Riding HEDGEFUNDIE’s excellent adventure

CanaBogle24 · Post by **CanaBogle24** » Tue Sep 29, 2020 5:19 am

Hi All,

I’m one of the many that have been very closely following the extensive HEDGEFUNDIE threads (Part I and Part II). The discussion has been tremendous and is some of the more intriguing content I’ve seen on this site – thank you to everyone who’s contributed!

However, while those threads have a technical bent, discussing relative merits of different approaches, views on a wide range of risks, the infinite potential modificaitons to the strategy, etc (and I'll certainly continue to follow Part II), what I’m proposing here is for folks who’ve decided to make it real to share their approach and results.

I’m starting here with a proposed template:

Start date: 9/28/20
Approach: Fixed allocation - 33/32/35 UPRO/TQQQ/TMF
Rebalancing frequency: Quarterly
Return (total / YTD): 0.79% / 0.79%

Initial contribution: $48k (~10% of invested assets)
Additional contributions: $1,500 / month (committed for at least first year, then re-evaluate)
Portfolio location: Taxable, M1Finance

Obviously, feel free to share as much or as little as you’d like. Would also love to hear about the emotional experience – were you able to hold steady through market fluctuations? Did you start, then decide to pull anything off the table?

Hope this’ll serve as a way to track how many of us have some skin in the game (ahem, adventure), observe the various permutations' performance, all while sharing in the thrills and commiserating through the inevitable dark days

Post by **LadyGeek** » Tue Sep 29, 2020 8:09 am

Thanks for starting the thread. Let me add up-front that this approach is intended for investing enthusiasts who have the ability, willingness, and need to take risk.

If you want to proceed, the advice is to use no more than 5% of your portfolio. Why? That's how much you can afford to lose and not ruin your retirement. This is your life's savings here, tread carefully.

To help new investors, I encourage members to post how much of their total portfolio is used for this approach. (There's nothing wrong with a simple 3-fund portfolio...)

The OP is riding this adventure with ~10% of invested assets (initial contribution).

GeraniumLover · Post by **GeraniumLover** » Tue Sep 29, 2020 8:13 am

Start date: 2/6/20
Approach: Fixed allocation - 55/45 UPRO/TMF
Rebalancing frequency: Quarterly
Return (total / YTD): 10.33% / 10.33%
Initial contribution: $1k
Additional contributions: future HSA contributions, if any
Portfolio location: HSA, Fidelity

coingaroo · Post by **coingaroo** » Tue Sep 29, 2020 9:37 am

Start date: 17/07/20
Approach: Fixed allocation - 60/40 TQQQ/TMF
Rebalancing frequency: With new contributions, plus discretionary rebalancing during high-vol periods
Return (total / YTD): 6.81% / 6.81%
Initial contribution: $25k
Additional contributions: ~$38k so far, plus on-going ad-hoc contributions
Portfolio location: IBKR

Volkl_One · Post by **Volkl_One** » Tue Sep 29, 2020 10:02 am

Start date: 2/28/20
Approach: Fixed allocation - 55/45 UPRO/TMF
Rebalancing frequency: Quarterly
Return (total / YTD): 17.11% / 17.11%
Initial contribution: $38k (~33% of invested assets)
Additional contributions: $6,000 / year (committed for at least first year, then re-evaluate)
Portfolio location: Roth, Charles Schwab

RovenSkyfall · Post by **RovenSkyfall** » Tue Sep 29, 2020 10:11 am

Start date: 9/4/2020
Approach: Fixed allocation - 55/45 UPRO/TMF (May transition to Q1mo MV)
Rebalancing frequency: Quarterly
Return (total / YTD): -3.3%/-3.3%
Initial contribution: $9k (4% of invested assets)
Additional contributions: None
Portfolio location: Fidelity Roth

rangerrick9211 · Post by **rangerrick9211** » Tue Sep 29, 2020 10:21 am

Revised adventure as TMF is not permitted due to employer independence reasons.

Start date: 09/2019
Approach: Fixed allocation - 35/20/45, UPRO/TQQQ/UBT
Rebalancing frequency: None - rebalance through DCA for now, but TQQQ needs to be reset
Return (total / YTD): 71.44%/31.75%
Initial contribution: $15k - 1.5% of assets
Additional contributions: $2,500/mo. (I accelerated through the dip)
Portfolio location: IB - Taxable

JamesDagan · Post by **JamesDagan** » Tue Sep 29, 2020 11:00 am

Might I suggest adding S&P 500 return over the equivalent duration for comparison?

Start date: 4/20/2020
Approach: Fixed allocation - 55/45 UPRO/TMF
Rebalancing frequency: Quarterly
Return (total / YTD): 26.83%
S&P 500 Return (total / YTD): 17.75%
Initial contribution: $13k (<5%)
Additional contributions: N/A - pending future evaluation
Portfolio location: Mega Backdoor Roth

HawkeyePierce · Post by **HawkeyePierce** » Tue Sep 29, 2020 11:30 am

Start date: 8/15/2019
Approach: Fixed allocation - 57/43 EDV/UPRO
Rebalancing frequency: Quarterly
Return (total / YTD): 48% / 31%
Initial contribution: $31k (<10%)
Additional contributions: $0
Portfolio location: Roth IRA, M1Finance

SaintExupery · Post by **SaintExupery** » Tue Sep 29, 2020 4:31 pm

Start date: 9/28/2020
Approach: Fixed allocation - 55/45 UPRO/TMF
Rebalancing frequency: Quarterly
Return (total / YTD): -0.46% / -0.46%
Initial contribution: $85k (~5% of invested assets)
Additional contributions: None
Portfolio location: Roth IRA, M1 Finance

rchmx1 · Post by **rchmx1** » Tue Sep 29, 2020 7:13 pm

I'd like to participate in this thread, even though I'd say it's more that I'm using HF's EA to inform my own EA with LEFTs, at least for the time being. There's still just so much volatility, profit taking when there's a pop and waiting to buy the dip, that I haven't yet felt comfortable sticking to a fixed AA (I also feel unsure about TMF or alternatives, which has kept me from accepting a fixed AA up to this point).

Start date(technically speaking, when I first dipped my toe in with a bit of UPRO): 1/21/20
Approach: AAA between UPRO/TQQQ/TMF/Cash (yes yes, mostly simple attempts at market timing, please put away the tar and feathers)
Rebalancing frequency: N/A
Return (total / YTD): 31.01%
Initial contribution: $611.68 (just that first initial toe dipping)
Additional contributions: $15,110.73 (added with an attempt towards opportunism to take advantage of market movements) (~20% of total portfolio)
Portfolio location: Individual 401k and taxable, Charles Schwab

physixfan · Post by **physixfan** » Tue Sep 29, 2020 8:18 pm

Start date: 9/1/19
Approach: Adaptive Allocation, which means the percentages of UPRO/TMF are adjusted according to their last 20 day's volatility.
Rebalancing frequency: Monthly
Performance: I'm using xueqiu.com (a Chinese website) to track my progress: https://xueqiu.com/P/ZH2021590. You can just ignore all the Chinese characters and only look at the chart. Currently it's near a 1 year mark, and it's up about 33%.

Uncorrelated · Post by **Uncorrelated** » Wed Sep 30, 2020 5:06 am

I mentioned this before, but I feel it's worth repeating.

HEDGEFUNDIE's adventure is based on reaching the highest CAGR with the combination of two funds: UPRO and TMF. After much debate, the asset allocation was settled on 55/45. This observation is valid in principle: if it is your goal to have the highest expected CAGR, then this asset allocation is probably close to optimal.

However, these observations start to fall apart if you start mixing HFEA with other assets. If you have an asset allocation of 50% total stock market and 50% UPRO/TMF, then the optimal ratio is not 55/45. No backtests have been performed by HEDGEFUNDIE (or anyone else, to my knowledge) to estimate the optimal ratio of UPRO/TMF when combined with other assets. My mean variance analyzer (If you're short on time, see the third image in that thread) can be used to estimate the correct asset allocation, it is clearly visible that any given ratio between UPRO/TMF is only optimal at a single point on the efficient frontier.

The 55/45 asset allocation is only valid if that is the only thing in your entire portfolio. If your portfolio contains other things, you should run new backtests or other calculations to determine the correct ratio in your circumstances. Otherwise, the portfolio will be inefficient (that means there are other portfolio's with the same variance but higher expected return).

(note: observations above, including the limited optimality of HEDGEFUNDIE's 55/45 AA, assume you have a human capital of exactly zero, have no real assets, will never receive social security, have no DB pension plan, spend no money, reject factor theory, and pay no taxes. All of which will affect the correct asset allocation to use).

keith6014 · Post by **keith6014** » Wed Sep 30, 2020 7:31 am

physixfan wrote: ↑Tue Sep 29, 2020 8:18 pm Start date: 9/1/19
Approach: Adaptive Allocation, which means the percentages of UPRO/TMF are adjusted according to their last 20 day's volatility.
Rebalancing frequency: Monthly
Performance: I'm using xueqiu.com (a Chinese website) to track my progress: https://xueqiu.com/P/ZH2021590. You can just ignore all the Chinese characters and only look at the chart. Currently it's near a 1 year mark, and it's up about 33%.

Great website. Is there an English equivalence? I am shocked there aren't that many good portfolio performance tracking tools.

Tingting1013 · Post by **Tingting1013** » Wed Sep 30, 2020 10:56 am

Uncorrelated wrote: ↑Wed Sep 30, 2020 5:06 am I mentioned this before, but I feel it's worth repeating.

HEDGEFUNDIE's adventure is based on reaching the highest CAGR with the combination of two funds: UPRO and TMF. After much debate, the asset allocation was settled on 55/45. This observation is valid in principle: if it is your goal to have the highest expected CAGR, then this asset allocation is probably close to optimal.

However, these observations start to fall apart if you start mixing HFEA with other assets. If you have an asset allocation of 50% total stock market and 50% UPRO/TMF, then the optimal ratio is not 55/45. No backtests have been performed by HEDGEFUNDIE (or anyone else, to my knowledge) to estimate the optimal ratio of UPRO/TMF when combined with other assets. My mean variance analyzer (If you're short on time, see the third image in that thread) can be used to estimate the correct asset allocation, it is clearly visible that any given ratio between UPRO/TMF is only optimal at a single point on the efficient frontier.

The 55/45 asset allocation is only valid if that is the only thing in your entire portfolio. If your portfolio contains other things, you should run new backtests or other calculations to determine the correct ratio in your circumstances. Otherwise, the portfolio will be inefficient (that means there are other portfolio's with the same variance but higher expected return).

(note: observations above, including the limited optimality of HEDGEFUNDIE's 55/45 AA, assume you have a human capital of exactly zero, have no real assets, will never receive social security, have no DB pension plan, spend no money, reject factor theory, and pay no taxes. All of which will affect the correct asset allocation to use).

I think what you’re missing is that the HFEA is an experiment with existential risks, with behavioral and downside protection value in bucketing it off from the rest of one’s portfolio.

ImUrHuckleberry · Post by **ImUrHuckleberry** » Wed Sep 30, 2020 11:21 am

LadyGeek wrote: ↑Tue Sep 29, 2020 8:09 am Thanks for starting the thread. Let me add up-front that this approach is intended for investing enthusiasts who have the ability, willingness, and need to take risk.

If you want to proceed, the advice is to use no more than 5% of your portfolio. Why? That's how much you can afford to lose and not ruin your retirement. This is your life's savings here, tread carefully.

To help new investors, I encourage members to post how much of their total portfolio is used for this approach. (There's nothing wrong with a simple 3-fund portfolio...)

The OP is riding this adventure with ~10% of invested assets (initial contribution).

How do you know how much somebody can afford to lose?

Forester · Post by **Forester** » Wed Sep 30, 2020 11:27 am

Probably a bad idea if the bond bull market topped out already earlier this year. Unlevered S&P 500 & LT bonds would make me nervous, let alone trebling down. In investing betting the house on what's done well recently never works out well.

rascott · Post by **rascott** » Wed Sep 30, 2020 11:31 am

Forester wrote: ↑Wed Sep 30, 2020 11:27 am Probably a bad idea if the bond bull market topped out already earlier this year. Unlevered S&P 500 & LT bonds would make me nervous, let alone trebling down. In investing betting the house on what's done well recently never works out well.

Well that was the same argument made over a year ago, and I'm up about 76% since that time when I started.

physixfan · Post by **physixfan** » Wed Sep 30, 2020 12:53 pm

keith6014 wrote: ↑Wed Sep 30, 2020 7:31 am
physixfan wrote: ↑Tue Sep 29, 2020 8:18 pm Start date: 9/1/19
Approach: Adaptive Allocation, which means the percentages of UPRO/TMF are adjusted according to their last 20 day's volatility.
Rebalancing frequency: Monthly
Performance: I'm using xueqiu.com (a Chinese website) to track my progress: https://xueqiu.com/P/ZH2021590. You can just ignore all the Chinese characters and only look at the chart. Currently it's near a 1 year mark, and it's up about 33%.
Great website. Is there an English equivalence? I am shocked there aren't that many good portfolio performance tracking tools.

I don't know any English equivalence... I think Personal Capital is a great portfolio performance tool for oneself, but it does not have a sharing feature which eliminates the absolute value and only shows the relative value to the public...

dsg123 · Post by **dsg123** » Wed Sep 30, 2020 1:22 pm

Starting Today 9/30/20--excited to try this strategy

Start date: 9/30/2020
Approach: Fixed allocation - 55/45 UPRO/TMF
Rebalancing frequency: Quarterly
Return (total / YTD): -n/a
Initial contribution: $10K
Additional contributions: None as of now, but will build 10K 1st of the year (starting Jan 2021)
Portfolio location: M1

So, this is to track along with y'all--I think the main thing is to stick to the plan no matter what. This is less than 10% of my net investments.
This is more a test in mental toughness (selling out) than the actual strategy, in my opinion.
But I may be wrong.

tchoupitoulas · Post by **tchoupitoulas** » Wed Sep 30, 2020 7:53 pm

Start date: 4/3/2019

Approach until 10/2019:
Fixed allocation - 40/60 UPRO/TMF
Rebalancing frequency: Quarterly

Approach 10/2019 to present:
Adaptive allocation (risk parity)
Rebalancing frequency: Monthly

Return (total / YTD): (77%/25%)
Initial contribution: $36K
Additional contributions: None
Portfolio location: M1

SVT · Post by **SVT** » Wed Sep 30, 2020 9:54 pm

Start date: 2/25/2019
Approach: Fixed allocation - 50/50 UPRO/TMF
Rebalancing frequency: Quarterly
Return (total / YTD): 127% / ?
Initial contribution: $81,820 (~10% of net worth at start)
Additional contributions: $0
Portfolio location: Roth IRA, M1Finance

BuffMaltese · Post by **BuffMaltese** » Thu Oct 01, 2020 1:10 am

Any thoughts from those of you doing AAA regarding the current high allocation of tmf. I’m doing minimum variance and the current recommended percentage of tmf is the highest it’s been in 10 years.

BullHouse_BearMarket · Thu Oct 01, 2020 7:43 am

Today is rebalance day for me.

Start date: 11/7/2019

Approach: Fixed allocation - 55/45 UPRO/TMF

Rebalancing frequency: Quarterly

Return (total / YTD): 61% / ?

Initial contribution: $10,000

Additional contributions: $4,000

Portfolio location: Roth IRA, M1Finance

Dr. Long · Post by **Dr. Long** » Thu Oct 01, 2020 7:46 am

Today is rebalance day for me too.

Start date: 7/3/2019

Approach: Fixed allocation - 60/40 UPRO/TMF

Rebalancing frequency: Quarterly

Return (total / YTD): 86.8% / ?

Initial contribution: $5,000

Additional contributions: $0

Portfolio location: Roth IRA, M1Finance

HawkeyePierce · Post by **HawkeyePierce** » Thu Oct 01, 2020 9:51 am

Rebal day for me too, though my allocation was off by <5%.

stimulacra · Post by **stimulacra** » Thu Oct 01, 2020 10:34 am

Start date: 9/25/20
Approach: Fixed allocation - 55/45 UPRO/TMF
Rebalancing frequency: Quarterly
Return (total / YTD): N/A
Initial contribution: $1k (<1% of portfolio)
Additional contributions: $5,500 for 2020 (when possible)
Portfolio location: Roth IRA, M1Finance

Total setup time for me was about 20 minutes spread over 2-3 days (for bank confirmation).

Alaric · Post by **Alaric** » Thu Oct 01, 2020 2:56 pm

Start date: 02/27/19
Approach: Fixed allocation: 02/27/19 — 10/01/19: 40/60 UPRO/TMF
10/01/19 — 02/11/20: 55/45 UPRO/TMF
02/11/20 — present: 30/25/45 UPRO/TQQQ/TMF
Rebalancing frequency: Quarterly (actual rebals 10/1/19; 2/11/20; 4/6/20; 7/1/20; 10/1/20)
Return (Total / CAGR): 74.0% / 80.3%
Initial contribution: $10k (~5% of invested assets)
Additional contributions: $28.5k added 2/11/20
HFEA weight of total portfolio: 17%
Portfolio location: Roth IRA, E*Trade

CRJPylote · Post by **CRJPylote** » Sat Oct 03, 2020 3:41 pm

Start date: 9/03/2020
Approach: UPRO/TMF 50/50(not adapted until end of sept)
Rebalancing frequency: Quarterly
Return (total / YTD) 3.83%/3.83%
Initial contribution: $200, although refined strategy(was in TQQQ for a bit in sept) and am up to $600 with $400 more on the way (5-6% of invested assets)
Additional contributions: TBD, will not exceed $300 per year (5% of roth space)
Portfolio location: Secondary Roth, Schwab
Rationale for strategy: Relatively low income at 24 years old but have contracted pay raises and expect higher income in 5-10 years. Would like to magnify long term returns at the cost of volatility.

corp_sharecropper · Post by **corp_sharecropper** » Sat Oct 03, 2020 5:12 pm

This thread makes me smile.

Looking forward to the day when we have a multi-asset, high leverage, risk parity thread that's multiple pages long.

I decided to put a little in the hedgefundie game so that I can track it, maintain interest, and compare it to my main strategy. Lucky for me, I decided this near the recent COVID lows (very early April). This fact may skew all "since inception" return comparisons I make if I don't contribute on a regular basis (I have not been contributing to it other than two times since the initial buy-in). Don't get me wrong, it's a wonderful problem to have, I'll just need to be mindful of this stroke of luck when I'm running the numbers years from now.

One thing I'm curious to hear what others think about is cheaper, more efficient/customizable alternatives. I'll explain where I'm coming from here. I'm assuming that anyone on bogleheads is already in the >95th percentile of both financial literacy, attention to the current financial landscape, and desire to learn more about finance/investing. So regardless of the simplicity and hands-off mantra of bogleheads, there seems to me both the aptitude and willingness/desire to be hands-ON and accepting of at least moderate "complexity". So why not implement this with futures? Either both sides of the equation (stocks & bonds) or one of them at least. My personal preference would be equities held outright, and treasury exposure through futures. The main hurdle with that is the notional value of a single treasury futures contract (going to be > $100K). It would definitely be cheaper than these LETFs. I also get the sense that people like the HF implementation (using LETFs), due to the fact that it uses implicit leverage, which means there's no margin to monitor on a personal level. That said, it would take a combination of the grossest of negligence along with incredibly bad market conditions to blow up an account using treasury futures, at the same amount of leverage as TMF. Honestly, you'd really have to try to blow it up, it's just not going to happen. I also adamantly believe having just long treasuries and stocks is sub optimal, and has potential to be a problem at some point but that's something for different thread. So I'm curious what some of you think on the points/opinions/assumptions I've described.

Keep calm and carry on.

whodidntante · Post by **whodidntante** » Tue Oct 20, 2020 9:38 pm

corp_sharecropper wrote: ↑Sat Oct 03, 2020 5:12 pm One thing I'm curious to hear what others think about is cheaper, more efficient/customizable alternatives. I'll explain where I'm coming from here. I'm assuming that anyone on bogleheads is already in the >95th percentile of both financial literacy, attention to the current financial landscape, and desire to learn more about finance/investing. So regardless of the simplicity and hands-off mantra of bogleheads, there seems to me both the aptitude and willingness/desire to be hands-ON and accepting of at least moderate "complexity". So why not implement this with futures? Either both sides of the equation (stocks & bonds) or one of them at least. My personal preference would be equities held outright, and treasury exposure through futures. The main hurdle with that is the notional value of a single treasury futures contract (going to be > $100K). It would definitely be cheaper than these LETFs. I also get the sense that people like the HF implementation (using LETFs), due to the fact that it uses implicit leverage, which means there's no margin to monitor on a personal level. That said, it would take a combination of the grossest of negligence along with incredibly bad market conditions to blow up an account using treasury futures, at the same amount of leverage as TMF. Honestly, you'd really have to try to blow it up, it's just not going to happen. I also adamantly believe having just long treasuries and stocks is sub optimal, and has potential to be a problem at some point but that's something for different thread. So I'm curious what some of you think on the points/opinions/assumptions I've described.

Most of the time I get no engagement when I suggest futures for various use cases that come up on this site. From the age of your unloved post, the same happened to you. I've just decided that is what happens when I bring up futures. I must be speaking gibberish. Do you want to join my gibberish club?

djeayzonne · Post by **djeayzonne** » Tue Oct 20, 2020 9:47 pm

whodidntante wrote: ↑Tue Oct 20, 2020 9:38 pm
corp_sharecropper wrote: ↑Sat Oct 03, 2020 5:12 pm One thing I'm curious to hear what others think about is cheaper, more efficient/customizable alternatives. I'll explain where I'm coming from here. I'm assuming that anyone on bogleheads is already in the >95th percentile of both financial literacy, attention to the current financial landscape, and desire to learn more about finance/investing. So regardless of the simplicity and hands-off mantra of bogleheads, there seems to me both the aptitude and willingness/desire to be hands-ON and accepting of at least moderate "complexity". So why not implement this with futures? Either both sides of the equation (stocks & bonds) or one of them at least. My personal preference would be equities held outright, and treasury exposure through futures. The main hurdle with that is the notional value of a single treasury futures contract (going to be > $100K). It would definitely be cheaper than these LETFs. I also get the sense that people like the HF implementation (using LETFs), due to the fact that it uses implicit leverage, which means there's no margin to monitor on a personal level. That said, it would take a combination of the grossest of negligence along with incredibly bad market conditions to blow up an account using treasury futures, at the same amount of leverage as TMF. Honestly, you'd really have to try to blow it up, it's just not going to happen. I also adamantly believe having just long treasuries and stocks is sub optimal, and has potential to be a problem at some point but that's something for different thread. So I'm curious what some of you think on the points/opinions/assumptions I've described.
Most of the time I get no engagement when I suggest futures for various use cases that come up on this site. From the age of your unloved post, the same happened to you. I've just decided that is what happens when I bring up futures. I must be speaking gibberish. Do you want to join my gibberish club?

I have about 35k dedicated to HFEA in Roth and HSA accounts.
I am using futures to do something similar in my taxable account.
I would like to engage in such conversation as I am not entirely sure I am doing things right.
Currently using 1 ES contract, 1 ZN contract, and about 6k worth of EDV.

whodidntante · Post by **whodidntante** » Tue Oct 20, 2020 10:11 pm

djeayzonne wrote: ↑Tue Oct 20, 2020 9:47 pm
whodidntante wrote: ↑Tue Oct 20, 2020 9:38 pm
corp_sharecropper wrote: ↑Sat Oct 03, 2020 5:12 pm One thing I'm curious to hear what others think about is cheaper, more efficient/customizable alternatives. I'll explain where I'm coming from here. I'm assuming that anyone on bogleheads is already in the >95th percentile of both financial literacy, attention to the current financial landscape, and desire to learn more about finance/investing. So regardless of the simplicity and hands-off mantra of bogleheads, there seems to me both the aptitude and willingness/desire to be hands-ON and accepting of at least moderate "complexity". So why not implement this with futures? Either both sides of the equation (stocks & bonds) or one of them at least. My personal preference would be equities held outright, and treasury exposure through futures. The main hurdle with that is the notional value of a single treasury futures contract (going to be > $100K). It would definitely be cheaper than these LETFs. I also get the sense that people like the HF implementation (using LETFs), due to the fact that it uses implicit leverage, which means there's no margin to monitor on a personal level. That said, it would take a combination of the grossest of negligence along with incredibly bad market conditions to blow up an account using treasury futures, at the same amount of leverage as TMF. Honestly, you'd really have to try to blow it up, it's just not going to happen. I also adamantly believe having just long treasuries and stocks is sub optimal, and has potential to be a problem at some point but that's something for different thread. So I'm curious what some of you think on the points/opinions/assumptions I've described.
Most of the time I get no engagement when I suggest futures for various use cases that come up on this site. From the age of your unloved post, the same happened to you. I've just decided that is what happens when I bring up futures. I must be speaking gibberish. Do you want to join my gibberish club?
I have about 35k dedicated to HFEA in Roth and HSA accounts.
I am using futures to do something similar in my taxable account.
I would like to engage in such conversation as I am not entirely sure I am doing things right.
Currently using 1 ES contract, 1 ZN contract, and about 6k worth of EDV.

I'm not sure what your concern is exactly but that sounds too stock heavy for risk parity. Your notional exposure is 172k stock, 106k bond. If that's what you wanted, party on.

corp_sharecropper · Post by **corp_sharecropper** » Tue Oct 20, 2020 11:19 pm

whodidntante wrote: ↑Tue Oct 20, 2020 9:38 pm
corp_sharecropper wrote: ↑Sat Oct 03, 2020 5:12 pm One thing I'm curious to hear what others think about is cheaper, more efficient/customizable alternatives. I'll explain where I'm coming from here. I'm assuming that anyone on bogleheads is already in the >95th percentile of both financial literacy, attention to the current financial landscape, and desire to learn more about finance/investing. So regardless of the simplicity and hands-off mantra of bogleheads, there seems to me both the aptitude and willingness/desire to be hands-ON and accepting of at least moderate "complexity". So why not implement this with futures? Either both sides of the equation (stocks & bonds) or one of them at least. My personal preference would be equities held outright, and treasury exposure through futures. The main hurdle with that is the notional value of a single treasury futures contract (going to be > $100K). It would definitely be cheaper than these LETFs. I also get the sense that people like the HF implementation (using LETFs), due to the fact that it uses implicit leverage, which means there's no margin to monitor on a personal level. That said, it would take a combination of the grossest of negligence along with incredibly bad market conditions to blow up an account using treasury futures, at the same amount of leverage as TMF. Honestly, you'd really have to try to blow it up, it's just not going to happen. I also adamantly believe having just long treasuries and stocks is sub optimal, and has potential to be a problem at some point but that's something for different thread. So I'm curious what some of you think on the points/opinions/assumptions I've described.
Most of the time I get no engagement when I suggest futures for various use cases that come up on this site. From the age of your unloved post, the same happened to you. I've just decided that is what happens when I bring up futures. I must be speaking gibberish. Do you want to join my gibberish club?

You had me at "Most of the time...", sign me up!

chrisdds98 · Post by **chrisdds98** » Tue Oct 20, 2020 11:49 pm

corp_sharecropper wrote: ↑Tue Oct 20, 2020 11:19 pm
whodidntante wrote: ↑Tue Oct 20, 2020 9:38 pm
corp_sharecropper wrote: ↑Sat Oct 03, 2020 5:12 pm One thing I'm curious to hear what others think about is cheaper, more efficient/customizable alternatives. I'll explain where I'm coming from here. I'm assuming that anyone on bogleheads is already in the >95th percentile of both financial literacy, attention to the current financial landscape, and desire to learn more about finance/investing. So regardless of the simplicity and hands-off mantra of bogleheads, there seems to me both the aptitude and willingness/desire to be hands-ON and accepting of at least moderate "complexity". So why not implement this with futures? Either both sides of the equation (stocks & bonds) or one of them at least. My personal preference would be equities held outright, and treasury exposure through futures. The main hurdle with that is the notional value of a single treasury futures contract (going to be > $100K). It would definitely be cheaper than these LETFs. I also get the sense that people like the HF implementation (using LETFs), due to the fact that it uses implicit leverage, which means there's no margin to monitor on a personal level. That said, it would take a combination of the grossest of negligence along with incredibly bad market conditions to blow up an account using treasury futures, at the same amount of leverage as TMF. Honestly, you'd really have to try to blow it up, it's just not going to happen. I also adamantly believe having just long treasuries and stocks is sub optimal, and has potential to be a problem at some point but that's something for different thread. So I'm curious what some of you think on the points/opinions/assumptions I've described.
Most of the time I get no engagement when I suggest futures for various use cases that come up on this site. From the age of your unloved post, the same happened to you. I've just decided that is what happens when I bring up futures. I must be speaking gibberish. Do you want to join my gibberish club?
You had me at "Most of the time...", sign me up!

I like the idea. Is there a guide for newbies of the mechanics of placing these trades at m1finance or fidelity. I'm assuming this isn't possible at vanguard

rascott · Post by **rascott** » Wed Oct 21, 2020 12:05 am

corp_sharecropper wrote: ↑Sat Oct 03, 2020 5:12 pm This thread makes me smile.

Looking forward to the day when we have a multi-asset, high leverage, risk parity thread that's multiple pages long.

I decided to put a little in the hedgefundie game so that I can track it, maintain interest, and compare it to my main strategy. Lucky for me, I decided this near the recent COVID lows (very early April). This fact may skew all "since inception" return comparisons I make if I don't contribute on a regular basis (I have not been contributing to it other than two times since the initial buy-in). Don't get me wrong, it's a wonderful problem to have, I'll just need to be mindful of this stroke of luck when I'm running the numbers years from now.

One thing I'm curious to hear what others think about is cheaper, more efficient/customizable alternatives. I'll explain where I'm coming from here. I'm assuming that anyone on bogleheads is already in the >95th percentile of both financial literacy, attention to the current financial landscape, and desire to learn more about finance/investing. So regardless of the simplicity and hands-off mantra of bogleheads, there seems to me both the aptitude and willingness/desire to be hands-ON and accepting of at least moderate "complexity". So why not implement this with futures? Either both sides of the equation (stocks & bonds) or one of them at least. My personal preference would be equities held outright, and treasury exposure through futures. The main hurdle with that is the notional value of a single treasury futures contract (going to be > $100K). It would definitely be cheaper than these LETFs. I also get the sense that people like the HF implementation (using LETFs), due to the fact that it uses implicit leverage, which means there's no margin to monitor on a personal level. That said, it would take a combination of the grossest of negligence along with incredibly bad market conditions to blow up an account using treasury futures, at the same amount of leverage as TMF. Honestly, you'd really have to try to blow it up, it's just not going to happen. I also adamantly believe having just long treasuries and stocks is sub optimal, and has potential to be a problem at some point but that's something for different thread. So I'm curious what some of you think on the points/opinions/assumptions I've described.

Keep calm and carry on.

I've been using futures for about a year+.... discussed it a lot in the HF thread back Aug/ Sept 2019. It's worked fine and is pretty simple to operate once per quarter. I use TD Ameritrade Think or Swim.

amk0408 · Post by **amk0408** » Wed Oct 21, 2020 8:55 am

Start date: 9/30/20
Approach: Fixed allocation - 25/20/55 UPRO/TQQQ/TMF
Return (total / YTD): -1.44% / -1.44%
Initial contribution: $15k (< 2% of invested assets)
Additional contributions: $5,000 quarterly
Rebalancing: Quarterly (1/1, 4/1, 7/1, 10/1) as needed
Location: M1Finance - Taxable

20% of new contribution dollars will be going to HEFA and the remaining 80% will be going to a traditional 40/60 bond/stock portfolio.

djeayzonne · Post by **djeayzonne** » Wed Oct 21, 2020 2:35 pm

whodidntante wrote: ↑Tue Oct 20, 2020 10:11 pm
djeayzonne wrote: ↑Tue Oct 20, 2020 9:47 pm
whodidntante wrote: ↑Tue Oct 20, 2020 9:38 pm
corp_sharecropper wrote: ↑Sat Oct 03, 2020 5:12 pm One thing I'm curious to hear what others think about is cheaper, more efficient/customizable alternatives. I'll explain where I'm coming from here. I'm assuming that anyone on bogleheads is already in the >95th percentile of both financial literacy, attention to the current financial landscape, and desire to learn more about finance/investing. So regardless of the simplicity and hands-off mantra of bogleheads, there seems to me both the aptitude and willingness/desire to be hands-ON and accepting of at least moderate "complexity". So why not implement this with futures? Either both sides of the equation (stocks & bonds) or one of them at least. My personal preference would be equities held outright, and treasury exposure through futures. The main hurdle with that is the notional value of a single treasury futures contract (going to be > $100K). It would definitely be cheaper than these LETFs. I also get the sense that people like the HF implementation (using LETFs), due to the fact that it uses implicit leverage, which means there's no margin to monitor on a personal level. That said, it would take a combination of the grossest of negligence along with incredibly bad market conditions to blow up an account using treasury futures, at the same amount of leverage as TMF. Honestly, you'd really have to try to blow it up, it's just not going to happen. I also adamantly believe having just long treasuries and stocks is sub optimal, and has potential to be a problem at some point but that's something for different thread. So I'm curious what some of you think on the points/opinions/assumptions I've described.
Most of the time I get no engagement when I suggest futures for various use cases that come up on this site. From the age of your unloved post, the same happened to you. I've just decided that is what happens when I bring up futures. I must be speaking gibberish. Do you want to join my gibberish club?
I have about 35k dedicated to HFEA in Roth and HSA accounts.
I am using futures to do something similar in my taxable account.
I would like to engage in such conversation as I am not entirely sure I am doing things right.
Currently using 1 ES contract, 1 ZN contract, and about 6k worth of EDV.
I'm not sure what your concern is exactly but that sounds too stock heavy for risk parity. Your notional exposure is 172k stock, 106k bond. If that's what you wanted, party on.

No, the ZN contract is at about 138k at the moment.
But, I read some other threads that were making the argument that you would need like 4 or 5 ZN contracts for each ES contract to match volatility.

sparksfly · Post by **sparksfly** » Wed Oct 21, 2020 3:00 pm

whodidntante wrote: ↑Tue Oct 20, 2020 9:38 pm
corp_sharecropper wrote: ↑Sat Oct 03, 2020 5:12 pm One thing I'm curious to hear what others think about is cheaper, more efficient/customizable alternatives. I'll explain where I'm coming from here. I'm assuming that anyone on bogleheads is already in the >95th percentile of both financial literacy, attention to the current financial landscape, and desire to learn more about finance/investing. So regardless of the simplicity and hands-off mantra of bogleheads, there seems to me both the aptitude and willingness/desire to be hands-ON and accepting of at least moderate "complexity". So why not implement this with futures? Either both sides of the equation (stocks & bonds) or one of them at least. My personal preference would be equities held outright, and treasury exposure through futures. The main hurdle with that is the notional value of a single treasury futures contract (going to be > $100K). It would definitely be cheaper than these LETFs. I also get the sense that people like the HF implementation (using LETFs), due to the fact that it uses implicit leverage, which means there's no margin to monitor on a personal level. That said, it would take a combination of the grossest of negligence along with incredibly bad market conditions to blow up an account using treasury futures, at the same amount of leverage as TMF. Honestly, you'd really have to try to blow it up, it's just not going to happen. I also adamantly believe having just long treasuries and stocks is sub optimal, and has potential to be a problem at some point but that's something for different thread. So I'm curious what some of you think on the points/opinions/assumptions I've described.
Most of the time I get no engagement when I suggest futures for various use cases that come up on this site. From the age of your unloved post, the same happened to you. I've just decided that is what happens when I bring up futures. I must be speaking gibberish. Do you want to join my gibberish club?

What would such a setup look like in terms of ratio of S&P 500 to treasury futures?

whodidntante · Post by **whodidntante** » Wed Oct 21, 2020 5:11 pm

sparksfly wrote: ↑Wed Oct 21, 2020 3:00 pm
whodidntante wrote: ↑Tue Oct 20, 2020 9:38 pm
corp_sharecropper wrote: ↑Sat Oct 03, 2020 5:12 pm One thing I'm curious to hear what others think about is cheaper, more efficient/customizable alternatives. I'll explain where I'm coming from here. I'm assuming that anyone on bogleheads is already in the >95th percentile of both financial literacy, attention to the current financial landscape, and desire to learn more about finance/investing. So regardless of the simplicity and hands-off mantra of bogleheads, there seems to me both the aptitude and willingness/desire to be hands-ON and accepting of at least moderate "complexity". So why not implement this with futures? Either both sides of the equation (stocks & bonds) or one of them at least. My personal preference would be equities held outright, and treasury exposure through futures. The main hurdle with that is the notional value of a single treasury futures contract (going to be > $100K). It would definitely be cheaper than these LETFs. I also get the sense that people like the HF implementation (using LETFs), due to the fact that it uses implicit leverage, which means there's no margin to monitor on a personal level. That said, it would take a combination of the grossest of negligence along with incredibly bad market conditions to blow up an account using treasury futures, at the same amount of leverage as TMF. Honestly, you'd really have to try to blow it up, it's just not going to happen. I also adamantly believe having just long treasuries and stocks is sub optimal, and has potential to be a problem at some point but that's something for different thread. So I'm curious what some of you think on the points/opinions/assumptions I've described.
Most of the time I get no engagement when I suggest futures for various use cases that come up on this site. From the age of your unloved post, the same happened to you. I've just decided that is what happens when I bring up futures. I must be speaking gibberish. Do you want to join my gibberish club?

What would such a setup look like in terms of ratio of S&P 500 to treasury futures?

It's the same as what you would do with 3x leveraged ETFs but you would use futures contracts to achieve the desired ratio of exposure. The futures contracts are a lot bigger. Probably pick the bond side first and then fill in with equity contracts, since those are more scalable with the new micro contracts. You should probably prefer the bond contracts over the note contracts to take more risk on the bond side if your goal is risk parity. But if you just want to build a diversified portfolio at a risk level you are comfortable with, you can substitute notes.

Raryn · Post by **Raryn** » Thu Oct 22, 2020 1:03 pm

All I know is that I should stop looking at the markets.

I joined HFEA with one smaller account in my portfolio on 10/1 with plans to hold 60/40 UPRO/TMF and rebalance quarterly, but I've been unable to not keep a close on it. During the 3 weeks since then, TMF has dropped 9% and UPRO has only come up 5.5% - for a total of -0.2%, despite the S&P being up ~3% - and honestly, TMF has been more volatile than UPRO!

Obviously 3 weeks does not a trend make and I should stop paying attention, but it's just... interesting.

stockmaster · Post by **stockmaster** » Thu Oct 22, 2020 1:06 pm

Why would anybody use TMF when TYD has fewer drawdowns and similar market correlation?

langlands · Post by **langlands** » Thu Oct 22, 2020 1:28 pm

stockmaster wrote: ↑Thu Oct 22, 2020 1:06 pm Why would anybody use TMF when TYD has fewer drawdowns and similar market correlation?

Yes, during that backtest at least TYD is clearly superior to TMF as a counterpart to UPRO (who knows what will happen going forward). I don't think that chart quite makes the point cleanly though since TMF does end up having the higher return. I've added a 2x leveraged TYD portfolio to the simulation that I think makes the comparison even more obvious (same market correlation as TMF and TYD, higher CAGR and lower stdev than TMF):

https://www.portfoliovisualizer.com/bac ... ion4_3=100

Basically there should be a market for more highly leveraged (5x or even 10x) ETFs on the shorter end of the yield curve. This theme has been echoed quite a few times and I guess is the impetus for all the treasury futures talk.

Uncorrelated · Post by **Uncorrelated** » Thu Oct 22, 2020 1:49 pm

stockmaster wrote: ↑Thu Oct 22, 2020 1:06 pm Why would anybody use TMF when TYD has fewer drawdowns and similar market correlation?

You should look at the total portfolio, not the individual parts. TYD is roughly comparable with 40% TMF. If you want to compare these ETF's directly, it would be better to compare TYD with 40% TMF.

I'm fairly certain I did some calculations on TMF/TYD somewhere, but I can't find it. If I remember correctly, the conclusion was that TYD is not worth using because there just isn't enough space in your portfolio to utilize it. If you want to keep the same ratio between equity risk and treasury risk as with 55/45 UPRO/TMF, you would be looking at 33/66 UPRO/TYD. For obvious reasons this has significantly lower expected return and risk. But if you want to take less risk, there are better ways to do that (see viewtopic.php?f=10&t=322366 for inspiration).

fwiw the evidence that treasuries and equities are negatively correlated over long term horizons is extremely weak.

make_a_better_world · Post by **make_a_better_world** » Thu Oct 22, 2020 2:14 pm

I went in July-August 2019. I used a rIRA, tIRA and regular brokerage to be able to take some out before I retire if it works in spectacular fashion and to balance without tax liability.

45% UPRO/55% TMF approach but I forgot to rebalance and UPRO is much heavier now.

As of now, total cash in $133,906 (~2% of portfolio). Total gain $63,348 or 47%. An incredible return but the account was not looking pretty in March when the market crashed. I stayed the course and I'm too chicken to put a large portion of my portfolio into this.

Semantics · Post by **Semantics** » Thu Oct 22, 2020 3:14 pm

Uncorrelated wrote: ↑Thu Oct 22, 2020 1:49 pm
stockmaster wrote: ↑Thu Oct 22, 2020 1:06 pm Why would anybody use TMF when TYD has fewer drawdowns and similar market correlation?

You should look at the total portfolio, not the individual parts. TYD is roughly comparable with 40% TMF. If you want to compare these ETF's directly, it would be better to compare TYD with 40% TMF.

I'm fairly certain I did some calculations on TMF/TYD somewhere, but I can't find it. If I remember correctly, the conclusion was that TYD is not worth using because there just isn't enough space in your portfolio to utilize it. If you want to keep the same ratio between equity risk and treasury risk as with 55/45 UPRO/TMF, you would be looking at 33/66 UPRO/TYD. For obvious reasons this has significantly lower expected return and risk. But if you want to take less risk, there are better ways to do that (see viewtopic.php?f=10&t=322366 for inspiration).

When I plugged TYD into your mean variance optimizer (using ter=1.1, itt=3), and set the returns to 3% for ITT and 4% for LTT to try and model expected returns ~10 years ago the solution actually included a decent amount of TYD for a gamma = 2 investor. It gave approx 30% UPRO / 20% TMF / 50% TYD. For gamma = 1 it was pretty close to 50% UPRO / 50% TMF as expected. So would it be fair to say that TYD may have made sense in the past, and may make sense in the future for some levels of risk aversion if treasury yields rise? It seems like there's a point where the LTT-ITT spread is too small to justify the extra risk for investors who are slightly more risk averse.

Uncorrelated · Post by **Uncorrelated** » Thu Oct 22, 2020 3:51 pm

Semantics wrote: ↑Thu Oct 22, 2020 3:14 pm
Uncorrelated wrote: ↑Thu Oct 22, 2020 1:49 pm
stockmaster wrote: ↑Thu Oct 22, 2020 1:06 pm Why would anybody use TMF when TYD has fewer drawdowns and similar market correlation?

You should look at the total portfolio, not the individual parts. TYD is roughly comparable with 40% TMF. If you want to compare these ETF's directly, it would be better to compare TYD with 40% TMF.

I'm fairly certain I did some calculations on TMF/TYD somewhere, but I can't find it. If I remember correctly, the conclusion was that TYD is not worth using because there just isn't enough space in your portfolio to utilize it. If you want to keep the same ratio between equity risk and treasury risk as with 55/45 UPRO/TMF, you would be looking at 33/66 UPRO/TYD. For obvious reasons this has significantly lower expected return and risk. But if you want to take less risk, there are better ways to do that (see viewtopic.php?f=10&t=322366 for inspiration).
When I plugged TYD into your mean variance optimizer (using ter=1.1, itt=3), and set the returns to 3% for ITT and 4% for LTT to try and model expected returns ~10 years ago the solution actually included a decent amount of TYD for a gamma = 2 investor. It gave approx 30% UPRO / 20% TMF / 50% TYD. For gamma = 1 it was pretty close to 50% UPRO / 50% TMF as expected. So would it be fair to say that TYD may have made sense in the past, and may make sense in the future for some levels of risk aversion if treasury yields rise? It seems like there's a point where the LTT-ITT spread is too small to justify the extra risk for investors who are slightly more risk averse.

I suppose that simulation didn't include total stock market? I always find decent allocations to total stock market at the high gamma's.

here without TYD. The gamma = 1 case might look like it changed significantly but the difference in expected utility is very small.

This simulation with ITT expected return = 1.66 annually and LTT expected return = 2.01 annually (1934-2018 average). I expected an allocation of 0% to TYD, maybe I misremembered. If I reduce the expected return on treasuries by 0.5% (ITT = 1.11%, LTT = 1.51%), the allocation to TYD becomes close to zero and we get the familiar mixture of total stock market, UPRO and TMF.

kim.gold · Post by **kim.gold** » Thu Oct 22, 2020 6:14 pm

djeayzonne wrote: ↑Wed Oct 21, 2020 2:35 pm
whodidntante wrote: ↑Tue Oct 20, 2020 10:11 pm
djeayzonne wrote: ↑Tue Oct 20, 2020 9:47 pm
whodidntante wrote: ↑Tue Oct 20, 2020 9:38 pm
corp_sharecropper wrote: ↑Sat Oct 03, 2020 5:12 pm One thing I'm curious to hear what others think about is cheaper, more efficient/customizable alternatives. I'll explain where I'm coming from here. I'm assuming that anyone on bogleheads is already in the >95th percentile of both financial literacy, attention to the current financial landscape, and desire to learn more about finance/investing. So regardless of the simplicity and hands-off mantra of bogleheads, there seems to me both the aptitude and willingness/desire to be hands-ON and accepting of at least moderate "complexity". So why not implement this with futures? Either both sides of the equation (stocks & bonds) or one of them at least. My personal preference would be equities held outright, and treasury exposure through futures. The main hurdle with that is the notional value of a single treasury futures contract (going to be > $100K). It would definitely be cheaper than these LETFs. I also get the sense that people like the HF implementation (using LETFs), due to the fact that it uses implicit leverage, which means there's no margin to monitor on a personal level. That said, it would take a combination of the grossest of negligence along with incredibly bad market conditions to blow up an account using treasury futures, at the same amount of leverage as TMF. Honestly, you'd really have to try to blow it up, it's just not going to happen. I also adamantly believe having just long treasuries and stocks is sub optimal, and has potential to be a problem at some point but that's something for different thread. So I'm curious what some of you think on the points/opinions/assumptions I've described.
Most of the time I get no engagement when I suggest futures for various use cases that come up on this site. From the age of your unloved post, the same happened to you. I've just decided that is what happens when I bring up futures. I must be speaking gibberish. Do you want to join my gibberish club?
I have about 35k dedicated to HFEA in Roth and HSA accounts.
I am using futures to do something similar in my taxable account.
I would like to engage in such conversation as I am not entirely sure I am doing things right.
Currently using 1 ES contract, 1 ZN contract, and about 6k worth of EDV.
I'm not sure what your concern is exactly but that sounds too stock heavy for risk parity. Your notional exposure is 172k stock, 106k bond. If that's what you wanted, party on.
No, the ZN contract is at about 138k at the moment.
But, I read some other threads that were making the argument that you would need like 4 or 5 ZN contracts for each ES contract to match volatility.

One ZN contract is equivalent to $46,838 of TLT.
$6000 of EDV equivalent to $7,500 of TLT.
So, "1 ES contract, 1 ZN contract, and about 6k worth of EDV" is 172,750 of SPY + $54,338 of TLT which is $227k invested in 76% stock / 24% bond portfolio.

When you calculate the Efficient Frontier for a SPY + TLT portfolio, you maximize the Sharpe Ratio for aprox. 52% SPY / 48% TLT.

To conclude this, an optimum allocation would require more bonds in the portofoliu. At today prices, one /ES contract would require 3.4 ZN contracts to maximize the Sharpe Ratio.

stockmaster · Post by **stockmaster** » Thu Oct 22, 2020 10:10 pm

Uncorrelated wrote: ↑Thu Oct 22, 2020 1:49 pm
stockmaster wrote: ↑Thu Oct 22, 2020 1:06 pm Why would anybody use TMF when TYD has fewer drawdowns and similar market correlation?

You should look at the total portfolio, not the individual parts. TYD is roughly comparable with 40% TMF. If you want to compare these ETF's directly, it would be better to compare TYD with 40% TMF.

I'm fairly certain I did some calculations on TMF/TYD somewhere, but I can't find it. If I remember correctly, the conclusion was that TYD is not worth using because there just isn't enough space in your portfolio to utilize it. If you want to keep the same ratio between equity risk and treasury risk as with 55/45 UPRO/TMF, you would be looking at 33/66 UPRO/TYD. For obvious reasons this has significantly lower expected return and risk. But if you want to take less risk, there are better ways to do that (see viewtopic.php?f=10&t=322366 for inspiration).

fwiw the evidence that treasuries and equities are negatively correlated over long term horizons is extremely weak.

I reran it with those parameters and still got similar results:

Note that the amount of cash added did not significantly alter the sortino or sharpe ratios.

stockmaster · Post by **stockmaster** » Fri Oct 23, 2020 2:23 am

Uncorrelated wrote: ↑Thu Oct 22, 2020 1:49 pm You should look at the total portfolio, not the individual parts. TYD is roughly comparable with 40% TMF. If you want to compare these ETF's directly, it would be better to compare TYD with 40% TMF.

I'm fairly certain I did some calculations on TMF/TYD somewhere, but I can't find it. If I remember correctly, the conclusion was that TYD is not worth using because there just isn't enough space in your portfolio to utilize it. If you want to keep the same ratio between equity risk and treasury risk as with 55/45 UPRO/TMF, you would be looking at 33/66 UPRO/TYD. For obvious reasons this has significantly lower expected return and risk. But if you want to take less risk, there are better ways to do that (see viewtopic.php?f=10&t=322366 for inspiration).

fwiw the evidence that treasuries and equities are negatively correlated over long term horizons is extremely weak.

Hey, I looked into it a bit more and I found that the optimum Sharpe portfolio for UPRO/TMF is 53/47, and for UPRO/TYD it's 33/66, so I got the same numbers you gave when chasing Sharpe ratio. Perhaps that's what you meant.

But I think it's a bit of an oversimplification to call TYD 40% TMF. It's more like 70% TMF, and even then it still behaves very differently. See below.