Page 2 of 2

Re: Large LTT purchase

Posted: Fri May 31, 2019 1:21 pm
by LittleDinghy
pmward wrote:
Thu May 23, 2019 2:41 pm
Personally what I would do is buy all of them at the next auction (next one is June 6). Then you can roll them a little each year. So if you want to target a 5 year rollover you sell and rollover 1/5 for the next 5 years, then after then they just roll automatically. If you do a 10 year rollover, you sell and rollover 1/10 every year for the next 10 years.
I am in the similar boat as the originator of this thread with similar size investments. I implemented the equity portions (20% S&P500 & 20% Small Cap Value) earlier this week and now want to implement the LTT 20% and the cash 20% (to be split between 0-12 month T-bills emergency funds in a taxable account and the bulk in 1-3 yr T-Notes). My question relates is whether (and why) it is better to buy the longest LTTs I can at Fidelity (April 15, 2049) and then rollover 1/5th for the next 5 years as suggested in the quote above, or to go ahead and set up a 5 rung ladder of LTTs at Fidelity maturing from 2045-2049? What are the pros and cons of each choice? Thank you!

Re: Large LTT purchase

Posted: Fri May 31, 2019 1:32 pm
by pmward
LittleDinghy wrote:
Fri May 31, 2019 1:21 pm
pmward wrote:
Thu May 23, 2019 2:41 pm
Personally what I would do is buy all of them at the next auction (next one is June 6). Then you can roll them a little each year. So if you want to target a 5 year rollover you sell and rollover 1/5 for the next 5 years, then after then they just roll automatically. If you do a 10 year rollover, you sell and rollover 1/10 every year for the next 10 years.
I am in the similar boat as the originator of this thread with similar size investments. I implemented the equity portions (20% S&P500 & 20% Small Cap Value) earlier this week and now want to implement the LTT 20% and the cash 20% (to be split between 0-12 month T-bills emergency funds in a taxable account and the bulk in 1-3 yr T-Notes). My question relates is whether (and why) it is better to buy the longest LTTs I can at Fidelity (April 15, 2049) and then rollover 1/5th for the next 5 years as suggested in the quote above, or to go ahead and set up a 5 rung ladder of LTTs at Fidelity maturing from 2045-2049? What are the pros and cons of each choice? Thank you!
Pro of buying all of the longest bond: You can buy them all at the next auction next week so it's simple and there is no spread you have to worry about.

Con of buying all the longest bond: Since you are doing the 1-3 year maturity on cash and a 5 year rollover on the long bonds, you would wind up with slightly longer average maturity (16 years vs 15) which means a bit added volatility (could also be a pro depending on if that volatility works in your favor or not). You can also go with all 1 year T Bills on the short end and wind up with 15.25 year average maturity when combined with purchasing all the 30 year up front, which is basically a wash on the average maturity and volatility as a whole.

Re: Large LTT purchase

Posted: Fri May 31, 2019 5:27 pm
by Kbg
This is in another thread...but very relevant.

https://portfoliocharts.com/2019/05/27/ ... convexity/

There's really no right or wrong here. It comes down to convenience and understanding the trade offs of your decision and being good with them.

If you read the article and study the chart I believe it is some ammo for what I was advocating on the long side of your purchases and also captures why I never mentioned anything about the short side. The long side will have serious repercussions (good or bad) to your portfolio. Therefore, I believe getting the average of whatever is going to happen is a more prudent option.

On a practical note regarding level of effort...you have no choice but to be active managing the short side. I don't see a whole lot more work doing same on the long side.

But again...see line #2.

Re: Large LTT purchase

Posted: Sat Jun 01, 2019 12:13 am
by ppnewbie
Kbg wrote:
Fri May 31, 2019 5:27 pm
This is in another thread...but very relevant.

https://portfoliocharts.com/2019/05/27/ ... convexity/

There's really no right or wrong here. It comes down to convenience and understanding the trade offs of your decision and being good with them.

If you read the article and study the chart I believe it is some ammo for what I was advocating on the long side of your purchases and also captures why I never mentioned anything about the short side. The long side will have serious repercussions (good or bad) to your portfolio. Therefore, I believe getting the average of whatever is going to happen is a more prudent option.

On a practical note regarding level of effort...you have no choice but to be active managing the short side. I don't see a whole lot more work doing same on the long side.

But again...see line #2.
Funny - I was actually starting to read Tyler's article and I saw your post as well.

Re: Large LTT purchase

Posted: Wed Jun 05, 2019 2:23 pm
by ppnewbie
Kbg wrote:
Thu May 23, 2019 11:04 pm
pmward wrote:
Thu May 23, 2019 2:59 pm
Kbg wrote:
Thu May 23, 2019 2:46 pm
Assuming you can buy open market bonds (e.g. those already issued) via a brokerage account(s) I'd definitely ladder. You want to be able to reinvest if interest rates ratchet up. (But that's an implicit forecast for sure)
Mathematically, is there really any difference between buying a ladder, or just buying all at auction up front? All changes in past yields should be prorated into the existing bonds.
Yes, because the coupons will all be different on day 1. Ladders at the end of the day are an averaging mechanism. If you know interest rates are heading straight south then you would never ladder and you would load up on 30 yr bonds. If you knew interest rates were headed straight north you would buy 4wk T-Bills. If you don’t claim to know any of that in advance, ladders get you the average over the time period you run the ladder at the bond duration level you target.

In the financial markets average is good. Most investors don’t achieve it.
Quick question on this. Isn't building in and harnessing the maximum volatility of one of the 4 asset classes a feature of the PP. Laddering in 30 year bonds over say 5 years seems to be a way to reduce some of the volatility. (Not being facetious here just trying to get a better handle on this) - but if you want to reduce volatility, 10 year bonds seem to be less volatile than a 30 year ladder. It would be great if there was a way to back test a PP portfolio with a close to 30 year average LTT maturities. Anyway thanks so much. I feel like I can start to see the forces working on the gyroscope with all the craziness happening in the economy at the moment.

Re: Large LTT purchase

Posted: Thu Jun 06, 2019 10:50 pm
by Kbg
ppnewbie wrote:
Wed Jun 05, 2019 2:23 pm
Quick question on this. Isn't building in and harnessing the maximum volatility of one of the 4 asset classes a feature of the PP. Laddering in 30 year bonds over say 5 years seems to be a way to reduce some of the volatility. (Not being facetious here just trying to get a better handle on this) - but if you want to reduce volatility, 10 year bonds seem to be less volatile than a 30 year ladder. It would be great if there was a way to back test a PP portfolio with a close to 30 year average LTT maturities. Anyway thanks so much. I feel like I can start to see the forces working on the gyroscope with all the craziness happening in the economy at the moment.
My personal take is the above is secondary. Primary is certain assets do better in different economic and monetary conditions. Volatility of the "on" asset helps mitigate the "off" assets. IDK, but I'm going to guess that when HBPP was working with the financial guy he mentioned in backtesting and constructing the PP they knew they needed the right assets to generate a decent real return and that probably wasn't going to happen with super mellow assets. It's counter-intuitive and I'm not going to go into the reasons here, but when constructing a portfolio too little volatility is a very undesirable characteristic. On the other end of the scale adding a dash of mellow to something volatile can actually improve not only risk adjusted but raw returns. If you're into matrix algebra, correlations and returns it's all out there in the google world.

You could get a pretty close answer to your question if you really wanted to. Take the dividend adjusted return history of TLT and do your own overlay using the bond duration formula and then back that into the PP's performance and rebalance annually.

Technical note...30 year bonds do not exist at all times. There are periods when they were not issued.

Re: Large LTT purchase

Posted: Fri Jun 07, 2019 4:47 pm
by ppnewbie
I am buying 1/4 of the LTT allocation to 30 year treasury bonds at auction
3/4 FNBGX with an average maturity of 24.1 years. So the total comes out to an average long term maturity of 25.575 years.
I may peel of a fourth of that FNBGX pile and get a 2048 bond. Anyway thanks for the help!