Modified versions of HFEA with ITT and Futures / Lifecycle Investing with Modern Portfolio Theory

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
impatientInv
Posts: 349
Joined: Fri Sep 03, 2021 1:26 pm

Re: Modified versions of HFEA with ITT and Futures / Lifecycle Investing with Modern Portfolio Theory

Post by impatientInv »

How are you managing losses from futures in a Roth IRA? If someone holds MES and ZF both in a Roth IRA it can be rough.

See the sample Example $100,000 in IRA in IBKR
target leverage 1.6x
4 MES - exposure $85k
1 ZN - 10 Year US Treasury
100 * VTI - $23k
800 * IXUS - $51k
US/INT - 68/32
~$4k cash

viewtopic.php?p=6635931#p6635931
No individual stocks.
Topic Author
skierincolorado
Posts: 2377
Joined: Sat Mar 21, 2020 10:56 am

Re: Modified versions of HFEA with ITT and Futures / Lifecycle Investing with Modern Portfolio Theory

Post by skierincolorado »

impatientInv wrote: Mon May 02, 2022 2:41 pm How are you managing losses from futures in a Roth IRA? If someone holds MES and ZF both in a Roth IRA it can be rough.

See the sample Example $100,000 in IRA in IBKR
target leverage 1.6x
4 MES - exposure $85k
1 ZN - 10 Year US Treasury
100 * VTI - $23k
800 * IXUS - $51k
US/INT - 68/32
~$4k cash

viewtopic.php?p=6635931#p6635931
I'm not sure what you're asking, but if asking how to maintain leverage without margin call, sell 20k vti and buy another mes.
Topic Author
skierincolorado
Posts: 2377
Joined: Sat Mar 21, 2020 10:56 am

Re: Modified versions of HFEA with ITT and Futures / Lifecycle Investing with Modern Portfolio Theory

Post by skierincolorado »

hiddenpower wrote: Sat Apr 30, 2022 3:45 pm Which of these seem preferable (taxable account)? I'd like to avoid futures, delay taxation, and aim to keep things simple.

1) 45 UPRO / 55 TYA
2) 45 UPRO / 55 TYD
3) 45 UPRO / 55 EDV

I'm actually leaning towards (1) since I could ultimately keep TYA around in my main portfolio if I ever sold off UPRO. But does the fact that TYA isn't daily rebalanced, whereas UPRO is daily rebalanced matter at all? I was also surprised that (3) seems to be doing the best in terms of gains and a slightly less drawdown but this could just be random.

Also for simulating things I went with VFITX, but is there a way to go back farther in time for backtesting ITTs?

Moreover, if I'm set on just optimizing returns, does it make sense to favor UPRO more than 55? Comparison https://imgur.com/a/JTRkT8C
Edv is similar duration exposure to tya and tyd. Long bonds have actually beaten shorter bonds recently on a duration basis. But long term that is not true at all. While a 45/55 upro/tyd would be similar returns to HFEA with less bond risk, 45/55 edv would not be close at all.

You can test all of this in the simba spreadsheet. But if using itt for tyd you must convert the duration exposure. Likewise tmf and ltt. Or if using ltt for edv.


Personally I would go 55/45 upro tyd. Tyd has something like 9 years bond duration. I am 145/170 but my bonds have only 5 year duration, which would be like a 35% tyd allocation.
45/55 upro/tyd is only a little less bond risk than hfea. Which we can see is too much for most people. 55/45 upro/tyd is a more durable aa that gets you through the 70s and early 80s better.

If that is too stocky then just do 10% vti 45 upro 45 tyd.

I really don't think tyd should be above 45. Anything above 45 is kind of a lotto that bond returns and term premiums look like the last 40 years and not the previous 30.

Or 60/40 upro/tyd. Or 45/15/40 upro/vti/tyd.

Also be sure tya or tyd have sufficient liquidity. From what j understand the spreads are wide but I don't know. Definitely don't day trade it.


What is the rebal frequency of tya? I don't think it matters too much as long as they rebal regularly.

Also consider that using fixed leverage etfs is not consistent with lifecycle investing. Lifecycle investing says that most under age 45 should increase leverage when they lose money in order to maintain a fixed percent of present and future lifetime savings invested (those under 30 should just always have 2x leverage in stocks). If you are using fixed ratios that doesn't happen. If using futures it happens naturally. When I lose money my leverage goes up.
User avatar
hiddenpower
Posts: 564
Joined: Tue Nov 17, 2020 11:24 pm

Re: Modified versions of HFEA with ITT and Futures / Lifecycle Investing with Modern Portfolio Theory

Post by hiddenpower »

skierincolorado wrote: Mon May 02, 2022 3:04 pm
hiddenpower wrote: Sat Apr 30, 2022 3:45 pm Which of these seem preferable (taxable account)? I'd like to avoid futures, delay taxation, and aim to keep things simple.

1) 45 UPRO / 55 TYA
2) 45 UPRO / 55 TYD
3) 45 UPRO / 55 EDV

I'm actually leaning towards (1) since I could ultimately keep TYA around in my main portfolio if I ever sold off UPRO. But does the fact that TYA isn't daily rebalanced, whereas UPRO is daily rebalanced matter at all? I was also surprised that (3) seems to be doing the best in terms of gains and a slightly less drawdown but this could just be random.

Also for simulating things I went with VFITX, but is there a way to go back farther in time for backtesting ITTs?

Moreover, if I'm set on just optimizing returns, does it make sense to favor UPRO more than 55? Comparison https://imgur.com/a/JTRkT8C
Edv is similar duration exposure to tya and tyd. Long bonds have actually beaten shorter bonds recently on a duration basis. But long term that is not true at all. While a 45/55 upro/tyd would be similar returns to HFEA with less bond risk, 45/55 edv would not be close at all.

You can test all of this in the simba spreadsheet. But if using itt for tyd you must convert the duration exposure. Likewise tmf and ltt. Or if using ltt for edv.


Personally I would go 55/45 upro tyd. Tyd has something like 9 years bond duration. I am 145/170 but my bonds have only 5 year duration, which would be like a 35% tyd allocation.
45/55 upro/tyd is only a little less bond risk than hfea. Which we can see is too much for most people. 55/45 upro/tyd is a more durable aa that gets you through the 70s and early 80s better.

If that is too stocky then just do 10% vti 45 upro 45 tyd.

I really don't think tyd should be above 45. Anything above 45 is kind of a lotto that bond returns and term premiums look like the last 40 years and not the previous 30.

Or 60/40 upro/tyd. Or 45/15/40 upro/vti/tyd.

Also be sure tya or tyd have sufficient liquidity. From what j understand the spreads are wide but I don't know. Definitely don't day trade it.


What is the rebal frequency of tya? I don't think it matters too much as long as they rebal regularly.

Also consider that using fixed leverage etfs is not consistent with lifecycle investing. Lifecycle investing says that most under age 45 should increase leverage when they lose money in order to maintain a fixed percent of present and future lifetime savings invested (those under 30 should just always have 2x leverage in stocks). If you are using fixed ratios that doesn't happen. If using futures it happens naturally. When I lose money my leverage goes up.
TYA is 2.25X rebalanced/rolled quarterly I believe. So wasn't sure if that would make a difference when paired with UPRO, given TYD is rebalanced daily too. Moreover, is there an effective way to actually do this in a taxable account without realizing lots of bond gains during an equities crash? I feel the tax drag of 1-2% stated in HFEA thread may be mis-understated.


As for lifecycle investing, it seems like LETFs are lifecycle investing to the far extreme since it reset when you lose money which is what Nayre's recommends (selling at a loss and resetting the leverage to say 2x(.
Topic Author
skierincolorado
Posts: 2377
Joined: Sat Mar 21, 2020 10:56 am

Re: Modified versions of HFEA with ITT and Futures / Lifecycle Investing with Modern Portfolio Theory

Post by skierincolorado »

hiddenpower wrote: Mon May 02, 2022 6:38 pm
skierincolorado wrote: Mon May 02, 2022 3:04 pm
hiddenpower wrote: Sat Apr 30, 2022 3:45 pm Which of these seem preferable (taxable account)? I'd like to avoid futures, delay taxation, and aim to keep things simple.

1) 45 UPRO / 55 TYA
2) 45 UPRO / 55 TYD
3) 45 UPRO / 55 EDV

I'm actually leaning towards (1) since I could ultimately keep TYA around in my main portfolio if I ever sold off UPRO. But does the fact that TYA isn't daily rebalanced, whereas UPRO is daily rebalanced matter at all? I was also surprised that (3) seems to be doing the best in terms of gains and a slightly less drawdown but this could just be random.

Also for simulating things I went with VFITX, but is there a way to go back farther in time for backtesting ITTs?

Moreover, if I'm set on just optimizing returns, does it make sense to favor UPRO more than 55? Comparison https://imgur.com/a/JTRkT8C
Edv is similar duration exposure to tya and tyd. Long bonds have actually beaten shorter bonds recently on a duration basis. But long term that is not true at all. While a 45/55 upro/tyd would be similar returns to HFEA with less bond risk, 45/55 edv would not be close at all.

You can test all of this in the simba spreadsheet. But if using itt for tyd you must convert the duration exposure. Likewise tmf and ltt. Or if using ltt for edv.


Personally I would go 55/45 upro tyd. Tyd has something like 9 years bond duration. I am 145/170 but my bonds have only 5 year duration, which would be like a 35% tyd allocation.
45/55 upro/tyd is only a little less bond risk than hfea. Which we can see is too much for most people. 55/45 upro/tyd is a more durable aa that gets you through the 70s and early 80s better.

If that is too stocky then just do 10% vti 45 upro 45 tyd.

I really don't think tyd should be above 45. Anything above 45 is kind of a lotto that bond returns and term premiums look like the last 40 years and not the previous 30.

Or 60/40 upro/tyd. Or 45/15/40 upro/vti/tyd.

Also be sure tya or tyd have sufficient liquidity. From what j understand the spreads are wide but I don't know. Definitely don't day trade it.


What is the rebal frequency of tya? I don't think it matters too much as long as they rebal regularly.

Also consider that using fixed leverage etfs is not consistent with lifecycle investing. Lifecycle investing says that most under age 45 should increase leverage when they lose money in order to maintain a fixed percent of present and future lifetime savings invested (those under 30 should just always have 2x leverage in stocks). If you are using fixed ratios that doesn't happen. If using futures it happens naturally. When I lose money my leverage goes up.
TYA is 2.25X rebalanced/rolled quarterly I believe. So wasn't sure if that would make a difference when paired with UPRO, given TYD is rebalanced daily too. Moreover, is there an effective way to actually do this in a taxable account without realizing lots of bond gains during an equities crash? I feel the tax drag of 1-2% stated in HFEA thread may be mis-understated.


As for lifecycle investing, it seems like LETFs are lifecycle investing to the far extreme since it reset when you lose money which is what Nayre's recommends (selling at a loss and resetting the leverage to say 2x(.
I don't think the less frequent rebalance would be a very big issue. The volatility of tya isn't that high anyways. I did not know they did this.

Sincd I have not used letf in taxable I have not looked into the tax drag heavily. I don't think it would hit cagr by 1%. They are taxes you have to pay eventually, you just have to pay it sooner which means you don't get the benefit of a free loan to invest it.

If someone is at 2x equity leverage (200% of their net worth invested in stock) then yes letf are consistent with lifecycle investing. But for the other phases of lifecycle investing one is not at 2x leverage and leverage goes up or down depending on how much one has invested relartive to expected lifetime savings. So if the market goes down, at that point you have relatively little invested relative to lifetime savings, so lifecycle investing says you must increase leverage (up to 2x) to try to maintain your investments.

Ayres and nalebuff somewhat arbitrarily limit leverage to 2x. But there is an argument for letting it go as high as 2.5x I think. The op of the lifecycle thread was doing that. That would make the time period of one's life longer where leveraging is going up and down. The 2.5x would end at a very young age. It would then float, and maybe go up to 2.5x again in a market crash.

You can't do this easily with letf if you are also investing in bond letf. Certainly not in taxable.
Topic Author
skierincolorado
Posts: 2377
Joined: Sat Mar 21, 2020 10:56 am

Re: Modified versions of HFEA with ITT and Futures / Lifecycle Investing with Modern Portfolio Theory

Post by skierincolorado »

hiddenpower wrote: Mon May 02, 2022 6:38 pm
skierincolorado wrote: Mon May 02, 2022 3:04 pm
hiddenpower wrote: Sat Apr 30, 2022 3:45 pm Which of these seem preferable (taxable account)? I'd like to avoid futures, delay taxation, and aim to keep things simple.

1) 45 UPRO / 55 TYA
2) 45 UPRO / 55 TYD
3) 45 UPRO / 55 EDV

I'm actually leaning towards (1) since I could ultimately keep TYA around in my main portfolio if I ever sold off UPRO. But does the fact that TYA isn't daily rebalanced, whereas UPRO is daily rebalanced matter at all? I was also surprised that (3) seems to be doing the best in terms of gains and a slightly less drawdown but this could just be random.

Also for simulating things I went with VFITX, but is there a way to go back farther in time for backtesting ITTs?

Moreover, if I'm set on just optimizing returns, does it make sense to favor UPRO more than 55? Comparison https://imgur.com/a/JTRkT8C
Edv is similar duration exposure to tya and tyd. Long bonds have actually beaten shorter bonds recently on a duration basis. But long term that is not true at all. While a 45/55 upro/tyd would be similar returns to HFEA with less bond risk, 45/55 edv would not be close at all.

You can test all of this in the simba spreadsheet. But if using itt for tyd you must convert the duration exposure. Likewise tmf and ltt. Or if using ltt for edv.


Personally I would go 55/45 upro tyd. Tyd has something like 9 years bond duration. I am 145/170 but my bonds have only 5 year duration, which would be like a 35% tyd allocation.
45/55 upro/tyd is only a little less bond risk than hfea. Which we can see is too much for most people. 55/45 upro/tyd is a more durable aa that gets you through the 70s and early 80s better.

If that is too stocky then just do 10% vti 45 upro 45 tyd.

I really don't think tyd should be above 45. Anything above 45 is kind of a lotto that bond returns and term premiums look like the last 40 years and not the previous 30.

Or 60/40 upro/tyd. Or 45/15/40 upro/vti/tyd.

Also be sure tya or tyd have sufficient liquidity. From what j understand the spreads are wide but I don't know. Definitely don't day trade it.


What is the rebal frequency of tya? I don't think it matters too much as long as they rebal regularly.

Also consider that using fixed leverage etfs is not consistent with lifecycle investing. Lifecycle investing says that most under age 45 should increase leverage when they lose money in order to maintain a fixed percent of present and future lifetime savings invested (those under 30 should just always have 2x leverage in stocks). If you are using fixed ratios that doesn't happen. If using futures it happens naturally. When I lose money my leverage goes up.
TYA is 2.25X rebalanced/rolled quarterly I believe. So wasn't sure if that would make a difference when paired with UPRO, given TYD is rebalanced daily too. Moreover, is there an effective way to actually do this in a taxable account without realizing lots of bond gains during an equities crash? I feel the tax drag of 1-2% stated in HFEA thread may be mis-understated.


As for lifecycle investing, it seems like LETFs are lifecycle investing to the far extreme since it reset when you lose money which is what Nayre's recommends (selling at a loss and resetting the leverage to say 2x(.
In addition to my last post, overall I'd say it doesn't sound like you quite get lifecycle investing yet. More than anything else by far it is what will increase lifetime returns and reduce risk and is the most important part to understand.

Are you an early or mid stage accumulator? If so I'd brush up on your lifecycle investing. I think it may also help some of the jitters I see you expressing about leverage. Listen to the audio book if you haven't and read the lifecycle thread. What really helped me was Steve's comments when asked if high leverage scared him he said he felt better knowing he was investing his future earnings sooner rather than later.

It's way more important to get the level of leverage right than the stock bond ratios. If unsure on bonds, just err on the side of stocks. 200/100 is a great aa for an early accumulator. You need that equity exposure. Whatever you have invested today will be nothing compared to what you have invested 20 years from now if you are working and saving that whole time.
DMoogle
Posts: 549
Joined: Sat Oct 31, 2020 10:24 pm

Re: Modified versions of HFEA with ITT and Futures / Lifecycle Investing with Modern Portfolio Theory

Post by DMoogle »

Brutal. Taxable current value at $270k vs. ATH of $480k. Haven't really had a firm de-levering strategy, but after each month of beatings, my leverage in my taxable has crept up to 235/410.

Selling my NTSX now that it's at a loss, selling a few ZN contracts, and unloading some random stocks I've had for ages. My target ratio is 165/250. Not yet sure if I'm going to delever to that or go a little higher. Will make a decision by EOD.
Last edited by DMoogle on Fri May 06, 2022 11:04 am, edited 1 time in total.
muffins14
Posts: 5528
Joined: Wed Oct 26, 2016 4:14 am
Location: New York

Re: Modified versions of HFEA with ITT and Futures / Lifecycle Investing with Modern Portfolio Theory

Post by muffins14 »

I'm sitting at like 115 / 80. Not really fun watching my treasuries futures continue to drain cash away from my portfolio. In taxable I have 1 TN, 2 ZN 2 ZF, and 1 ZT.

I guess I am going to roll and keep all of them when the time comes, and just hope the higher yields do something nice for me
Crom laughs at your Four Winds
Topic Author
skierincolorado
Posts: 2377
Joined: Sat Mar 21, 2020 10:56 am

Re: Modified versions of HFEA with ITT and Futures / Lifecycle Investing with Modern Portfolio Theory

Post by skierincolorado »

DMoogle wrote: Fri May 06, 2022 9:31 am Brutal. Taxable current value at $270k vs. ATH of $480k. Haven't really had a firm de-levering strategy, but after each month of beatings, my leverage in my taxable has crept up to 235/410.

Selling my NTSX now that it's at a loss, selling a few ZN contracts, and unloading some random stocks I've had for ages. My target ratio is 165/250. Not yet if I'm going to delever to that or go a little higher. Will make a decision by EOD.
Brutal indeed.
klaus14
Posts: 953
Joined: Sun Nov 25, 2018 6:43 pm

Re: Modified versions of HFEA with ITT and Futures / Lifecycle Investing with Modern Portfolio Theory

Post by klaus14 »

muffins14 wrote: Fri May 06, 2022 9:45 am I'm sitting at like 115 / 80. Not really fun watching my treasuries futures continue to drain cash away from my portfolio. In taxable I have 1 TN, 2 ZN 2 ZF, and 1 ZT.

I guess I am going to roll and keep all of them when the time comes, and just hope the higher yields do something nice for me
Higher yields doesn't automatically mean good return for strategy because as yields rise financing cost of futures also rise. A reasonable return expectation for treasury futures is the term premium. And as of today, latest ACM term premium estimation is still below financing cost spread (~0.25) for 10Y (0.069) and 7Y (0.176). So expected return of TN and ZN is still negative.

And there is the negative correlation benefit between treasuries and stocks. That is not holding up either. On the contrary, low performance of the bonds (rising yields) is the reason for low performance of stocks.

I am planning to not have any bonds or bond futures until
- Real yield is comfortably above zero.
- Term premium prediction is comfortably above financing cost spread.
- Positive correlation reverses (I don't need to exactly time this, some lag is fine)
My investment algorithm: https://www.bogleheads.org/forum/viewtopic.php?f=10&t=351899&p=6112869#p6112869
Topic Author
skierincolorado
Posts: 2377
Joined: Sat Mar 21, 2020 10:56 am

Re: Modified versions of HFEA with ITT and Futures / Lifecycle Investing with Modern Portfolio Theory

Post by skierincolorado »

klaus14 wrote: Fri May 06, 2022 3:33 pm
muffins14 wrote: Fri May 06, 2022 9:45 am I'm sitting at like 115 / 80. Not really fun watching my treasuries futures continue to drain cash away from my portfolio. In taxable I have 1 TN, 2 ZN 2 ZF, and 1 ZT.

I guess I am going to roll and keep all of them when the time comes, and just hope the higher yields do something nice for me
Higher yields doesn't automatically mean good return for strategy because as yields rise financing cost of futures also rise. A reasonable return expectation for treasury futures is the term premium. And as of today, latest ACM term premium estimation is still below financing cost spread (~0.25) for 10Y (0.069) and 7Y (0.176). So expected return of TN and ZN is still negative.

And there is the negative correlation benefit between treasuries and stocks. That is not holding up either. On the contrary, low performance of the bonds (rising yields) is the reason for low performance of stocks.

I am planning to not have any bonds or bond futures until
- Real yield is comfortably above zero.
- Term premium prediction is comfortably above financing cost spread.
- Positive correlation reverses (I don't need to exactly time this, some lag is fine)
How will you define the correlation? The daily correlation has been negative YTD. I'm not sure what the correlation of monthly returns has been.
muffins14
Posts: 5528
Joined: Wed Oct 26, 2016 4:14 am
Location: New York

Re: Modified versions of HFEA with ITT and Futures / Lifecycle Investing with Modern Portfolio Theory

Post by muffins14 »

I see, so if I take your estimates to be correct, then by not selling my futures now, I'm basically hoping that in the event of market crash people still flock to treasuries, increase the price to overcome the negative expected return

I'm down about $50k on the futures, so debating whether to
A) call this my lesson, stop using futures, and carry over my losses, so the loss feels more like $25k long term
B) HODL, maybe there will be some price spike due to a market crash
C) stay the course, because I was treating this like a leveraged 70/30 portfolio plus lifecycle investing, such that I could basically pretend I had a 2M portfolio rather than a 1.2M portfolio. If I really had 2M at 70/30 today, I wouldn't sell the bonds based on near-term real return expectations, so why should I sell these treasury futures?
Crom laughs at your Four Winds
Topic Author
skierincolorado
Posts: 2377
Joined: Sat Mar 21, 2020 10:56 am

Re: Modified versions of HFEA with ITT and Futures / Lifecycle Investing with Modern Portfolio Theory

Post by skierincolorado »

muffins14 wrote: Fri May 06, 2022 4:58 pm I see, so if I take your estimates to be correct, then by not selling my futures now, I'm basically hoping that in the event of market crash people still flock to treasuries, increase the price to overcome the negative expected return

I'm down about $50k on the futures, so debating whether to
A) call this my lesson, stop using futures, and carry over my losses, so the loss feels more like $25k long term
B) HODL, maybe there will be some price spike due to a market crash
C) stay the course, because I was treating this like a leveraged 70/30 portfolio plus lifecycle investing, such that I could basically pretend I had a 2M portfolio rather than a 1.2M portfolio. If I really had 2M at 70/30 today, I wouldn't sell the bonds based on near-term real return expectations, so why should I sell these treasury futures?
Klaus' points are worth considering. You can see some of our previous posts regarding these points but to summarize: the term premium is an estimate not a known quantity, we don't know when the next drop in yields could occur like March 2020, and it seems unlikely that bond futures would have negative return at the same time as positive equity correlation over a longer time horizon. There would be no incentive to take the risk. I just don't see reason to believe things have radically changed from history or theory, especially with bond yields near ten year highs.

The primary reason I think to own treasury futures has always been that they historically provide significant return in a major equities crash. I can't see any reason that has changed and am feeling better in some ways that they have more room to fall before hitting the zero lower bound and the fed has to consider negative rates.
klaus14
Posts: 953
Joined: Sun Nov 25, 2018 6:43 pm

Re: Modified versions of HFEA with ITT and Futures / Lifecycle Investing with Modern Portfolio Theory

Post by klaus14 »

skierincolorado wrote: Fri May 06, 2022 4:57 pm
klaus14 wrote: Fri May 06, 2022 3:33 pm
muffins14 wrote: Fri May 06, 2022 9:45 am I'm sitting at like 115 / 80. Not really fun watching my treasuries futures continue to drain cash away from my portfolio. In taxable I have 1 TN, 2 ZN 2 ZF, and 1 ZT.

I guess I am going to roll and keep all of them when the time comes, and just hope the higher yields do something nice for me
Higher yields doesn't automatically mean good return for strategy because as yields rise financing cost of futures also rise. A reasonable return expectation for treasury futures is the term premium. And as of today, latest ACM term premium estimation is still below financing cost spread (~0.25) for 10Y (0.069) and 7Y (0.176). So expected return of TN and ZN is still negative.

And there is the negative correlation benefit between treasuries and stocks. That is not holding up either. On the contrary, low performance of the bonds (rising yields) is the reason for low performance of stocks.

I am planning to not have any bonds or bond futures until
- Real yield is comfortably above zero.
- Term premium prediction is comfortably above financing cost spread.
- Positive correlation reverses (I don't need to exactly time this, some lag is fine)
How will you define the correlation? The daily correlation has been negative YTD. I'm not sure what the correlation of monthly returns has been.
Not the perfect method but i think we can look at NTSX (SP500 + bond futures) vs VOO (SP500) monthly returns. (PV)

What you want is: When VOO is negative, NTSX is less negative or positive. Which was the case until 2020/10. After that regime changes and what you see is if VOO is negative then NTSX is even more negative (except 2021/11)
Last edited by klaus14 on Sat May 07, 2022 2:36 pm, edited 1 time in total.
My investment algorithm: https://www.bogleheads.org/forum/viewtopic.php?f=10&t=351899&p=6112869#p6112869
klaus14
Posts: 953
Joined: Sun Nov 25, 2018 6:43 pm

Re: Modified versions of HFEA with ITT and Futures / Lifecycle Investing with Modern Portfolio Theory

Post by klaus14 »

muffins14 wrote: Fri May 06, 2022 4:58 pm I see, so if I take your estimates to be correct, then by not selling my futures now, I'm basically hoping that in the event of market crash people still flock to treasuries, increase the price to overcome the negative expected return

I'm down about $50k on the futures, so debating whether to
A) call this my lesson, stop using futures, and carry over my losses, so the loss feels more like $25k long term
B) HODL, maybe there will be some price spike due to a market crash
C) stay the course, because I was treating this like a leveraged 70/30 portfolio plus lifecycle investing, such that I could basically pretend I had a 2M portfolio rather than a 1.2M portfolio. If I really had 2M at 70/30 today, I wouldn't sell the bonds based on near-term real return expectations, so why should I sell these treasury futures?
Expected return in lower duration seems positive so on the whole probably whole bond future curve expected return is ~0.

I suspect estimation being harder for lower duration (because you have less prior duration to estimate embedded t-bill rate trajectory) so i wouldn't use this to time different durations. But i'd be much more comfortable if whole curve was above financing spread.

If expected return is zero, it means if future bond rate trajectory happens as market expects then you'll get zero from your treasury futures. I guess happy case for treasury futures is: market crashes -> FED capitulates and bonds recover (violating expectations). But i think it's not wise to try to time these things. in the long term up and down surprises should cancel each other and you should get the expected return. Same with correlations, they should revert to long term trend at some point. I am trying to market time my entry point. Wish me luck :)
Last edited by klaus14 on Fri May 06, 2022 6:54 pm, edited 2 times in total.
My investment algorithm: https://www.bogleheads.org/forum/viewtopic.php?f=10&t=351899&p=6112869#p6112869
Topic Author
skierincolorado
Posts: 2377
Joined: Sat Mar 21, 2020 10:56 am

Re: Modified versions of HFEA with ITT and Futures / Lifecycle Investing with Modern Portfolio Theory

Post by skierincolorado »

klaus14 wrote: Fri May 06, 2022 6:34 pm

Not the perfect method but i think we can look at NTSX (SP500 + bond futures) vs VOO (SP500) monthly returns. (PV)

What you want is: When VOO is negative, NTSX is less negative or positive. Which was the case until 2020/10. After that regime changes and what you see is if VOO is negative then NTSX is even more negative (except 2022/11)
Yes so that would probably be more related to the correlation of monthly returns
Dry-Drink
Posts: 530
Joined: Fri Jan 14, 2022 9:50 am

Re: Modified versions of HFEA with ITT and Futures / Lifecycle Investing with Modern Portfolio Theory

Post by Dry-Drink »

muffins14 wrote: Fri May 06, 2022 9:45 am I'm sitting at like 115 / 80.
Is your target actually 100/40 (based on sig)? As in, you have 2x the bond exposure your strategic portfolio calls for?
muffins14
Posts: 5528
Joined: Wed Oct 26, 2016 4:14 am
Location: New York

Re: Modified versions of HFEA with ITT and Futures / Lifecycle Investing with Modern Portfolio Theory

Post by muffins14 »

Dry-Drink wrote: Sat May 07, 2022 12:07 am
muffins14 wrote: Fri May 06, 2022 9:45 am I'm sitting at like 115 / 80.
Is your target actually 100/40 (based on sig)? As in, you have 2x the bond exposure your strategic portfolio calls for?
More like 115/50 now for the goal. When yields had changed a lot I bought the ZT, which adds a lot of exposure but also is pretty low volatility, so it was somewhat impulsive of me to get it. Selling would mean a loss of $300 there and get me to 115/57. So, still over that previous decision, but it’s less bond risk than most in this thread at least

Overall I want a 70/30 portfolio, but leveraged
Crom laughs at your Four Winds
Dry-Drink
Posts: 530
Joined: Fri Jan 14, 2022 9:50 am

Re: Modified versions of HFEA with ITT and Futures / Lifecycle Investing with Modern Portfolio Theory

Post by Dry-Drink »

muffins14 wrote: Sat May 07, 2022 8:59 am
Dry-Drink wrote: Sat May 07, 2022 12:07 am
muffins14 wrote: Fri May 06, 2022 9:45 am I'm sitting at like 115 / 80.
Is your target actually 100/40 (based on sig)? As in, you have 2x the bond exposure your strategic portfolio calls for?
More like 115/50 now for the goal. When yields had changed a lot I bought the ZT, which adds a lot of exposure but also is pretty low volatility, so it was somewhat impulsive of me to get it. Selling would mean a loss of $300 there and get me to 115/57. So, still over that previous decision, but it’s less bond risk than most in this thread at least

Overall I want a 70/30 portfolio, but leveraged
You are right on your new stock long-term target so that’sa nice coincidence, no need to rebalance that. But you’re pretty overweight bonds, that would definitely cross my rebalance band. What band do you use to rebalance?
muffins14
Posts: 5528
Joined: Wed Oct 26, 2016 4:14 am
Location: New York

Re: Modified versions of HFEA with ITT and Futures / Lifecycle Investing with Modern Portfolio Theory

Post by muffins14 »

Dry-Drink wrote: Sat May 07, 2022 9:19 am
You are right on your new stock long-term target so that’sa nice coincidence, no need to rebalance that. But you’re pretty overweight bonds, that would definitely cross my rebalance band. What band do you use to rebalance?
I’ll sell the ZT on Monday, and then with 3 months of contributions I would be at 115/52 if prices all remain the same, so I’ll probably just do that rather than sell another futures contract, and I’ll re-evaluate then.

5% bands for rebalancing is the idea
Crom laughs at your Four Winds
Topic Author
skierincolorado
Posts: 2377
Joined: Sat Mar 21, 2020 10:56 am

Re: Modified versions of HFEA with ITT and Futures / Lifecycle Investing with Modern Portfolio Theory

Post by skierincolorado »

muffins14 wrote: Sat May 07, 2022 9:58 am
Dry-Drink wrote: Sat May 07, 2022 9:19 am
You are right on your new stock long-term target so that’sa nice coincidence, no need to rebalance that. But you’re pretty overweight bonds, that would definitely cross my rebalance band. What band do you use to rebalance?
I’ll sell the ZT on Monday, and then with 3 months of contributions I would be at 115/52 if prices all remain the same, so I’ll probably just do that rather than sell another futures contract, and I’ll re-evaluate then.

5% bands for rebalancing is the idea
Is the 115/75 duration weighted?
muffins14
Posts: 5528
Joined: Wed Oct 26, 2016 4:14 am
Location: New York

Re: Modified versions of HFEA with ITT and Futures / Lifecycle Investing with Modern Portfolio Theory

Post by muffins14 »

No, it’s currently 1ZT, 2 ZF, 2 ZN, 1TN
Crom laughs at your Four Winds
BayStater
Posts: 97
Joined: Sun Mar 29, 2020 11:57 pm

Re: Modified versions of HFEA with ITT and Futures / Lifecycle Investing with Modern Portfolio Theory

Post by BayStater »

If I were not able to use futures, to maintain the 3:1 ratio with levered ETFs the most I could do is either

50% VOO and 50% TYD (150% ITT)
or 45% VOO and 55% TYA (137% ITT)

Which isn't enough stock exposure to be comparable with HFEA. So without futures, this strategy is inaccessible?
muffins14
Posts: 5528
Joined: Wed Oct 26, 2016 4:14 am
Location: New York

Re: Modified versions of HFEA with ITT and Futures / Lifecycle Investing with Modern Portfolio Theory

Post by muffins14 »

You could use:

Box spread financing
Portfolio margin
Leveraged ETFs
Futures
Crom laughs at your Four Winds
Topic Author
skierincolorado
Posts: 2377
Joined: Sat Mar 21, 2020 10:56 am

Re: Modified versions of HFEA with ITT and Futures / Lifecycle Investing with Modern Portfolio Theory

Post by skierincolorado »

BayStater wrote: Mon May 09, 2022 8:30 am If I were not able to use futures, to maintain the 3:1 ratio with levered ETFs the most I could do is either

50% VOO and 50% TYD (150% ITT)
or 45% VOO and 55% TYA (137% ITT)

Which isn't enough stock exposure to be comparable with HFEA. So without futures, this strategy is inaccessible?
You could use UPRO. 55% UPRO 45% TYA would be pretty similar to HFEA but improved. Regular HFEA has roo much exposure to bonds and the bonds that it does have exposure to have poor risk adjusted returns. Replacing TMF with TYD the two strategies have the same return from 1955 to 2011, but the drawdown on the tyd portfolio is much much less. There is no benefit for all the extra bond risk of tmf.

You can also use box spread financing if this is a taxable account.
BayStater
Posts: 97
Joined: Sun Mar 29, 2020 11:57 pm

Re: Modified versions of HFEA with ITT and Futures / Lifecycle Investing with Modern Portfolio Theory

Post by BayStater »

skierincolorado wrote: Mon May 09, 2022 9:29 am You could use UPRO. 55% UPRO 45% TYA would be pretty similar to HFEA but improved. Regular HFEA has roo much exposure to bonds and the bonds that it does have exposure to have poor risk adjusted returns. Replacing TMF with TYD the two strategies have the same return from 1955 to 2011, but the drawdown on the tyd portfolio is much much less. There is no benefit for all the extra bond risk of tmf.
I'm definitely interested in the efficiency of the middle of the curve. I found that paper by PIMCO on the original post really interesting.

Isn't the argument that TYD won't have enough oomph to balance out drawdowns in UPRO? This is why mHFEA takes on additional leverage in ITTs via futures (over what's available through TYD or TYA).
Topic Author
skierincolorado
Posts: 2377
Joined: Sat Mar 21, 2020 10:56 am

Re: Modified versions of HFEA with ITT and Futures / Lifecycle Investing with Modern Portfolio Theory

Post by skierincolorado »

BayStater wrote: Mon May 09, 2022 11:52 am
skierincolorado wrote: Mon May 09, 2022 9:29 am You could use UPRO. 55% UPRO 45% TYA would be pretty similar to HFEA but improved. Regular HFEA has roo much exposure to bonds and the bonds that it does have exposure to have poor risk adjusted returns. Replacing TMF with TYD the two strategies have the same return from 1955 to 2011, but the drawdown on the tyd portfolio is much much less. There is no benefit for all the extra bond risk of tmf.
I'm definitely interested in the efficiency of the middle of the curve. I found that paper by PIMCO on the original post really interesting.

Isn't the argument that TYD won't have enough oomph to balance out drawdowns in UPRO? This is why mHFEA takes on additional leverage in ITTs via futures (over what's available through TYD or TYA).
If you are just looking at 1982 to 2020 the correct amount of oomph is probably closer to tmf than tyd. But that is a period of exceptional bond returns. If you look at a broader range of economic scenarios, tyd is plenty of oomph.

As I said, a portfolio beginning in 1955 using tyd was ahead of one using tmf until 2011. I think that's pretty clear evidence that tyd had plenty of oomph.

What we really care about is the efficient frontier... which is the more specific term for what you are getting at with the term oomph. Tmf was only on the efficient frontier 1982 to 2020. While I don't expect bonds to look as bad as they did before 1982, probably not as good as pre 2021 either. Bonds probably were not on the efficient frontier at all from 1955 to 1982. If we look at the full period the efficient frontier would use tyd not tmf.

This all assumes we can only use letf. If you use futures, the efficient frontier 1982 to present was ITT but more than you could get from tyd. Which is why hfea used tmf.
User avatar
hiddenpower
Posts: 564
Joined: Tue Nov 17, 2020 11:24 pm

Re: Modified versions of HFEA with ITT and Futures / Lifecycle Investing with Modern Portfolio Theory

Post by hiddenpower »

skierincolorado wrote: Mon May 09, 2022 1:07 pm
BayStater wrote: Mon May 09, 2022 11:52 am
skierincolorado wrote: Mon May 09, 2022 9:29 am You could use UPRO. 55% UPRO 45% TYA would be pretty similar to HFEA but improved. Regular HFEA has roo much exposure to bonds and the bonds that it does have exposure to have poor risk adjusted returns. Replacing TMF with TYD the two strategies have the same return from 1955 to 2011, but the drawdown on the tyd portfolio is much much less. There is no benefit for all the extra bond risk of tmf.
I'm definitely interested in the efficiency of the middle of the curve. I found that paper by PIMCO on the original post really interesting.

Isn't the argument that TYD won't have enough oomph to balance out drawdowns in UPRO? This is why mHFEA takes on additional leverage in ITTs via futures (over what's available through TYD or TYA).
If you are just looking at 1982 to 2020 the correct amount of oomph is probably closer to tmf than tyd. But that is a period of exceptional bond returns. If you look at a broader range of economic scenarios, tyd is plenty of oomph.

As I said, a portfolio beginning in 1955 using tyd was ahead of one using tmf until 2011. I think that's pretty clear evidence that tyd had plenty of oomph.

What we really care about is the efficient frontier... which is the more specific term for what you are getting at with the term oomph. Tmf was only on the efficient frontier 1982 to 2020. While I don't expect bonds to look as bad as they did before 1982, probably not as good as pre 2021 either. Bonds probably were not on the efficient frontier at all from 1955 to 1982. If we look at the full period the efficient frontier would use tyd not tmf.

This all assumes we can only use letf. If you use futures, the efficient frontier 1982 to present was ITT but more than you could get from tyd. Which is why hfea used tmf.
So how do we keep a pulse on what's the current efficient frontier? It sounds like it's always in flux
Topic Author
skierincolorado
Posts: 2377
Joined: Sat Mar 21, 2020 10:56 am

Re: Modified versions of HFEA with ITT and Futures / Lifecycle Investing with Modern Portfolio Theory

Post by skierincolorado »

hiddenpower wrote: Thu May 12, 2022 3:19 pm
skierincolorado wrote: Mon May 09, 2022 1:07 pm
BayStater wrote: Mon May 09, 2022 11:52 am
skierincolorado wrote: Mon May 09, 2022 9:29 am You could use UPRO. 55% UPRO 45% TYA would be pretty similar to HFEA but improved. Regular HFEA has roo much exposure to bonds and the bonds that it does have exposure to have poor risk adjusted returns. Replacing TMF with TYD the two strategies have the same return from 1955 to 2011, but the drawdown on the tyd portfolio is much much less. There is no benefit for all the extra bond risk of tmf.
I'm definitely interested in the efficiency of the middle of the curve. I found that paper by PIMCO on the original post really interesting.

Isn't the argument that TYD won't have enough oomph to balance out drawdowns in UPRO? This is why mHFEA takes on additional leverage in ITTs via futures (over what's available through TYD or TYA).
If you are just looking at 1982 to 2020 the correct amount of oomph is probably closer to tmf than tyd. But that is a period of exceptional bond returns. If you look at a broader range of economic scenarios, tyd is plenty of oomph.

As I said, a portfolio beginning in 1955 using tyd was ahead of one using tmf until 2011. I think that's pretty clear evidence that tyd had plenty of oomph.

What we really care about is the efficient frontier... which is the more specific term for what you are getting at with the term oomph. Tmf was only on the efficient frontier 1982 to 2020. While I don't expect bonds to look as bad as they did before 1982, probably not as good as pre 2021 either. Bonds probably were not on the efficient frontier at all from 1955 to 1982. If we look at the full period the efficient frontier would use tyd not tmf.

This all assumes we can only use letf. If you use futures, the efficient frontier 1982 to present was ITT but more than you could get from tyd. Which is why hfea used tmf.
So how do we keep a pulse on what's the current efficient frontier? It sounds like it's always in flux
It is always different depending on what time period in history or in the future.

Look at a longer period of history.
Look for a durable ef that doesn't flunk other periods of history.
Make predictions of the underlying variables like return and risk.
Generally speaking owning more than one asset type will be more efficient even if you are not exactly on the ef. As long as the asset has returns above the risk free rate.
Being exactly on the ef in the future would be incredibly lucky.
comeinvest
Posts: 2709
Joined: Mon Mar 12, 2012 6:57 pm

Re: Modified versions of HFEA with ITT and Futures / Lifecycle Investing with Modern Portfolio Theory

Post by comeinvest »

The 20-year treasury future contract was removed from the treasury analytics page?! https://www.cmegroup.com/tools-informat ... ytics.html
unemployed_pysicist
Posts: 220
Joined: Sat Oct 09, 2021 2:32 pm
Location: Amsterdam

Re: Modified versions of HFEA with ITT and Futures / Lifecycle Investing with Modern Portfolio Theory

Post by unemployed_pysicist »

comeinvest wrote: Fri May 13, 2022 7:31 pm The 20-year treasury future contract was removed from the treasury analytics page?! https://www.cmegroup.com/tools-informat ... ytics.html
I see it now listed as 20 year, right underneath the 3 year note contract.
couldn't afford the h | | BUY BONDS | WEAR DIAMONDS
comeinvest
Posts: 2709
Joined: Mon Mar 12, 2012 6:57 pm

Re: Modified versions of HFEA with ITT and Futures / Lifecycle Investing with Modern Portfolio Theory

Post by comeinvest »

unemployed_pysicist wrote: Sat May 14, 2022 1:38 am
comeinvest wrote: Fri May 13, 2022 7:31 pm The 20-year treasury future contract was removed from the treasury analytics page?! https://www.cmegroup.com/tools-informat ... ytics.html
I see it now listed as 20 year, right underneath the 3 year note contract.
You are right. I could swear it was gone yesterday. But why not order by maturity range, where they had the 20y before?!
unemployed_pysicist
Posts: 220
Joined: Sat Oct 09, 2021 2:32 pm
Location: Amsterdam

Re: Modified versions of HFEA with ITT and Futures / Lifecycle Investing with Modern Portfolio Theory

Post by unemployed_pysicist »

I also remember seeing the 20 year down with the other longer maturities.

I guess they are doing some work on the website, but I don't understand why they put the 20 year with the shorter, note contracts now. I thought they used to order the contracts by increasing maturity.
couldn't afford the h | | BUY BONDS | WEAR DIAMONDS
User avatar
hiddenpower
Posts: 564
Joined: Tue Nov 17, 2020 11:24 pm

Re: Modified versions of HFEA with ITT and Futures / Lifecycle Investing with Modern Portfolio Theory

Post by hiddenpower »

Hey Skier, I was going to PM this but figured might as well put it in the open.

With inflation and rates rising, it sounds like normal HFEA is way more risk than one should take on. I'm wondering how ITTs faired or if one should just hold off all-together for now. Originally I was thinking of using my Roth (10% NW) for HFEA and my traditional IRA (20% NW) for PSLDX. But I found a post of yours mentioning the outperformance of ITTs and lower bond usage during the 70s and thus am considering just going with something like 55 UPRO / 45 TYA between the entire two accounts.

Do by chance have drawdowns for something like a 55 UPRO / 45 TYA (165 S&P / 112.5 ITT) config or 45 UPRO / 55 TYA (165 S&P / 123.75 ITT) config that goes back through the 70s and 80s? Seeing how this all faired would be illuminating. I could see myself using the normal HFEA strat during less uncertain times.

TYA is a 2.5x quarterly rebalanced fund targeting exposure of TLT. It seems to track pretty close to 2.5X IEF.

I found Simba's spreadsheet and started using it for the first time but in google sheet's I'm getting loads of NAs and can't figure out how to get backtest >= 100% equities and see drawdowns for this period.
peturano
Posts: 7
Joined: Fri May 21, 2021 4:18 pm

Re: Modified versions of HFEA with ITT and Futures / Lifecycle Investing with Modern Portfolio Theory

Post by peturano »

Image

But note that this is an annual rebalancing. I don't know how to do a quarterly rebalancing backtest.
comeinvest
Posts: 2709
Joined: Mon Mar 12, 2012 6:57 pm

Re: Modified versions of HFEA with ITT and Futures / Lifecycle Investing with Modern Portfolio Theory

Post by comeinvest »

FYI: I just noticed that Google Finance now has equity index futures and treasury futures charts going back to 2004.

I'm curious what precisely the futures roll stitching method is.

I think Google Finance does not have total return charts of stocks and ETFs, which means comparative analysis is limited.
unemployed_pysicist
Posts: 220
Joined: Sat Oct 09, 2021 2:32 pm
Location: Amsterdam

Re: Modified versions of HFEA with ITT and Futures / Lifecycle Investing with Modern Portfolio Theory

Post by unemployed_pysicist »

At a glance, the T bond futures data from Google finance looks similar to the continuous futures from yahoo finance.

When I last looked at the yahoo finance data, the pattern in the volume seemed to suggest that it tracks the front month almost all the way to the last trading day. The volume is large right before the delivery month and then there is very little volume during the delivery month. The volume picks up dramatically after the last trading date in the delivery month, after which it goes back to a high level. So the contract does not appear to roll when everyone else does, and always tracks the nearest ctd.

I have been assuming that this means that the price gets stitched at the last trading date of the front month, but I'm not entirely sure. If someone could support or refute this, that would be great since I would also like to know how yahoo/Google stitches their continuous futures.
couldn't afford the h | | BUY BONDS | WEAR DIAMONDS
User avatar
LTCM
Posts: 412
Joined: Wed Sep 09, 2020 3:58 am

Re: Modified versions of HFEA with ITT and Futures / Lifecycle Investing with Modern Portfolio Theory

Post by LTCM »

Is there a updated version of millennialmillions spreadsheet with 2022 YTD data (or close to it)?

Or is there a quick and easy way to update it using Simba's?

Thanks
55% VUG - 20% VEA - 20% EDV - 5% BNDX
comeinvest
Posts: 2709
Joined: Mon Mar 12, 2012 6:57 pm

Re: Modified versions of HFEA with ITT and Futures / Lifecycle Investing with Modern Portfolio Theory

Post by comeinvest »

Some nice anticorrelation lately. Unfortunately I didn't rebalance, although the fluctuations would have hit 5% stock/bond rebalance bands I think.

Image

Image
Investing Lawyer
Posts: 51
Joined: Wed Apr 08, 2020 9:35 pm

Re: Modified versions of HFEA with ITT and Futures / Lifecycle Investing with Modern Portfolio Theory

Post by Investing Lawyer »

Since I recently discovered that yahoo finance offers a continuous future contract variant from a post on this forum, I wanted to see how closely bond fund returns tracked the futures returns. I compared the weekly log returns for the past 1 year of PRTIX (intermediate treasury fund, ~6.25 duration) and ZN (currently 6.290 duration). The correlation was a hair over 97%. With this, i'm fairly comfortable using bond funds in portfolio visualizer as substitutes as the futures.

P.S. If weekly returns from one year appear an odd selection, it's mostly because a five year time series on the futures with monthly returns had a decent amount of missing data, I was too lazy to try to clean and match it up up. And, apparently yahoo finance doesn't store historical data for UB to cross check TLT as well.
Topic Author
skierincolorado
Posts: 2377
Joined: Sat Mar 21, 2020 10:56 am

Re: Modified versions of HFEA with ITT and Futures / Lifecycle Investing with Modern Portfolio Theory

Post by skierincolorado »

Investing Lawyer wrote: Thu Jun 02, 2022 9:28 pm Since I recently discovered that yahoo finance offers a continuous future contract variant from a post on this forum, I wanted to see how closely bond fund returns tracked the futures returns. I compared the weekly log returns for the past 1 year of PRTIX (intermediate treasury fund, ~6.25 duration) and ZN (currently 6.290 duration). The correlation was a hair over 97%. With this, i'm fairly comfortable using bond funds in portfolio visualizer as substitutes as the futures.

P.S. If weekly returns from one year appear an odd selection, it's mostly because a five year time series on the futures with monthly returns had a decent amount of missing data, I was too lazy to try to clean and match it up up. And, apparently yahoo finance doesn't store historical data for UB to cross check TLT as well.
If you missed it there was a pretty thorough analysis of this sort of issue earlier in the thread that cites some papers and did a comparison of funds vs futures returns after standardizing by the variance of returns
Topic Author
skierincolorado
Posts: 2377
Joined: Sat Mar 21, 2020 10:56 am

Re: Modified versions of HFEA with ITT and Futures / Lifecycle Investing with Modern Portfolio Theory

Post by skierincolorado »

hiddenpower wrote: Thu May 26, 2022 12:13 pm Hey Skier, I was going to PM this but figured might as well put it in the open.

With inflation and rates rising, it sounds like normal HFEA is way more risk than one should take on. I'm wondering how ITTs faired or if one should just hold off all-together for now. Originally I was thinking of using my Roth (10% NW) for HFEA and my traditional IRA (20% NW) for PSLDX. But I found a post of yours mentioning the outperformance of ITTs and lower bond usage during the 70s and thus am considering just going with something like 55 UPRO / 45 TYA between the entire two accounts.

Do by chance have drawdowns for something like a 55 UPRO / 45 TYA (165 S&P / 112.5 ITT) config or 45 UPRO / 55 TYA (165 S&P / 123.75 ITT) config that goes back through the 70s and 80s? Seeing how this all faired would be illuminating. I could see myself using the normal HFEA strat during less uncertain times.

TYA is a 2.5x quarterly rebalanced fund targeting exposure of TLT. It seems to track pretty close to 2.5X IEF.

I found Simba's spreadsheet and started using it for the first time but in google sheet's I'm getting loads of NAs and can't figure out how to get backtest >= 100% equities and see drawdowns for this period.
I'd have to start by saying I object to the premise that anybody could have any skill in timing movement between HFEA and mHFEA. I'd also repeat what I've said before about looking at AA across entire net-worth. Bucketing leads to inefficiencies if you do not rebalance across the buckets. Holding a highly leveraged portfolio in one bucket and un/low-leveraged portfolio in another bucket is mathematically guaranteed to underperform and/or lead to increased risk exposure over time if you do not rebalance between the two buckets. You'd be killing yourself with more volatility decay than necessary. I see the psychological appeal of bucketing, but the costs are too great.

The main advantages of a mHFEA-like portfolio (55 UPRO 45 TYA/TYD) over a 1/3 HFEA and 2/3 PSLDX are:

1. Reduced interest rate risk. The latter has nearly twice the interst rate risk of the former. The former would perform better in a longer term backtest through a rising rate period.

2. Historically ITT have had higher return per unit of interest rate risk than the LTT used in HFEA and PSLDX.

To run leverage in Simba use the compare portfolio tab and enter allocations to 2x or 3x leveraged funds, or use unleveraged funds and short T-Bills (found near the top) by using a negative % allocation. If you're still having trouble I could run a scenario or two for you.

One other consideration is that fixed leverage strategies are not entirely consistent with lifecycle investing, which is one of the primary justifications for the use of leverage. This may be the most important consideration of all when choosing an IPS.
Last edited by skierincolorado on Fri Jun 03, 2022 9:55 am, edited 4 times in total.
User avatar
hiddenpower
Posts: 564
Joined: Tue Nov 17, 2020 11:24 pm

Re: Modified versions of HFEA with ITT and Futures / Lifecycle Investing with Modern Portfolio Theory

Post by hiddenpower »

skierincolorado wrote: Fri Jun 03, 2022 12:27 am
hiddenpower wrote: Thu May 26, 2022 12:13 pm Hey Skier, I was going to PM this but figured might as well put it in the open.

With inflation and rates rising, it sounds like normal HFEA is way more risk than one should take on. I'm wondering how ITTs faired or if one should just hold off all-together for now. Originally I was thinking of using my Roth (10% NW) for HFEA and my traditional IRA (20% NW) for PSLDX. But I found a post of yours mentioning the outperformance of ITTs and lower bond usage during the 70s and thus am considering just going with something like 55 UPRO / 45 TYA between the entire two accounts.

Do by chance have drawdowns for something like a 55 UPRO / 45 TYA (165 S&P / 112.5 ITT) config or 45 UPRO / 55 TYA (165 S&P / 123.75 ITT) config that goes back through the 70s and 80s? Seeing how this all faired would be illuminating. I could see myself using the normal HFEA strat during less uncertain times.

TYA is a 2.5x quarterly rebalanced fund targeting exposure of TLT. It seems to track pretty close to 2.5X IEF.

I found Simba's spreadsheet and started using it for the first time but in google sheet's I'm getting loads of NAs and can't figure out how to get backtest >= 100% equities and see drawdowns for this period.
I'd have to start by saying I object to the premise that anybody could have any skill in timing movement between HFEA and mHFEA. I'd also repeat what I've said before about looking at AA across entire net-worth. Bucketing leads to inefficiencies if you do not rebalance across the buckets. Holding a highly leveraged portfolio in one bucket and un/low-leveraged portfolio in another bucket is mathematically guaranteed to underperform and/or lead to increased risk exposure over time if you do not rebalance between the two buckets.

The main advantages of a mHFEA-like portfolio (55 UPRO 45 TYA/TYD) over a 1/3 HFEA and 2/3 PSLDX are:

1. Reduced interest rate risk. The latter has nearly twice the interst rate risk of the former. This would perform better in a longer term backtest.

2. Historically ITT have had higher return per unit of interest rate risk than the LTT used in HFEA and PSLDX.

To run leverage in Simba use the compare portfolio tab and enter allocations to 2x or 3x leveraged funds, or use unleveraged funds and short T-Bills (found near the top) by using a negative % allocation. If you're still having trouble I could run a scenario or two for you.

One other consideration is that fixed leverage strategies are not entirely consistent with lifecycle investing, which is one of the primary justifications for the use of leverage. This may be the most important consideration of all when choosing an IPS.
I would be super interested to see how some comparisons like how 55 UPRO / 45 TYA did in the 70s/80s etc, compared to something like a 55/45 UPRO/TMF and 70/30 UPRO/TMF split. Maybe less TMF would also be sufficient. Unfortunately I can't run simba in google sheets and don't have excel so it would be much appreciated.

I tried to test out 55 UPRO / 45 TMF using siamonds data with some hacked up code but my results differ from the typical inflation HFEA screenshot going around. Could have made a mistake but was wondering if anyone has run some tests on this themselves.
mmKay
Posts: 3
Joined: Sat Jun 04, 2022 6:51 pm

Re: Modified versions of HFEA with ITT and Futures / Lifecycle Investing with Modern Portfolio Theory

Post by mmKay »

Hi all, I'm a long time lurker. I should start off with a massive thank you to all the contributors to this thread (and the admins). I have learned a lot from it. :happy

I'm wondering if anyone using treasury futures for ITTs has considered adding short positions on other treasury futures (including longer term US and international) that are in contango (i.e., futures prices above cash / spot price) as a kind of a hedge? I saw comeinvest mentioned something similar a while back:
comeinvest wrote: Mon Nov 29, 2021 8:24 pm I happen to have a *short* position in German government bonds via the 2-year future (GBS) for a while. The German yield curve formed a trough in the 2-3y area for several months this year, which seemed too enticing to pass, but it's more playing than a proven strategy.
I have been experimenting with adding this to my mHFEA (which is a very small amount of my wealth). Last quarter I took short positions in ZB and long-term UK and Canada bonds, and that worked out as a nice hedge. This quarter I have short positions in Canadian long-term and German 30yr and 10yr.

Because these are often longer-term (at least at the moment) it seems like a bet against beta, which I see has been mentioned on this thread a few times.

It seems to have gone well for me so far, but it's early days and I guess in a bond bull market it would feel like a real drag. Because these positions are in contango, though, shouldn't there still be a positive expected return for short positions, based on negative roll yield? I'm still very new to this and keen to learn more. I know the risk is of a yield curve flattening, but I'm wondering if I'm missing something else from those who are more knowledgeable than I am.

EDIT: I reduced my exposure this strategy of shorting treasuries that are in contango as a hedge after I saw a post from skierincolorado in the HFEA thread late last year saying that expected changes in future yields could also be priced into current futures prices, which I guess could be a cause of this contango. Having this extra unknown variable makes this strategy seem less likely to work. I'm keen to learn more about futures pricing and what causes this contango, but it seems like the way the academic literature goes of calculating carry (as outlined in this thread) is a better way to go if one wants to take short positions.
Last edited by mmKay on Sun Jul 31, 2022 1:30 am, edited 1 time in total.
comeinvest
Posts: 2709
Joined: Mon Mar 12, 2012 6:57 pm

Re: Modified versions of HFEA with ITT and Futures / Lifecycle Investing with Modern Portfolio Theory

Post by comeinvest »

mmKay wrote: Sat Jun 04, 2022 7:34 pm Hi all, I'm a long time lurker. I should start off with a massive thank you to all the contributors to this thread (and the admins). I have learned a lot from it. :happy

I'm wondering if anyone using treasury futures for ITTs has considered adding short positions on other treasury futures (including longer term US and international) that are in contango (i.e., futures prices above cash / spot price) as a kind of a hedge? I saw comeinvest mentioned something similar a while back:
comeinvest wrote: Mon Nov 29, 2021 8:24 pm I happen to have a *short* position in German government bonds via the 2-year future (GBS) for a while. The German yield curve formed a trough in the 2-3y area for several months this year, which seemed too enticing to pass, but it's more playing than a proven strategy.
I have been experimenting with adding this to my mHFEA (which is a very small amount of my wealth). Last quarter I took short positions in ZB and long-term UK and Canada bonds, and that worked out as a nice hedge. This quarter I have short positions in Canadian long-term and German 30yr and 10yr.

Because these are often longer-term (at least at the moment) it seems like a bet against beta, which I see has been mentioned on this thread a few times.

It seems to have gone well for me so far, but it's early days and I guess in a bond bull market it would feel like a real drag. Because these positions are in contango, though, shouldn't there still be a positive expected return for short positions, based on negative roll yield? I'm still very new to this and keen to learn more. I know the risk is of a yield curve flattening, but I'm wondering if I'm missing something else from those who are more knowledgeable than I am.
I'm not sure if I understand your strategy. I'm also not sure if you understand the pricing of treasury futures. I'm not sure if "contango" is a signal of mispricing or trading opportunity.

I currently still have positions that bet in a duration-neutral way on the disappearance of the current zig-zag in the U.S. yield curve. (In short words, short the troughs and long the peaks.) I got burned so far, as the zig-zag intensified. So if you start a position now, you might get some good risk/return premium as the zig-zag is quite big and not very common historically. My strategy is an overlay over mHFEA in hope to extract a risk premium from anomalies in the slope of the curve that is relatively independent from both equity and interest rate risk (interest rate risk = parallel shifts of the yield curve). The more independent sources of return the better.

You can also read the papers on carry return investing on the yield curve. You could backtest a long/short strategy as an overlay or extension of the mHFEA backtesting. mHFEA suggests a preference for ITT vs LTT, but is basically limited to long futures positions. We could generalize the optimization problem by allowing both long and short positions on the yield curve (for example long ZF, short UB as UB is hated in the mHFEA thread as it historically had lower return per duration risk than ZF). Nothing indicates that the optimal solution of the optimization problem with only positive positions (for example 170% ITT / 140% equities) is the optimal global solution that allows short positions.
User avatar
LTCM
Posts: 412
Joined: Wed Sep 09, 2020 3:58 am

Re: Modified versions of HFEA with ITT and Futures / Lifecycle Investing with Modern Portfolio Theory

Post by LTCM »

comeinvest wrote: Sun Jun 05, 2022 4:43 am You could backtest a long/short strategy as an overlay or extension of the mHFEA backtesting.
This fits my obsession with round numbers and is very on brand for my user name...

100% Stocks
250% LTT
2500% STT
-1250% ITT
-1500% T-Bills

Since 1871...

Portfolio CAGR: 13.74%
Max. Drawdown: 58.19%
Sharpe: 0.48
Ulcer Perf: 0.91
US Market correlation: 0.70

What could possibly go wrong?!
55% VUG - 20% VEA - 20% EDV - 5% BNDX
comeinvest
Posts: 2709
Joined: Mon Mar 12, 2012 6:57 pm

Re: Modified versions of HFEA with ITT and Futures / Lifecycle Investing with Modern Portfolio Theory

Post by comeinvest »

LTCM wrote: Sun Jun 05, 2022 5:50 pm
comeinvest wrote: Sun Jun 05, 2022 4:43 am You could backtest a long/short strategy as an overlay or extension of the mHFEA backtesting.
This fits my obsession with round numbers and is very on brand for my user name...

100% Stocks
250% LTT
2500% STT
-1250% ITT
-1500% T-Bills

Since 1871...

Portfolio CAGR: 13.74%
Max. Drawdown: 58.19%
Sharpe: 0.48
Ulcer Perf: 0.91
US Market correlation: 0.70

What could possibly go wrong?!
Not sure if you actually tested something or if you are joking, but if you have long and short positions on the treasury yield curve, you can have very large positions magnitudes higher than your NAV, that have quite low total fluctuations / risk, as the duration exposures cancel each other out. You can also get margin offsets at CME.
DMoogle
Posts: 549
Joined: Sat Oct 31, 2020 10:24 pm

Re: Modified versions of HFEA with ITT and Futures / Lifecycle Investing with Modern Portfolio Theory

Post by DMoogle »

LTCM wrote: Sun Jun 05, 2022 5:50 pm
comeinvest wrote: Sun Jun 05, 2022 4:43 am You could backtest a long/short strategy as an overlay or extension of the mHFEA backtesting.
This fits my obsession with round numbers and is very on brand for my user name...

100% Stocks
250% LTT
2500% STT
-1250% ITT
-1500% T-Bills

Since 1871...

Portfolio CAGR: 13.74%
Max. Drawdown: 58.19%
Sharpe: 0.48
Ulcer Perf: 0.91
US Market correlation: 0.70

What could possibly go wrong?!
Taking you seriously, spreads+commissions on treasury futures would actually start to have a noticeable effect. IIRC it ends up being like 0.05% expense annualized (might be off on this). With that much leverage, you're going to be looking at something like 2% expense/annum.
User avatar
LTCM
Posts: 412
Joined: Wed Sep 09, 2020 3:58 am

Re: Modified versions of HFEA with ITT and Futures / Lifecycle Investing with Modern Portfolio Theory

Post by LTCM »

DMoogle wrote: Mon Jun 06, 2022 8:43 am
Taking you seriously, spreads+commissions on treasury futures would actually start to have a noticeable effect. IIRC it ends up being like 0.05% expense annualized (might be off on this). With that much leverage, you're going to be looking at something like 2% expense/annum.
Do you have specific numbers for those commissions? I can plug them in to the back test sheet just to see what happens.

Looking at IB commissions and assuming a 100k account I get approx $15 per year for the 250% LTT. $85 for the STT. $75 for the ITT. So $175 total which is only 0.175 ER for the entire scheme. (T-bill cost is already factored in but VTI ER would need adding).

I thought it was interesting that ITT needs shorting to get the max Sharpe on a short/long treasury position. Is it even possible to short ITT in a tax deferred account or would that need to be taxable?
55% VUG - 20% VEA - 20% EDV - 5% BNDX
User avatar
LTCM
Posts: 412
Joined: Wed Sep 09, 2020 3:58 am

Re: Modified versions of HFEA with ITT and Futures / Lifecycle Investing with Modern Portfolio Theory

Post by LTCM »

comeinvest wrote: Mon Jun 06, 2022 5:44 am Not sure if you actually tested something or if you are joking, but if you have long and short positions on the treasury yield curve, you can have very large positions magnitudes higher than your NAV, that have quite low total fluctuations / risk, as the duration exposures cancel each other out. You can also get margin offsets at CME.
I back tested it. That's where I got the returns and Sharpe ratios from. I was hoping to be able to get to 0.5 Sharpe over the 150 year period but couldn't quite manage it. I was sticking to 100% stocks and only trying to manipulate the short/long Treasury position though.

I'm not joking about being interested but 'm not about to actually do it because it's terrifying.

The returns on the position since 2020 are mildly amusing. I'll post a link when I'm back at my computer.
55% VUG - 20% VEA - 20% EDV - 5% BNDX
Post Reply