn't aware of this.

There are years like 2014 when you'd probably be selling some bonds because they've appreciated more than gold and stocks. After a bad year for bonds, you'd be buying them back up to 25%. So, over time you get bonds of different maturities... it's not a neat ladder as you would get if you set it up intentionally, but it works more or less the same.

Harry Browne's advice was to sell US 30-year bonds when they only had 20 years until maturity. He was really concerned with keeping the average bond duration long, but he simplified things by just using maturity.

I am curious what others on here think you should do if the longest maturity you can buy there in Australia is only 24 years. Wait until bonds are at 16 or 17 years to maturity? I'm not sure. A 24-year bond is close to a 30-year bond, but it's not exactly the same thing.

[MachineGhost]

You can't look at it in duration terms because it is completely mismatched in the portfolio with various bond maturities and to the equity (there are no 50-year duration bonds!).  Browne was just using duration as a proxy for the risk contribution.  What you really want is a risk parity where each of the three assets contribute equal risk, whether that be volatility or some other metric, and preferably which takes long-term, covariances into account.  But that goes against the simplistic 1/n weighting scheme.  Some bond durations will just be too low to offer enough risk contribution to match equity and gold.

[economicsjunkie]

Actually I just thought of a third option. So it's:

1) Continue purchasing 2037-maturity bonds until it's time to roll over the entire lot,
2) Purchase bonds with the longest maturity no matter what (in this case 2039-maturity bonds), or
3) Roll over your entire 2037-maturity holding into the bonds that mature in 2039, and also purchase 2039-maturity bonds.

Option 2) would be orthodox PP investing I believe, plus it makes sense in particular if you're paying fees. There's nothing wrong with having some bonds that mature later than others, as long as you replace those that have only 20 years left to go with the longest you can get.