HEDGEFUNDIE's excellent adventure [risk parity strategy using 3x leveraged ETFs]

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
samsdad
Posts: 516
Joined: Sat Jan 02, 2016 6:20 pm

Re: HEDGEFUNDIE's excellent adventure [risk parity strategy using 3x leveraged ETFs]

Post by samsdad » Mon Feb 11, 2019 11:51 pm

HEDGEFUNDIE wrote:
Mon Feb 11, 2019 11:47 pm
skeptical wrote:
Mon Feb 11, 2019 11:33 pm
I am very interested in hearing from those who are implementing a UPRO/TMF strategy, specifically related to the bid/ask spread and how it impacts returns.

For example, according to M*, the bid/ask for TMF right now is 18.10/20.00, with a current "price" of 19.54, and the ETF has about 100M in assets.

I am not familiar with the details of ETF trading (I have only been using mutual funds). When the quarterly rebalancing takes place, and you need to sell of some of the TMF, does this mean you need to place a market order to sell and only get $18.10 ? Which is a 10% haircut on the sale ? Or, do you place a limit order (not sure if I am using the correct terminology) for $19.54, and hope to get that price, and if not keep trying ?

Right now, UPRO has a bid/ask spread of .1%, so it seems a lot more liquid (with $1.2B in assets).

Not sure how big an effects this is on the "backtesting" and simulations, but doing this 4X a year could mean a lot of friction, especially during periods of market stress and lack of liquidity.
I have no idea how M* is calculating that ridiculous bid ask spread. Perhaps it’s the after-market spread?

Fidelity says TMF’s bid ask spread over the past month averaged 0.08%. ETF.com says it’s 0.06% over the past 60 days.

https://screener.fidelity.com/ftgw/etf/ ... ymbols=TMF

https://www.etf.com/TMF#overview
It’s indeed the aftermarket spread. Check back tomorrow during trading hours.

badapu
Posts: 2
Joined: Mon Feb 11, 2019 11:24 pm

Re: HEDGEFUNDIE's excellent adventure [risk parity strategy using 3x leveraged ETFs]

Post by badapu » Tue Feb 12, 2019 12:06 am

Long time stalker, first time poster.
Really love this adventure. Thanks a lot Hedgefundie!

I do not have room in tax advantaged space. Any suggestions about decreasing taxes in brokerage account? (highest tax bracket)

Also from my modeling -> adding small slices of 3x international markets, 3x REITs or 3x emerging markets -> does not improve portfolio performance? Anyone else with different findings?

Seems UPRO and TMF provides the most efficient portfolio among 3x funds.

skeptical
Posts: 54
Joined: Fri Jul 18, 2014 12:24 pm

Re: HEDGEFUNDIE's excellent adventure [risk parity strategy using 3x leveraged ETFs]

Post by skeptical » Tue Feb 12, 2019 8:45 am

For example, according to M*, the bid/ask for TMF right now is 18.10/20.00, with a current "price" of 19.54, and the ETF has about 100M in assets.
I have no idea how M* is calculating that ridiculous bid ask spread. Perhaps it’s the after-market spread?
It’s indeed the aftermarket spread. Check back tomorrow during trading hours.
Yes, it went down to a a negligible overnight.

staythecourse
Posts: 6520
Joined: Mon Jan 03, 2011 9:40 am

Re: HEDGEFUNDIE's excellent adventure [risk parity strategy using 3x leveraged ETFs]

Post by staythecourse » Tue Feb 12, 2019 9:18 am

SVT wrote:
Mon Feb 11, 2019 10:31 pm
staythecourse wrote:
Mon Feb 11, 2019 10:17 pm
SVT wrote:
Mon Feb 11, 2019 7:57 pm
Well, I just emailed M1 Finance to initiate rolling over my Roth IRAs from Merrill Edge and Vanguard. Portfolio at M1 is set. Going to invest my entire $80k Roth IRA in this. I'll take the risk.
What is the process like at M1? I registered and find the website not so obvious. Maybe because there is no money sitting in it already. I opened a roth IRA there online. Do I just email them a statement from the current firm holding my roth ira and that is it? Is there any paperwork that needs to be done (i'm assuming there should be, but didnt see a link on line for it). Do they transfer the current funds in kind or do they liquidate to cash?

For those who have used their platform is the buying easy? Do you just go to your pie and start buying?

Just trying to figure out the mechanics.

Thanks in advance.

Good luck.

p.s. Again it should be noted for less experienced investor the limits should same as play money, i.e. <10% of your account.
Since you already created the account, from there you can just click on the funding link near the top, then scroll all the way down to the bottom of the page and click on "Start a Transfer". I chose "Transfer account from another brokerage", which gives you instructions on what to do, which includes emailing them a copy of your statements. Apparently, they take care of everything from there. There's also a FAQ link there as well, which is here: https://www.m1finance.com/Direct_Accoun ... rs_FAQ.pdf

It's my understanding that the money I currently have invested in my current Roths will automatically go into the funds I have in the pie I created.

I just emailed them right after the end of normal business hours today, so I haven't received a response yet. I imagine I will tomorrow.
That was my impression while navigating the website. So it would seem from the FAQ that they contact you after the transfer before they drop it in your account to see how you want it allocated. Guess that is when you tell them do it like my pie. Also, when setting up my pie they make you pick 3 slices. How do you alter that down to 2 funds only (UPRO and TMF)?

Good luck.
"The stock market [fluctuation], therefore, is noise. A giant distraction from the business of investing.” | -Jack Bogle

staythecourse
Posts: 6520
Joined: Mon Jan 03, 2011 9:40 am

Re: HEDGEFUNDIE's excellent adventure [risk parity strategy using 3x leveraged ETFs]

Post by staythecourse » Tue Feb 12, 2019 9:20 am

I have read some of links of other sites that analyzed the same approach and when it came to rebalancing I only saw test done by set frequency, i.e. annual, semi, quarterly, etc... Has anyone run the rebalancing using a standard boglehead 5% drift method?

Good luck.
"The stock market [fluctuation], therefore, is noise. A giant distraction from the business of investing.” | -Jack Bogle

skeptic42
Posts: 4
Joined: Mon Feb 11, 2019 5:27 pm

Re: HEDGEFUNDIE's excellent adventure [risk parity strategy using 3x leveraged ETFs]

Post by skeptic42 » Tue Feb 12, 2019 12:39 pm

HEDGEFUNDIE wrote:
Mon Feb 11, 2019 12:36 pm
PluckyDucky wrote:
Mon Feb 11, 2019 11:24 am
Has there been any discussion on the optimal amount of leveraged ETF?

I read a paper a year or 2 ago about mixing 2x and 3x leverage ETF to get the optimal amount such as 2.7x. The optimization was based on various factors that I don't remember. Of course past performance does not mean future results. :happy
It's a good question, and the answer has a lot to do with volatility drag.

Here is the return profile of UPRO at 3x leverage:

Image

And here is the same for SSO at 2x leverage:

Image

The long run volatility of the S&P 500 is 15%; for Long Treasuries it's 10%. At those levels of volatility you are still earning extra return for 3x leverage.

I should note that on a risk-adjusted basis 3x leverage will almost always lose to 2x leverage. But as the old saying goes, "you can't eat Sharpe ratios".
The "Estimated Fund Returns" shown in these tables can be calculated for I the "One Year Index" and V the "One Year Volatility Rate" with:

(1 + I)^3 * exp(-3 * V^2) - 1 for daily 3x leveraged and
(1 + I)^2 * exp(-V^2) - 1 for daily 2x leveraged.

This is just the compounding of the daily leverage and the volatility decay. The cost of borrowing, the expense ratio, and trading costs will reduce returns. The realized returns of leveraged funds are clearly lower than the estimated returns from these tables, from portfolio visualizer:

Dec 31, 2010 - Jan 31, 2019
GDX CAGR -11.01% Stdev 35.97%, estimated daily 3x leveraged CAGR -52.2%, NUGT CAGR -55.53%

Jun 30, 2009 - Jan 31, 2019
VFINX CAGR 14.12% Stdev 12.58%, estimated daily 2x leveraged CAGR 28.2%, SSO CAGR 25.29%
estimated daily 3x leveraged CAGR 41.7%, UPRO CAGR 35.38%

Nov 30, 1997 - Jan 31, 2019
VFINX CAGR 6.93% Stdev 14.91%, estimated daily 2x leveraged CAGR 11.8%, ULPIX CAGR 5.07%

PluckyDucky
Posts: 22
Joined: Tue Jan 15, 2019 8:29 pm

Re: HEDGEFUNDIE's excellent adventure [risk parity strategy using 3x leveraged ETFs]

Post by PluckyDucky » Tue Feb 12, 2019 2:06 pm

WisdomTree has an article on this topic that references a prior article in 1996. Since then, the strategy from the 1996 article out-performed 100% S&P500.
the levered 60/40 portfolio utilizing the same approach he outlined in the paper, actually saw portfolio returns outperform in the following 25-year period compared to its historical back test. The levered 60/40 portfolio returned 12.2% for the 1994–2018 period, 260 bps ahead of the 100% equity line, compared with only 80 bps during his original test.
https://www.wisdomtree.com/blog/2018-08 ... 0-equities

The 1996 article is available online.

ETA: The WisdomTree fund NTSX seeks to achieve a "90/60" equity to stock ratio by being 90/10 then levering the rest of the bond portion to achieve 90/60 according to the website: https://www.wisdomtree.com/-/media/us-m ... (ntsx).pdf

Are there any other funds that do this?

EfficientInvestor
Posts: 92
Joined: Thu Nov 01, 2018 7:02 pm

Re: HEDGEFUNDIE's excellent adventure [risk parity strategy using 3x leveraged ETFs]

Post by EfficientInvestor » Tue Feb 12, 2019 3:09 pm

PluckyDucky wrote:
Tue Feb 12, 2019 2:06 pm
WisdomTree has an article on this topic that references a prior article in 1996. Since then, the strategy from the 1996 article out-performed 100% S&P500.
the levered 60/40 portfolio utilizing the same approach he outlined in the paper, actually saw portfolio returns outperform in the following 25-year period compared to its historical back test. The levered 60/40 portfolio returned 12.2% for the 1994–2018 period, 260 bps ahead of the 100% equity line, compared with only 80 bps during his original test.
https://www.wisdomtree.com/blog/2018-08 ... 0-equities

The 1996 article is available online.

ETA: The WisdomTree fund NTSX seeks to achieve a "90/60" equity to stock ratio by being 90/10 then levering the rest of the bond portion to achieve 90/60 according to the website: https://www.wisdomtree.com/-/media/us-m ... (ntsx).pdf

Are there any other funds that do this?
I'm not aware of any other funds that do this. The article seems to infer that they were the first such ETF. Regardless, I don't think applying leverage to a 60/40 portfolio is the best practice. Something closer to the 40/60 portfolio that has been discussed on this thread is probably closer to optimal because you are applying leverage to a more efficient portfolio. Consider the backtest at the link below that compares a 2X 60/40 fund, a 2X 40/60 fund, and the S&P 500. These use 7-10 year treasuries since the 2X version used has been around since 2006. Both 2X portfolios handily beat the S&P 500. While the 2X 60/40 fund had better return over time than the 2X 40/60, it had twice the max drawdown and had a higher SD than the S&P 500. The 2X 40/60 portfolio still drastically outperformed the S&P 500, but with less SD. If you want to go for increased returns, it is better to go from 2X leverage to 3X leverage on a 40/60 than to go from 2X 40/60 to 2X 60/40. If 3X funds had been around since 2006, a 3X 40/60 fund would have had a CAGR over the period around 18% with a max DD of approximately -30%. Much better returns than the 2X 60/40 portfolio but with less drawdown. Once again, the driver being that you are applying leverage to a more efficient asset allocation.

Jul 2006 - Jan 2019
2X 60/40 - CAGR = 12.8%, SD = 16.1%, Max DD = -43.7%
2X 40/60 - CAGR = 11.8%, SD = 12.2%, Max DD = -21.7%
S&P 500 - CAGR = 8.3%, SD = 14.5%, Max DD = -51.0%

https://www.portfoliovisualizer.com/bac ... tion2_2=60

SVT
Posts: 244
Joined: Mon Oct 13, 2008 8:56 am

Re: HEDGEFUNDIE's excellent adventure [risk parity strategy using 3x leveraged ETFs]

Post by SVT » Tue Feb 12, 2019 3:34 pm

staythecourse wrote:
Tue Feb 12, 2019 9:18 am
SVT wrote:
Mon Feb 11, 2019 10:31 pm
staythecourse wrote:
Mon Feb 11, 2019 10:17 pm
SVT wrote:
Mon Feb 11, 2019 7:57 pm
Well, I just emailed M1 Finance to initiate rolling over my Roth IRAs from Merrill Edge and Vanguard. Portfolio at M1 is set. Going to invest my entire $80k Roth IRA in this. I'll take the risk.
What is the process like at M1? I registered and find the website not so obvious. Maybe because there is no money sitting in it already. I opened a roth IRA there online. Do I just email them a statement from the current firm holding my roth ira and that is it? Is there any paperwork that needs to be done (i'm assuming there should be, but didnt see a link on line for it). Do they transfer the current funds in kind or do they liquidate to cash?

For those who have used their platform is the buying easy? Do you just go to your pie and start buying?

Just trying to figure out the mechanics.

Thanks in advance.

Good luck.

p.s. Again it should be noted for less experienced investor the limits should same as play money, i.e. <10% of your account.
Since you already created the account, from there you can just click on the funding link near the top, then scroll all the way down to the bottom of the page and click on "Start a Transfer". I chose "Transfer account from another brokerage", which gives you instructions on what to do, which includes emailing them a copy of your statements. Apparently, they take care of everything from there. There's also a FAQ link there as well, which is here: https://www.m1finance.com/Direct_Accoun ... rs_FAQ.pdf

It's my understanding that the money I currently have invested in my current Roths will automatically go into the funds I have in the pie I created.

I just emailed them right after the end of normal business hours today, so I haven't received a response yet. I imagine I will tomorrow.
That was my impression while navigating the website. So it would seem from the FAQ that they contact you after the transfer before they drop it in your account to see how you want it allocated. Guess that is when you tell them do it like my pie. Also, when setting up my pie they make you pick 3 slices. How do you alter that down to 2 funds only (UPRO and TMF)?

Good luck.
That's just for signing up. Once you're signed up you can build your pie with just 2 funds.

jaj2276
Posts: 382
Joined: Sat Apr 16, 2011 5:13 pm

Re: HEDGEFUNDIE's excellent adventure [risk parity strategy using 3x leveraged ETFs]

Post by jaj2276 » Tue Feb 12, 2019 8:36 pm

staythecourse wrote:
Tue Feb 12, 2019 9:20 am
I have read some of links of other sites that analyzed the same approach and when it came to rebalancing I only saw test done by set frequency, i.e. annual, semi, quarterly, etc... Has anyone run the rebalancing using a standard boglehead 5% drift method?

Good luck.
I've run it but the bands look materially worse than simply picking a time-frame. My simulation tests 40/60 UPRO/TMF from Feb 2010 to present with a $50k buy-in.

No rebalance gives an ending balance of $325k with a max portfolio balance of $405k.
A quarterly rebalance was best with an ending balance of $473k and a max portfolio balance of $504k.

A 5% band gives values of $414k/$442k.
A 10% band gives values of $396k/$427k.

Maybe a combo band of time + outperformance might yield better numbers. Quarterly rebal and 15% bands gave better numbers but of course that's not really actionable.

staythecourse
Posts: 6520
Joined: Mon Jan 03, 2011 9:40 am

Re: HEDGEFUNDIE's excellent adventure [risk parity strategy using 3x leveraged ETFs]

Post by staythecourse » Tue Feb 12, 2019 11:23 pm

jaj2276 wrote:
Tue Feb 12, 2019 8:36 pm
staythecourse wrote:
Tue Feb 12, 2019 9:20 am
I have read some of links of other sites that analyzed the same approach and when it came to rebalancing I only saw test done by set frequency, i.e. annual, semi, quarterly, etc... Has anyone run the rebalancing using a standard boglehead 5% drift method?

Good luck.
I've run it but the bands look materially worse than simply picking a time-frame. My simulation tests 40/60 UPRO/TMF from Feb 2010 to present with a $50k buy-in.

No rebalance gives an ending balance of $325k with a max portfolio balance of $405k.
A quarterly rebalance was best with an ending balance of $473k and a max portfolio balance of $504k.

A 5% band gives values of $414k/$442k.
A 10% band gives values of $396k/$427k.

Maybe a combo band of time + outperformance might yield better numbers. Quarterly rebal and 15% bands gave better numbers but of course that's not really actionable.
Thanks for the information.

One of the posters above was kind enough to link several sites where this approach has already been thought of and used. Some of them that was linked showed different rebalancing methods using frequency. Sometimes the best was annual and other time quarterly. I am thinking it ALL matters which is best based on sequence of returns during the time period examined. Just curious if bands fell into that same "trap" so to speak.

Maybe Mr. Swedroe's opinion into retirement on rebalancing would be useful. Where you rebalance OUT of stocks, but not into stocks? That way you are not doubling down into a bear market when stocks are getting pummeled for a long period of time. Sort of take gains off the table at 10% band on the upswing of UPRO, but not rebalance back into UPRO if it hits the skids?

Just thinking out loud for anyone who has an opinion.

Good luck.
"The stock market [fluctuation], therefore, is noise. A giant distraction from the business of investing.” | -Jack Bogle

BanditKing
Posts: 584
Joined: Tue Oct 29, 2013 11:11 pm

Re: HEDGEFUNDIE's excellent adventure [risk parity strategy using 3x leveraged ETFs]

Post by BanditKing » Wed Feb 13, 2019 12:00 am

I may have missed it, how how does this adventure fare if executed just prior to a market downturn? That would be my concern - I'd worry that the recovery time would be significant.

Any backtests for starting close to 2000 and 2008 type of plummets?

User avatar
mrspock
Posts: 183
Joined: Tue Feb 13, 2018 2:49 am
Location: Vulcan

Re: HEDGEFUNDIE's excellent adventure [risk parity strategy using 3x leveraged ETFs]

Post by mrspock » Wed Feb 13, 2019 12:25 am

BanditKing wrote:
Wed Feb 13, 2019 12:00 am
I may have missed it, how how does this adventure fare if executed just prior to a market downturn? That would be my concern - I'd worry that the recovery time would be significant.

Any backtests for starting close to 2000 and 2008 type of plummets?
For what it's worth, I tried simulating this at the worst possible time last fall (~Sep 17th), from peak to plunge to Feb 12th... you'd be ahead of simply holding VTI through the same period. I didn't rebalance in this scenario as I was assuming an annual rebalance. Not too shabby... I hear you though, this strategy seems waaaay more interesting right after a solid bear market (>35% plunge from the peak), sure it could plunge more...but risk/reward ratio becomes almost irresistible.

Again, this is a *risky* strategy so I'd bet a very modest % of your portfolio (I'm betting <<5%) if you plan on joining the rest of us fools on this adventure :D , I can't help but think the idea is really good, but 8 or 9 years into a bull market is probably asking for trouble. That said, the backtests showed it faired better than my current 80/20 allocation, so I guess we will find out!

BanditKing
Posts: 584
Joined: Tue Oct 29, 2013 11:11 pm

Re: HEDGEFUNDIE's excellent adventure [risk parity strategy using 3x leveraged ETFs]

Post by BanditKing » Wed Feb 13, 2019 12:31 am

mrspock wrote:
Wed Feb 13, 2019 12:25 am
Again, this is a *risky* strategy so I'd bet a very modest % of your portfolio (I'm betting <<5%) if you plan on joining the rest of us fools on this adventure :D , I can't help but think the idea is really good, but 8 or 9 years into a bull market is probably asking for trouble. That said, the backtests showed it faired better than my current 80/20 allocation, so I guess we will find out!
I'm very interested in doing this, and would certainly be comfortable with 10% of my portfolio here, but the extended bull run we're in now, and general prognostications, makes me feel I should wait a bit. Yea, that's market timing, I know.

I'd still love to see some 2000 and 2008 graphs - could change my mind.

hohum
Posts: 22
Joined: Thu Jan 03, 2019 2:34 pm

Re: HEDGEFUNDIE's excellent adventure [risk parity strategy using 3x leveraged ETFs]

Post by hohum » Wed Feb 13, 2019 12:37 am

BanditKing wrote:
Wed Feb 13, 2019 12:31 am
mrspock wrote:
Wed Feb 13, 2019 12:25 am
Again, this is a *risky* strategy so I'd bet a very modest % of your portfolio (I'm betting <<5%) if you plan on joining the rest of us fools on this adventure :D , I can't help but think the idea is really good, but 8 or 9 years into a bull market is probably asking for trouble. That said, the backtests showed it faired better than my current 80/20 allocation, so I guess we will find out!
I'm very interested in doing this, and would certainly be comfortable with 10% of my portfolio here, but the extended bull run we're in now, and general prognostications, makes me feel I should wait a bit. Yea, that's market timing, I know.

I'd still love to see some 2000 and 2008 graphs - could change my mind.
1998-2011
20% vti 20% ees 10% ijs 20% dls 10% edv 20% iau

Topic Author
HEDGEFUNDIE
Posts: 1754
Joined: Sun Oct 22, 2017 2:06 pm

Re: HEDGEFUNDIE's excellent adventure [risk parity strategy using 3x leveraged ETFs]

Post by HEDGEFUNDIE » Wed Feb 13, 2019 2:43 am

I have updated the original post with a section describing why the strategy should work in theory. Hopefully it is easy to follow. Suggestions welcome.

samsdad
Posts: 516
Joined: Sat Jan 02, 2016 6:20 pm

Re: HEDGEFUNDIE's excellent adventure [risk parity strategy using 3x leveraged ETFs]

Post by samsdad » Wed Feb 13, 2019 7:39 am

hohum wrote:
Wed Feb 13, 2019 12:37 am
BanditKing wrote:
Wed Feb 13, 2019 12:31 am
mrspock wrote:
Wed Feb 13, 2019 12:25 am
Again, this is a *risky* strategy so I'd bet a very modest % of your portfolio (I'm betting <<5%) if you plan on joining the rest of us fools on this adventure :D , I can't help but think the idea is really good, but 8 or 9 years into a bull market is probably asking for trouble. That said, the backtests showed it faired better than my current 80/20 allocation, so I guess we will find out!
I'm very interested in doing this, and would certainly be comfortable with 10% of my portfolio here, but the extended bull run we're in now, and general prognostications, makes me feel I should wait a bit. Yea, that's market timing, I know.

I'd still love to see some 2000 and 2008 graphs - could change my mind.
1998-2011
PV doesn’t allow custom benchmarks like these to be shared unless the recipient also happens to have the same ones in their PV, hence the many pics I posted upthread. So when someone else clicks on your link they see an error.

hohum
Posts: 22
Joined: Thu Jan 03, 2019 2:34 pm

Re: HEDGEFUNDIE's excellent adventure [risk parity strategy using 3x leveraged ETFs]

Post by hohum » Wed Feb 13, 2019 10:06 am

samsdad wrote:
Wed Feb 13, 2019 7:39 am
hohum wrote:
Wed Feb 13, 2019 12:37 am
BanditKing wrote:
Wed Feb 13, 2019 12:31 am
mrspock wrote:
Wed Feb 13, 2019 12:25 am
Again, this is a *risky* strategy so I'd bet a very modest % of your portfolio (I'm betting <<5%) if you plan on joining the rest of us fools on this adventure :D , I can't help but think the idea is really good, but 8 or 9 years into a bull market is probably asking for trouble. That said, the backtests showed it faired better than my current 80/20 allocation, so I guess we will find out!
I'm very interested in doing this, and would certainly be comfortable with 10% of my portfolio here, but the extended bull run we're in now, and general prognostications, makes me feel I should wait a bit. Yea, that's market timing, I know.

I'd still love to see some 2000 and 2008 graphs - could change my mind.
1998-2011
PV doesn’t allow custom benchmarks like these to be shared unless the recipient also happens to have the same ones in their PV, hence the many pics I posted upthread. So when someone else clicks on your link they see an error.
Ha, thanks. I did test the link, and of course it worked for me. What you said makes sense.
20% vti 20% ees 10% ijs 20% dls 10% edv 20% iau

User avatar
HomerJ
Posts: 12291
Joined: Fri Jun 06, 2008 12:50 pm

Re: HEDGEFUNDIE's excellent adventure [risk parity strategy using 3x leveraged ETFs]

Post by HomerJ » Wed Feb 13, 2019 10:28 am

BanditKing wrote:
Wed Feb 13, 2019 12:31 am
I'd still love to see some 2000 and 2008 graphs - could change my mind.
You're making a critical assumption that the next crash will be similar to 2000 and 2008.
The J stands for Jay

BanditKing
Posts: 584
Joined: Tue Oct 29, 2013 11:11 pm

Re: HEDGEFUNDIE's excellent adventure [risk parity strategy using 3x leveraged ETFs]

Post by BanditKing » Wed Feb 13, 2019 12:12 pm

HomerJ wrote:
Wed Feb 13, 2019 10:28 am
BanditKing wrote:
Wed Feb 13, 2019 12:31 am
I'd still love to see some 2000 and 2008 graphs - could change my mind.
You're making a critical assumption that the next crash will be similar to 2000 and 2008.
You are absolutely right, but at the same time you have to have some reference.

dalbright
Posts: 66
Joined: Sun Oct 21, 2018 6:23 am

Re: HEDGEFUNDIE's excellent adventure [risk parity strategy using 3x leveraged ETFs]

Post by dalbright » Wed Feb 13, 2019 12:19 pm

PluckyDucky wrote:
Tue Feb 12, 2019 2:06 pm
WisdomTree has an article on this topic that references a prior article in 1996. Since then, the strategy from the 1996 article out-performed 100% S&P500.
the levered 60/40 portfolio utilizing the same approach he outlined in the paper, actually saw portfolio returns outperform in the following 25-year period compared to its historical back test. The levered 60/40 portfolio returned 12.2% for the 1994–2018 period, 260 bps ahead of the 100% equity line, compared with only 80 bps during his original test.
https://www.wisdomtree.com/blog/2018-08 ... 0-equities

The 1996 article is available online.

ETA: The WisdomTree fund NTSX seeks to achieve a "90/60" equity to stock ratio by being 90/10 then levering the rest of the bond portion to achieve 90/60 according to the website: https://www.wisdomtree.com/-/media/us-m ... (ntsx).pdf

Are there any other funds that do this?
That's actually the first fund i've seen that does something if this nature, thanks! Wish it had more trading volume...

samsdad
Posts: 516
Joined: Sat Jan 02, 2016 6:20 pm

Re: HEDGEFUNDIE's excellent adventure [risk parity strategy using 3x leveraged ETFs]

Post by samsdad » Wed Feb 13, 2019 1:03 pm

BanditKing wrote:
Wed Feb 13, 2019 12:31 am
mrspock wrote:
Wed Feb 13, 2019 12:25 am
Again, this is a *risky* strategy so I'd bet a very modest % of your portfolio (I'm betting <<5%) if you plan on joining the rest of us fools on this adventure :D , I can't help but think the idea is really good, but 8 or 9 years into a bull market is probably asking for trouble. That said, the backtests showed it faired better than my current 80/20 allocation, so I guess we will find out!
I'm very interested in doing this, and would certainly be comfortable with 10% of my portfolio here, but the extended bull run we're in now, and general prognostications, makes me feel I should wait a bit. Yea, that's market timing, I know.

I'd still love to see some 2000 and 2008 graphs - could change my mind.
EDIT: I’ve deleted a large part of this post for the reasons set forth below:
***
If you are following the other thread viewtopic.php?f=10&t=272640&newpost=437 ... ead#unread about simulating the expenses (and therefore, the returns) of these leveraged funds you will see that they have some moving parts in the background that drag on returns, some of which can be estimated based on reference to one-month LIBOR rates, etc., and some of which, at this point, are pure conjecture (swap spreads, short-term-lending rates.)

YOU HAVE TO TAKE THE PREVIOUSLY-POSTED SIMULATED DATA WITH A HUGE GRAIN OF SALT. It does not take into account interest rates that the funds themselves may have been subject to had they been around at the time. The simulated bond fund is sketchy at best as the timeline gets longer too.

To give you an idea, here's the LIBOR rates https://www.macrotrends.net/1433/histor ... ates-chart going back to 1986:

Image

But, I don't want to derail this thread by getting into this too much. Other eyeballs are very much welcome on that other thread: viewtopic.php?f=10&t=272640&newpost=437 ... ead#unread

At this point, I don't think posting anymore pictures or return data based on simulations is appropriate until we have a better set of numbers.

GrowthSeeker
Posts: 338
Joined: Tue May 15, 2018 10:14 pm

Re: HEDGEFUNDIE's excellent adventure [risk parity strategy using 3x leveraged ETFs]

Post by GrowthSeeker » Wed Feb 13, 2019 3:59 pm

hdas wrote:
Sun Feb 10, 2019 10:59 pm
GrowthSeeker wrote:
Sun Feb 10, 2019 10:40 pm
Any thoughts on adding an extra layer of complexity such as a momentum-based timing scheme to switch from UPRO to TMF? Clearly this would be market timing and there is a danger of excessive curve fitting when identifying parameters, but if successful, it might partial protection against the possibility of a wipe out. To me, a worthy goal would be to dramatically decrease the likelihood of a wipeout rather than tweaking to go from 25% CAGR to 30%.

If so, then one could do a weighted average of a fixed ratio and a switching scheme. For example, instead of UPRO/TMF = 40/60, the ratio might vary from 30/70 to 50/50; it could be for example UPRO/TMF/variable = 30/50/20 where "variable" is either UPRO or TMF depending on which one the timing algorithm designates "the winner".

Thoughts?
An option presented upthread :greedy
I'm thinking about a filter that would lessen the odds of a wipe out.

In the upthread example, hdas used PV's Rolling Portfolio Optimization using the Maximize Sharpe Ratio method with a 12 month look-back. OK, that's one way, but to some extent, the PV optimizers are a black box to me. I'm not sure exactly how they are calculating what. To be functional how would I use this? Would I go to PV once a month or quarter and run it? Will this website be there in 20-30 years? Maybe run Sharpe ratio calculations on one's own spreadsheet. Then you could check other parameters, say different look back periods instead of 12 months, and graph out CAGR and Sharpe ratio vs number of look back months (or look back trading days).

What about a different way: PV also has Portfolio Optimization which I assume simply checks different weights over a time period and reports the winner. I think the look back period is the entire data sample. Here is a graph using existing UPRO and TMF data. This got better results than 12-month rolling optimization. But I did not use reconstructed data going further back in time.
https://www.portfoliovisualizer.com/opt ... ymbol2=TMF

Another way: PV's "Dual Momentum" picks the "n" best performers from a group of "m" tickers and goes long in the winners. The idea is to increase the chance that your out of the riskiest leveraged ETF if/when it crashes. I just tried a few ideas and did not find a successful one, but maybe an approach along these lines with different details might. I don't know.
For example I tried using 4 leveraged ETFs (both 3x and 2x versions for stocks and bonds) with quarterly trading and a 3 month look back period to pick the best 1 out of 4. For a benchmark, I used UPRO itself rather than the S&P 500 to see if the UPRO corrections get filtered out.
UPRO, TMF, SSO (2x S&P 500), UBT (2x long bond)
https://www.portfoliovisualizer.com/tes ... odWeight=0

But what about something like a moving average crossover applied to the 40:60 UPRO:TMF portfolio? Here the idea would be to be out of the market in cash (or at least something else which is not leveraged) during times of highest risk. Maybe tune it to be in the market 80% or 90% of the time. Would this be worthy of research, the whole point being to decrease the likelihood you get caught by a crash of one of these leveraged ETFs? I think it would be impossible to KNOW that whatever technique is chosen would lessen the chance of a wipe out. Maybe if the filter were successful in avoiding some of the rapid drops in portfolio value, this would imply that a sudden wipe out would also be filtered out.
Just because you're paranoid doesn't mean they're NOT out to get you.

jaj2276
Posts: 382
Joined: Sat Apr 16, 2011 5:13 pm

Re: HEDGEFUNDIE's excellent adventure [risk parity strategy using 3x leveraged ETFs]

Post by jaj2276 » Wed Feb 13, 2019 4:14 pm

For those using M1 Finance, a couple of questions:

1) Any concern about the company going belly up in the future? Something like MF Global (commingling company and investor funds). Not sure how it works with securities held in book form so might not be relevant.

2) When you rebalance your pie, when does the rebalance happen? Do they put aggressive orders in the market at the prevailing prices? Do they put on close orders?

EfficientInvestor
Posts: 92
Joined: Thu Nov 01, 2018 7:02 pm

Re: HEDGEFUNDIE's excellent adventure [risk parity strategy using 3x leveraged ETFs]

Post by EfficientInvestor » Wed Feb 13, 2019 4:14 pm

GrowthSeeker wrote:
Wed Feb 13, 2019 3:59 pm
hdas wrote:
Sun Feb 10, 2019 10:59 pm
GrowthSeeker wrote:
Sun Feb 10, 2019 10:40 pm
Any thoughts on adding an extra layer of complexity such as a momentum-based timing scheme to switch from UPRO to TMF? Clearly this would be market timing and there is a danger of excessive curve fitting when identifying parameters, but if successful, it might partial protection against the possibility of a wipe out. To me, a worthy goal would be to dramatically decrease the likelihood of a wipeout rather than tweaking to go from 25% CAGR to 30%.

If so, then one could do a weighted average of a fixed ratio and a switching scheme. For example, instead of UPRO/TMF = 40/60, the ratio might vary from 30/70 to 50/50; it could be for example UPRO/TMF/variable = 30/50/20 where "variable" is either UPRO or TMF depending on which one the timing algorithm designates "the winner".

Thoughts?
An option presented upthread :greedy
I'm thinking about a filter that would lessen the odds of a wipe out.

In the upthread example, hdas used PV's Rolling Portfolio Optimization using the Maximize Sharpe Ratio method with a 12 month look-back. OK, that's one way, but to some extent, the PV optimizers are a black box to me. I'm not sure exactly how they are calculating what. To be functional how would I use this? Would I go to PV once a month or quarter and run it? Will this website be there in 20-30 years? Maybe run Sharpe ratio calculations on one's own spreadsheet. Then you could check other parameters, say different look back periods instead of 12 months, and graph out CAGR and Sharpe ratio vs number of look back months (or look back trading days).

What about a different way: PV also has Portfolio Optimization which I assume simply checks different weights over a time period and reports the winner. I think the look back period is the entire data sample. Here is a graph using existing UPRO and TMF data. This got better results than 12-month rolling optimization. But I did not use reconstructed data going further back in time.
https://www.portfoliovisualizer.com/opt ... ymbol2=TMF

Another way: PV's "Dual Momentum" picks the "n" best performers from a group of "m" tickers and goes long in the winners. The idea is to increase the chance that your out of the riskiest leveraged ETF if/when it crashes. I just tried a few ideas and did not find a successful one, but maybe an approach along these lines with different details might. I don't know.
For example I tried using 4 leveraged ETFs (both 3x and 2x versions for stocks and bonds) with quarterly trading and a 3 month look back period to pick the best 1 out of 4. For a benchmark, I used UPRO itself rather than the S&P 500 to see if the UPRO corrections get filtered out.
UPRO, TMF, SSO (2x S&P 500), UBT (2x long bond)
https://www.portfoliovisualizer.com/tes ... odWeight=0

But what about something like a moving average crossover applied to the 40:60 UPRO:TMF portfolio? Here the idea would be to be out of the market in cash (or at least something else which is not leveraged) during times of highest risk. Maybe tune it to be in the market 80% or 90% of the time. Would this be worthy of research, the whole point being to decrease the likelihood you get caught by a crash of one of these leveraged ETFs? I think it would be impossible to KNOW that whatever technique is chosen would lessen the chance of a wipe out. Maybe if the filter were successful in avoiding some of the rapid drops in portfolio value, this would imply that a sudden wipe out would also be filtered out.
If worried about a wipeout, you could consider the use of options. Either selling covered calls or buying puts. The current Jan 2020 (338 days) 20 strike put for UPRO is $0.90/share. UPRO is currently at $45.43. Therefore, you could lock in a max drawdown (for the 3x stock fund) of -56% for 2% of the current price of UPRO. Comparable protection for TMF would be even cheaper.

Topic Author
HEDGEFUNDIE
Posts: 1754
Joined: Sun Oct 22, 2017 2:06 pm

Re: HEDGEFUNDIE's excellent adventure [risk parity strategy using 3x leveraged ETFs]

Post by HEDGEFUNDIE » Wed Feb 13, 2019 5:04 pm

jaj2276 wrote:
Wed Feb 13, 2019 4:14 pm
2) When you rebalance your pie, when does the rebalance happen? Do they put aggressive orders in the market at the prevailing prices? Do they put on close orders?
https://support.m1finance.com/hc/en-us/ ... ing-Window

sparksfly
Posts: 22
Joined: Mon Oct 16, 2017 9:57 pm

Re: HEDGEFUNDIE's excellent adventure [risk parity strategy using 3x leveraged ETFs]

Post by sparksfly » Thu Feb 14, 2019 11:27 am

My M1 Finance Roth IRA was opened and funded yesterday. Auto invest triggered the buy orders this morning in the portfolio. Very small part of the portfolio but I'm in :sharebeer

staythecourse
Posts: 6520
Joined: Mon Jan 03, 2011 9:40 am

Re: HEDGEFUNDIE's excellent adventure [risk parity strategy using 3x leveraged ETFs]

Post by staythecourse » Thu Feb 14, 2019 11:59 am

sparksfly wrote:
Thu Feb 14, 2019 11:27 am
My M1 Finance Roth IRA was opened and funded yesterday. Auto invest triggered the buy orders this morning in the portfolio. Very small part of the portfolio but I'm in :sharebeer
How long from when you emailed them your statement to when the transactions occurred? I'm trying to clean up some accounts right now before making this move and just trying to get an idea of timeline?

Good luck.
"The stock market [fluctuation], therefore, is noise. A giant distraction from the business of investing.” | -Jack Bogle

sparksfly
Posts: 22
Joined: Mon Oct 16, 2017 9:57 pm

Re: HEDGEFUNDIE's excellent adventure [risk parity strategy using 3x leveraged ETFs]

Post by sparksfly » Thu Feb 14, 2019 12:29 pm

staythecourse wrote:
Thu Feb 14, 2019 11:59 am
sparksfly wrote:
Thu Feb 14, 2019 11:27 am
My M1 Finance Roth IRA was opened and funded yesterday. Auto invest triggered the buy orders this morning in the portfolio. Very small part of the portfolio but I'm in :sharebeer
How long from when you emailed them your statement to when the transactions occurred? I'm trying to clean up some accounts right now before making this move and just trying to get an idea of timeline?

Good luck.
I did not transfer an account from another institution so I may not have the right answer if that's what you are asking.

I opened a new Roth IRA on Tuesday. My credit records are frozen and they couldn't verify the identity so they asked me to send a utility bill and DL which I sent same night. They verified and opened the account yesterday. I made a transfer from bank yesterday. This morning, money was in the account and trades took place.

columbia
Posts: 1280
Joined: Tue Aug 27, 2013 5:30 am

Re: HEDGEFUNDIE's excellent adventure [risk parity strategy using 3x leveraged ETFs]

Post by columbia » Thu Feb 14, 2019 12:42 pm

I wonder if this thread is inflating the value of both ETNs. :mrgreen:

jaj2276
Posts: 382
Joined: Sat Apr 16, 2011 5:13 pm

Re: HEDGEFUNDIE's excellent adventure [risk parity strategy using 3x leveraged ETFs]

Post by jaj2276 » Thu Feb 14, 2019 12:59 pm

staythecourse wrote:
Thu Feb 14, 2019 11:59 am
sparksfly wrote:
Thu Feb 14, 2019 11:27 am
My M1 Finance Roth IRA was opened and funded yesterday. Auto invest triggered the buy orders this morning in the portfolio. Very small part of the portfolio but I'm in :sharebeer
How long from when you emailed them your statement to when the transactions occurred? I'm trying to clean up some accounts right now before making this move and just trying to get an idea of timeline?

Good luck.
I transferred a Roth IRA from Vanguard. Started the process on Feb 6, had the money available today (Feb 14). First trade will actually go through tomorrow during their trading window.

Maybe because I'm not a millenial, but M1 Finance's website isn't entirely intuitive to me. I had created a Pie (called My Pie) that was 45/55 but after thinking about it I wanted to change it to 40/60. However I couldn't find a way to edit 'My Pie' and so I created another pie. All somewhat confusing but I think I'm on track.

So as of tomorrow, 2.5% of my portfolio will be invested in this strategy.

User avatar
aj76er
Posts: 642
Joined: Tue Dec 01, 2015 11:34 pm
Location: Portland, OR

Re: HEDGEFUNDIE's excellent adventure [risk parity strategy using 3x leveraged ETFs]

Post by aj76er » Thu Feb 14, 2019 1:17 pm

I decided to give this a try with a very small portion of my portfolio (~ 2%) in my Roth IRA @ Fidelity. I think the logic behind the strategy is compelling, and obviously, recent bull market results have been very good. I think the real risks to this strategy have to do with the counterparty risk inherent in using these ETFs (e.g. UPRO, TMF). Most concerning are the deep risks as follows:
  • Ability of UPRO and TMF to faithfully track a 3X leveraged index in all market conditions (i.e. "manager risk")
  • The long-term prospects of UPRO and TMF staying open (i.e. "fund closure risk" or "business risk")
These funds are designed to be short-term trading tools; and as such, the fund manager's interests are not aligned with mine. If these funds decide their operations are not profitable (or not profitable enough), then they could close. If these funds lose enough AUM in a sharp, downward market, then they could close.

I'm now essentially taking on idiosyncratic risks that index investing is traditionally designed to avoid.

Because of these deep risks, I have added the following proposal into my personal IPS:
Limit the use of leveraged index funds to no more than 5% of total portfolio
My entire Roth is currently <4% of my portfolio; so after some time, I may convert it entirely into this strategy. If it rises above 5% of my portfolio (e.g. 7.5%->10%) of portfolio, then I plan to remove profits into FXAIX (Fidelity's S&P 500 index fund).

FWIW, I opened an account with M1, but never finished the process (Utility Bill + DL), as I decided to stay with Fidelity for the following reasons:
  • Fidelity as a known, documented, and established Roth backdoor process
  • M1 platform is new and comes with it's own risks (buyout, closure, fee hikes, etc...)
  • I didn't want to deal with a new account, new security protocols, etc..
I want to personally thank everyone for contributing on this strategy; especially hedgefundie, efficientinvestor, saimond, and samsdad

I will be following the 3X leveraged data reconstruction thread with much interest :).
"Buy-and-hold, long-term, all-market-index strategies, implemented at rock-bottom cost, are the surest of all routes to the accumulation of wealth" - John C. Bogle

EfficientInvestor
Posts: 92
Joined: Thu Nov 01, 2018 7:02 pm

Re: HEDGEFUNDIE's excellent adventure [risk parity strategy using 3x leveraged ETFs]

Post by EfficientInvestor » Thu Feb 14, 2019 1:37 pm

aj76er wrote:
Thu Feb 14, 2019 1:17 pm
I decided to give this a try with a very small portion of my portfolio (~ 2%) in my Roth IRA @ Fidelity. I think the logic behind the strategy is compelling, and obviously, recent bull market results have been very good. I think the real risks to this strategy have to do with the counterparty risk inherent in using these ETFs (e.g. UPRO, TMF). Most concerning are the deep risks as follows:
  • Ability of UPRO and TMF to faithfully track a 3X leveraged index in all market conditions (i.e. "manager risk")
  • The long-term prospects of UPRO and TMF staying open (i.e. "fund closure risk" or "business risk")
These funds are designed to be short-term trading tools; and as such, the fund manager's interests are not aligned with mine. If these funds decide their operations are not profitable (or not profitable enough), then they could close. If these funds lose enough AUM in a sharp, downward market, then they could close.

I'm now essentially taking on idiosyncratic risks that index investing is traditionally designed to avoid.

Because of these deep risks, I have added the following proposal into my personal IPS:
Limit the use of leveraged index funds to no more than 5% of total portfolio
My entire Roth is currently <4% of my portfolio; so after some time, I may convert it entirely into this strategy. If it rises above 5% of my portfolio (e.g. 7.5%->10%) of portfolio, then I plan to remove profits into FXAIX (Fidelity's S&P 500 index fund).

FWIW, I opened an account with M1, but never finished the process (Utility Bill + DL), as I decided to stay with Fidelity for the following reasons:
  • Fidelity as a known, documented, and established Roth backdoor process
  • M1 platform is new and comes with it's own risks (buyout, closure, fee hikes, etc...)
  • I didn't want to deal with a new account, new security protocols, etc..
I want to personally thank everyone for contributing on this strategy; especially hedgefundie, efficientinvestor, saimond, and samsdad

I will be following the 3X leveraged data reconstruction thread with much interest :).
Rock on. However, I would challenge you about moving profits away from the leveraged portfolio after it gets above 5% of your funds. If we assume the leveraged portfolio gets 20% going forward and the other 95% of your funds get 8%, you would be limiting yourself to essentially an 8.6% return going forward. Alternatively, you could make the decision today that the 2% you are moving over is now a sunk cost. It will either grow like we think it will, or it will go to zero. If you let it continue to grow, it could eventually become a nice chunk of change. If it goes to zero, it doesn't matter because it is already a sunk cost to you.

User avatar
aj76er
Posts: 642
Joined: Tue Dec 01, 2015 11:34 pm
Location: Portland, OR

Re: HEDGEFUNDIE's excellent adventure [risk parity strategy using 3x leveraged ETFs]

Post by aj76er » Thu Feb 14, 2019 1:57 pm

EfficientInvestor wrote:
Thu Feb 14, 2019 1:37 pm
aj76er wrote:
Thu Feb 14, 2019 1:17 pm
I decided to give this a try with a very small portion of my portfolio (~ 2%) in my Roth IRA @ Fidelity. I think the logic behind the strategy is compelling, and obviously, recent bull market results have been very good. I think the real risks to this strategy have to do with the counterparty risk inherent in using these ETFs (e.g. UPRO, TMF). Most concerning are the deep risks as follows:
  • Ability of UPRO and TMF to faithfully track a 3X leveraged index in all market conditions (i.e. "manager risk")
  • The long-term prospects of UPRO and TMF staying open (i.e. "fund closure risk" or "business risk")
These funds are designed to be short-term trading tools; and as such, the fund manager's interests are not aligned with mine. If these funds decide their operations are not profitable (or not profitable enough), then they could close. If these funds lose enough AUM in a sharp, downward market, then they could close.

I'm now essentially taking on idiosyncratic risks that index investing is traditionally designed to avoid.

Because of these deep risks, I have added the following proposal into my personal IPS:
Limit the use of leveraged index funds to no more than 5% of total portfolio
My entire Roth is currently <4% of my portfolio; so after some time, I may convert it entirely into this strategy. If it rises above 5% of my portfolio (e.g. 7.5%->10%) of portfolio, then I plan to remove profits into FXAIX (Fidelity's S&P 500 index fund).

FWIW, I opened an account with M1, but never finished the process (Utility Bill + DL), as I decided to stay with Fidelity for the following reasons:
  • Fidelity as a known, documented, and established Roth backdoor process
  • M1 platform is new and comes with it's own risks (buyout, closure, fee hikes, etc...)
  • I didn't want to deal with a new account, new security protocols, etc..
I want to personally thank everyone for contributing on this strategy; especially hedgefundie, efficientinvestor, saimond, and samsdad

I will be following the 3X leveraged data reconstruction thread with much interest :).
Rock on. However, I would challenge you about moving profits away from the leveraged portfolio after it gets above 5% of your funds. If we assume the leveraged portfolio gets 20% going forward and the other 95% of your funds get 8%, you would be limiting yourself to essentially an 8.6% return going forward. Alternatively, you could make the decision today that the 2% you are moving over is now a sunk cost. It will either grow like we think it will, or it will go to zero. If you let it continue to grow, it could eventually become a nice chunk of change. If it goes to zero, it doesn't matter because it is already a sunk cost to you.
Thanks for this perspective. I've gone back and forth about treating it as a sunk cost vs part of a balanced portfolio. One perspective is to treat this as an investment that is somehow different than the rest of my portfolio, such as a house or antique art, etc. I think that is a valid thing to do. A house or art can increase in value dramatically, or it can go to zero (e.g. fire), so it has similar idiosyncratic risks involved. However, with both a house or art, I can buy insurance to protect against these risks. So under this mental model, perhaps I should buy put options once the portfolio gets large enough? (as opposed to simply de-leveraging). I accept that the portfolio can grow large or go to zero, but the sequence of first growing large and then going to zero would be a real bummer.
"Buy-and-hold, long-term, all-market-index strategies, implemented at rock-bottom cost, are the surest of all routes to the accumulation of wealth" - John C. Bogle

EfficientInvestor
Posts: 92
Joined: Thu Nov 01, 2018 7:02 pm

Re: HEDGEFUNDIE's excellent adventure [risk parity strategy using 3x leveraged ETFs]

Post by EfficientInvestor » Thu Feb 14, 2019 2:05 pm

aj76er wrote:
Thu Feb 14, 2019 1:57 pm
Thanks for this perspective. I've gone back and forth about treating it as a sunk cost vs part of a balanced portfolio. One perspective is to treat this as an investment that is somehow different than the rest of my portfolio, such as a house or antique art, etc. I think that is a valid thing to do. A house or art can increase in value dramatically, or it can go to zero (e.g. fire), so it has similar idiosyncratic risks involved. However, with both a house or art, I can buy insurance to protect against these risks. So under this mental model, perhaps I should buy put options once the portfolio gets large enough? (as opposed to simply de-leveraging). I accept that the portfolio can grow large or go to zero, but the sequence of first growing large and then going to zero would be a real bummer.
If the account were to grow so much that it would hurt that bad to see it go to zero, then obviously something is working right with the portfolio. Therefore, instead of moving funds away from it, I would agree with the use of the put options. You would probably have to give up ~2% per year to protect against a 50% drawdown, but that's better than giving up 12% per year (using earlier numbers) by moving it from the 20% return fund to the 8% return fund.

interestediniras
Posts: 2
Joined: Thu Jan 10, 2019 2:06 am

Re: HEDGEFUNDIE's excellent adventure [risk parity strategy using 3x leveraged ETFs]

Post by interestediniras » Thu Feb 14, 2019 2:12 pm

EfficientInvestor: What do you think about simply replicating the asset allocation of the All-Weather Fund, i.e. 30% equities, 40% long term Treasury bonds, 15% intermediate term Treasury bonds, 7.5% gold, and 7.5% commodities?

EfficientInvestor
Posts: 92
Joined: Thu Nov 01, 2018 7:02 pm

Re: HEDGEFUNDIE's excellent adventure [risk parity strategy using 3x leveraged ETFs]

Post by EfficientInvestor » Thu Feb 14, 2019 2:26 pm

interestediniras wrote:
Thu Feb 14, 2019 2:12 pm
EfficientInvestor: What do you think about simply replicating the asset allocation of the All-Weather Fund, i.e. 30% equities, 40% long term Treasury bonds, 15% intermediate term Treasury bonds, 7.5% gold, and 7.5% commodities?
I think it's a great idea because that mix has historically produced a very high Sharpe ratio. That was the whole point of Dalio/Bridgewater coming up with that asset allocation. See the historical efficient frontier graph since 1972 at the link below using stock, 10-year treasuries, and gold. The tangency point (highest Sharpe ratio) is basically the same mix as the All-Weather fund.
https://www.portfoliovisualizer.com/eff ... sset3=Gold
If you are going to use leverage, you should apply it to an AA with the highest Sharpe ratio possible. That being said, the leveraged funds for commodities are limited, so you may need to skip those and just use gold as the inflation hedge. Therefore, you could consider 35% equities, 50% treasuries, and 15% gold. I like to add in 10% REIT for extra diversification, so my AA is currently around 35% stock, 10% REIT, 45% LTT, 10% gold. You could sub out the LTT for 7-10 year treasuries if you are concerned about rising interest rates leading to poor performance on the longer bonds.

interestediniras
Posts: 2
Joined: Thu Jan 10, 2019 2:06 am

Re: HEDGEFUNDIE's excellent adventure [risk parity strategy using 3x leveraged ETFs]

Post by interestediniras » Thu Feb 14, 2019 2:34 pm

EfficientInvestor wrote:
Thu Feb 14, 2019 2:26 pm
interestediniras wrote:
Thu Feb 14, 2019 2:12 pm
EfficientInvestor: What do you think about simply replicating the asset allocation of the All-Weather Fund, i.e. 30% equities, 40% long term Treasury bonds, 15% intermediate term Treasury bonds, 7.5% gold, and 7.5% commodities?
I think it's a great idea because that mix has historically produced a very high Sharpe ratio. That was the whole point of Dalio/Bridgewater coming up with that asset allocation. See the historical efficient frontier graph since 1972 at the link below using stock, 10-year treasuries, and gold. The tangency point (highest Sharpe ratio) is basically the same mix as the All-Weather fund.
https://www.portfoliovisualizer.com/eff ... sset3=Gold
If you are going to use leverage, you should apply it to an AA with the highest Sharpe ratio possible. That being said, the leveraged funds for commodities are limited, so you may need to skip those and just use gold as the inflation hedge. Therefore, you could consider 35% equities, 50% treasuries, and 15% gold. I like to add in 10% REIT for extra diversification, so my AA is currently around 35% stock, 10% REIT, 45% LTT, 10% gold. You could sub out the LTT for 7-10 year treasuries if you are concerned about rising interest rates leading to poor performance on the longer bonds.
Thanks for the comments. Yes, in general I like the approach of first choosing an asset allocation which is efficient, and then levering up to obtain desired returns, which is a more theoretically principled approach than just sliding the scale between 0/100 and 100/0 with unlevered indices, which only gives you one dimension of variation even though you want to optimize for two outcomes: returns and efficiency. Is this in line with your own thinking?

Can I also ask what proportion of your overall portfolio is invested using a levered approach, and also if you are still including sector-specific levered funds in your 35% stock? Personally, I'm inclined to shift over a majority of my portfolio to a levered approach, albeit not necessarily all at 3X.

EfficientInvestor
Posts: 92
Joined: Thu Nov 01, 2018 7:02 pm

Re: HEDGEFUNDIE's excellent adventure [risk parity strategy using 3x leveraged ETFs]

Post by EfficientInvestor » Thu Feb 14, 2019 3:11 pm

interestediniras wrote:
Thu Feb 14, 2019 2:34 pm
EfficientInvestor wrote:
Thu Feb 14, 2019 2:26 pm
interestediniras wrote:
Thu Feb 14, 2019 2:12 pm
EfficientInvestor: What do you think about simply replicating the asset allocation of the All-Weather Fund, i.e. 30% equities, 40% long term Treasury bonds, 15% intermediate term Treasury bonds, 7.5% gold, and 7.5% commodities?
I think it's a great idea because that mix has historically produced a very high Sharpe ratio. That was the whole point of Dalio/Bridgewater coming up with that asset allocation. See the historical efficient frontier graph since 1972 at the link below using stock, 10-year treasuries, and gold. The tangency point (highest Sharpe ratio) is basically the same mix as the All-Weather fund.
https://www.portfoliovisualizer.com/eff ... sset3=Gold
If you are going to use leverage, you should apply it to an AA with the highest Sharpe ratio possible. That being said, the leveraged funds for commodities are limited, so you may need to skip those and just use gold as the inflation hedge. Therefore, you could consider 35% equities, 50% treasuries, and 15% gold. I like to add in 10% REIT for extra diversification, so my AA is currently around 35% stock, 10% REIT, 45% LTT, 10% gold. You could sub out the LTT for 7-10 year treasuries if you are concerned about rising interest rates leading to poor performance on the longer bonds.
Thanks for the comments. Yes, in general I like the approach of first choosing an asset allocation which is efficient, and then levering up to obtain desired returns, which is a more theoretically principled approach than just sliding the scale between 0/100 and 100/0 with unlevered indices, which only gives you one dimension of variation even though you want to optimize for two outcomes: returns and efficiency. Is this in line with your own thinking?

Can I also ask what proportion of your overall portfolio is invested using a levered approach, and also if you are still including sector-specific levered funds in your 35% stock? Personally, I'm inclined to shift over a majority of my portfolio to a levered approach, albeit not necessarily all at 3X.
Yes, that is in line with my thinking. As for the portion of my portfolio invested using the leveraged approach...like others, I limit it to an amount that if it went to zero, it wouldn't greatly affect me. I don't believe this will happen because I believe in the theory and believe the funds operate well enough to stay in line with the theory. But, it is better to be safe than sorry. All that said, I currently have about 10% of my retirement accounts in the leveraged approach. I will not put any more into the leveraged portion of my retirement account, but I also won't remove any of the profits. If the idea works, it will be a nice sum in 30 years. If it doesn't, I still have the other 90% along with all the 401k contributions I will continue to make towards my normal workplace retirement plan. In addition, I also have leveraged funds in my non-retirement account. This represents about 25% of my investible money and is essentially my retire early fund. If the concept works, then I can retire much earlier than I otherwise planned on. If it crashes and goes to zero, I am still on path with my normal retirement accounts to retire by 60.

As for sector-specific funds...yes I'm still in TQQQ (tech), CURE (healthcare), and RETL (consumer cyclical). I prefer tilts towards the stock sectors that tend to do the best when stocks are doing well. Therefore, I weed out utilities and consumer staples by going towards these sectors. I'm sure many on this forum would balk at that logic and would say to just go with the broader index. If I wasn't doing this, I would probably split between large, mid, and small cap indexes.

staythecourse
Posts: 6520
Joined: Mon Jan 03, 2011 9:40 am

Re: HEDGEFUNDIE's excellent adventure [risk parity strategy using 3x leveraged ETFs]

Post by staythecourse » Thu Feb 14, 2019 4:02 pm

jaj2276 wrote:
Thu Feb 14, 2019 12:59 pm
staythecourse wrote:
Thu Feb 14, 2019 11:59 am
sparksfly wrote:
Thu Feb 14, 2019 11:27 am
My M1 Finance Roth IRA was opened and funded yesterday. Auto invest triggered the buy orders this morning in the portfolio. Very small part of the portfolio but I'm in :sharebeer
How long from when you emailed them your statement to when the transactions occurred? I'm trying to clean up some accounts right now before making this move and just trying to get an idea of timeline?

Good luck.
I transferred a Roth IRA from Vanguard. Started the process on Feb 6, had the money available today (Feb 14). First trade will actually go through tomorrow during their trading window.

Maybe because I'm not a millenial, but M1 Finance's website isn't entirely intuitive to me. I had created a Pie (called My Pie) that was 45/55 but after thinking about it I wanted to change it to 40/60. However I couldn't find a way to edit 'My Pie' and so I created another pie. All somewhat confusing but I think I'm on track.

So as of tomorrow, 2.5% of my portfolio will be invested in this strategy.
Much thanks so about 7-10 days. Sounds reasonable. Agree their website is NOT user friendly. It should be a bit easier, but maybe it is once you have some money in the pies?

Good luck.
"The stock market [fluctuation], therefore, is noise. A giant distraction from the business of investing.” | -Jack Bogle

samsdad
Posts: 516
Joined: Sat Jan 02, 2016 6:20 pm

Re: HEDGEFUNDIE's excellent adventure [risk parity strategy using 3x leveraged ETFs]

Post by samsdad » Thu Feb 14, 2019 5:35 pm

Interesting reading for those who are concerned about counterparty risk. The figures and tables referenced in the paper are in the back by the way.

https://efmaefm.org/0efmameetings/EFMA% ... lpaper.pdf
Abstract
As most Exchange-Traded Funds (ETFs) engage in securities lending or are based on total return swaps, they expose their investors to counterparty risk. In this paper, we present a framework to study counterparty risk and provide empirical estimates for a sample of physical and synthetic funds. Our findings contradict the allegations made by international agencies about the poor quality of the collateral used by ETFs. We find that the counterparty risk exposure is higher for synthetic ETFs but that investors are compensated for bearing this risk. Finally, we theoretically show how to construct an optimal collateral portfolio for an ETF.
Further into the paper:
In order to quantify the counterparty risk exposure of investors, we propose two original risk metrics: (1) the probability for a fund of not having enough collateral on the following day and (2) the magnitude of the collateral expected shortfall conditional on not having enough collateral. Both measures are computed conditionally on the default of the fund counterparty. Overall we find that the estimated probability of being undercollateralized is substantial but that the collateral shortfall remains moderate. However, counterparty risk exposure turns out to be higher for inverse funds and for funds tracking commodities or exchange rates. When contrasting the level of counterparty risk of investors investing in synthetic and physical ETFs, we find that counterparty risk exposure is higher for synthetic funds but that investors are compensated for bearing this risk through lower tracking errors but similar fees.
To the best of our knowledge, our study is the first attempt to assess empirically the quality of the collateral for a large and representative sample of ETFs.
Perhaps those of you with bigger brains and a fondness of math that looks like hieroglyphics can tell us within a few decimal points what the risk is, exactly.
_______
EDIT: It looks like some 96% of the expected synthetic ETF collateral shortfalls in the study are between 0 and 5% of the NAV (with about 50% of that being between 0 and 0.5%); 85% of all synthetic ETF collateral shortfalls in the study would be expected to be $5MM USD or less. See pg. 16 and Table 4 at end.

In contrast, about 99% of the physical ETFs have an expected collateral shortfall of between 0 and 0.5% of NAV. The remaining 1% or so would be between 0.5-1%; about 87% of all physical ETF collateral shortfalls in the study would be expected to be $1MM USD or less.
Last edited by samsdad on Thu Feb 14, 2019 6:00 pm, edited 1 time in total.

jaj2276
Posts: 382
Joined: Sat Apr 16, 2011 5:13 pm

Re: HEDGEFUNDIE's excellent adventure [risk parity strategy using 3x leveraged ETFs]

Post by jaj2276 » Thu Feb 14, 2019 5:53 pm

Since I plan to use this thread share in the joy and despair (hopefully more joy) of this strategy, I wanted to easily be able to get back to my initial investment. $54,741 Roth IRA. I will continue to direct future Roth IRA contributions to this strategy.

Edit 1: Trade has been completed. I bought 1683.4941 shares of TMF @ $19.51. I bought 472.92958 shares of UPRO at $46.30. Down $60 as of this edit :shock:

Edit 2: Last edit I'll make on this post and my strategy performance until first rebal. After day 1, +0.18%. I can smell the $5m :P
Last edited by jaj2276 on Fri Feb 15, 2019 4:11 pm, edited 2 times in total.

finvestor
Posts: 30
Joined: Thu Apr 20, 2017 1:22 pm

Re: HEDGEFUNDIE's excellent adventure [risk parity strategy using 3x leveraged ETFs]

Post by finvestor » Fri Feb 15, 2019 7:06 am

Has anyone implemented (or considered implementing) this strategy in Europe? If so, which ETFs are you using? These days most europeans do not have access to US domiciled ETFs, so I was wondering if anyone knows good Europe domiciled ETFs to implement this strategy?

LawnGnomeGenome
Posts: 8
Joined: Sun Feb 03, 2013 5:41 pm

Re: HEDGEFUNDIE's excellent adventure [risk parity strategy using 3x leveraged ETFs]

Post by LawnGnomeGenome » Fri Feb 15, 2019 8:34 am

Hi everyone,

To piggy-back on what aj76er said, a big thank you for a very interesting and enlightening conversation on this topic to the aforementioned contributors - HEDGEFUNDIE, EfficientInvestor, et. al. I have decided to join in the fun, but a bit differently. I liked the idea hdas proposed of the Maximizing Sharpe Ratio (MSR) strategy, so I am implementing a mix of 10% UPRO, 45% TMF, and 45% TQQQ with quarterly rebalancing. The current MSR allocation calls for 13.41% UPRO, 44.66% TMF, and 41.93% TQQQ based on data starting March 2010, but since I am currently investing in a taxable Robinhood account, that's a PITA to try to adhere to with such a relative small amount invested. Rather surprisingly, PV says that the 10/45/45 has had a better return (37.00% vs 36.64% CAGR) over the same time period for the same Sharpe (1.45) and Sortino Ratio (2.85).

I plan to move to M1 once I get to the full allocation of funds, which I expect to be 1-2% of my current portfolio, so I can keep my tax lots easier to manage. This is absolutely subject to change since the current PV analysis can only backtest to March 2010 due to TQQQ's inception, and I'm open to switching to the 40% UPRO / 60% TMF as market conditions dictate.

I will watch HEDGEFUNDIE's 40% UPRO / 60% TMF, and would be interested to see how our portfolios behave. Would anyone be interested in posting return comparisons, with perhaps weekly/monthly updates? Best of luck (and hopefully good fortune) to all that participate.

For science!

unknownfuture
Posts: 1
Joined: Fri Feb 15, 2019 3:30 pm

Re: HEDGEFUNDIE's excellent adventure [risk parity strategy using 3x leveraged ETFs]

Post by unknownfuture » Fri Feb 15, 2019 3:51 pm

First of all, thank you for this thread; I appreciate it.

1) Some people in this thread have expressed fear that if too many people pile on this strategy, it will stop working. How realistic is this fear? The underlying S&P 500 stocks and LTT bonds have enormous market cap, so changing the UPRO and TMF prices would require a very large force. It is therefore my understanding that strategy will keep working unless investors with a very large amount of money (hundreds of billions of dollars?) starts exploiting it; which in this case seems unlikely. Is this understanding correct? Or are there reasons to think that this strategy will stop working even when a small fraction of the market will act on it? For example, could a small number of actors use options to create enormous leverage and remove this inefficiency?

2) Is there a place where I can find simulated time series data for the UPRO and TMF funds?

Thanks!
Last edited by unknownfuture on Fri Feb 15, 2019 4:35 pm, edited 1 time in total.

samsdad
Posts: 516
Joined: Sat Jan 02, 2016 6:20 pm

Re: HEDGEFUNDIE's excellent adventure [risk parity strategy using 3x leveraged ETFs]

Post by samsdad » Fri Feb 15, 2019 4:35 pm

unknownfuture wrote:
Fri Feb 15, 2019 3:51 pm
First of all, thank you for this thread; I appreciate it.

1) Some people in this thread have expressed fear that if too many people pile on this strategy, it will stop working. How realistic is this fear? The underlying S&P 500 stocks and LTT bonds have enormous market cap, so changing the UPRO and TMF prices would require a very large force. It is therefore my understanding that the "inefficiency" will be remain unless a very large number of high-cap actors act upon it. Is this understanding correct? Or are there reasons to think that this strategy will stop working even when a small fraction of the market will act on it? For example, could a small number of actors use options to create enormous leverage and remove this inefficiency?

2) Is there a place where I can find simulated time series data for the UPRO and TMF funds?

Thanks!
1. I'll leave others to discuss no. 1.
2. It's being worked on here: viewtopic.php?f=10&t=272640.

Jags4186
Posts: 2882
Joined: Wed Jun 18, 2014 7:12 pm

Re: HEDGEFUNDIE's excellent adventure [risk parity strategy using 3x leveraged ETFs]

Post by Jags4186 » Fri Feb 15, 2019 4:37 pm

unknownfuture wrote:
Fri Feb 15, 2019 3:51 pm
First of all, thank you for this thread; I appreciate it.

1) Some people in this thread have expressed fear that if too many people pile on this strategy, it will stop working. How realistic is this fear? The underlying S&P 500 stocks and LTT bonds have enormous market cap, so changing the UPRO and TMF prices would require a very large force. It is therefore my understanding that strategy will keep working unless investors with a very large amount of money (hundreds of billions of dollars?) starts exploiting it; which in this case seems unlikely. Is this understanding correct? Or are there reasons to think that this strategy will stop working even when a small fraction of the market will act on it? For example, could a small number of actors use options to create enormous leverage and remove this inefficiency?

Thanks!
Considering this isn’t a new strategy and at most couple dozen people here are now throwing maybe a total of a couple $100,000 into it...doubtful.

staythecourse
Posts: 6520
Joined: Mon Jan 03, 2011 9:40 am

Re: HEDGEFUNDIE's excellent adventure [risk parity strategy using 3x leveraged ETFs]

Post by staythecourse » Fri Feb 15, 2019 4:46 pm

Jags4186 wrote:
Fri Feb 15, 2019 4:37 pm
unknownfuture wrote:
Fri Feb 15, 2019 3:51 pm
First of all, thank you for this thread; I appreciate it.

1) Some people in this thread have expressed fear that if too many people pile on this strategy, it will stop working. How realistic is this fear? The underlying S&P 500 stocks and LTT bonds have enormous market cap, so changing the UPRO and TMF prices would require a very large force. It is therefore my understanding that strategy will keep working unless investors with a very large amount of money (hundreds of billions of dollars?) starts exploiting it; which in this case seems unlikely. Is this understanding correct? Or are there reasons to think that this strategy will stop working even when a small fraction of the market will act on it? For example, could a small number of actors use options to create enormous leverage and remove this inefficiency?

Thanks!
Considering this isn’t a new strategy and at most couple dozen people here are now throwing maybe a total of a couple $100,000 into it...doubtful.
Even more then that is that the products themselves are NOT meant for long term investors. The fact they are geared towards active investors means there really won't be many folks even considering holding this concept as a buy/ hold strategy. The few that do there will be likely even fewer who will hold for a period of even >10 years. So, i just don't see how a concept like this can have a market impact on the underlying funds.

Those of us that will be trying this REALLY are the fish swimming upstream.

Good luck.
"The stock market [fluctuation], therefore, is noise. A giant distraction from the business of investing.” | -Jack Bogle

FIREmeup
Posts: 22
Joined: Thu Jun 15, 2017 4:00 pm

Re: HEDGEFUNDIE's excellent adventure [risk parity strategy using 3x leveraged ETFs]

Post by FIREmeup » Fri Feb 15, 2019 5:14 pm

staythecourse wrote:
Fri Feb 15, 2019 4:46 pm
Jags4186 wrote:
Fri Feb 15, 2019 4:37 pm
unknownfuture wrote:
Fri Feb 15, 2019 3:51 pm
First of all, thank you for this thread; I appreciate it.

1) Some people in this thread have expressed fear that if too many people pile on this strategy, it will stop working. How realistic is this fear? The underlying S&P 500 stocks and LTT bonds have enormous market cap, so changing the UPRO and TMF prices would require a very large force. It is therefore my understanding that strategy will keep working unless investors with a very large amount of money (hundreds of billions of dollars?) starts exploiting it; which in this case seems unlikely. Is this understanding correct? Or are there reasons to think that this strategy will stop working even when a small fraction of the market will act on it? For example, could a small number of actors use options to create enormous leverage and remove this inefficiency?

Thanks!
Considering this isn’t a new strategy and at most couple dozen people here are now throwing maybe a total of a couple $100,000 into it...doubtful.
Even more then that is that the products themselves are NOT meant for long term investors. The fact they are geared towards active investors means there really won't be many folks even considering holding this concept as a buy/ hold strategy. The few that do there will be likely even fewer who will hold for a period of even >10 years. So, i just don't see how a concept like this can have a market impact on the underlying funds.

Those of us that will be trying this REALLY are the fish swimming upstream.

Good luck.
And even more, the concern of too many people ruining this long term trade is zero. Not non zero, actual zero. These products are based on the bond market and the stock market, millions of dollars wouldn't be enough to influence them as it is easily hedgeable.

That being said I wouldn't buy or sell these ETNs without a limit price, like any stock because a large order can execute at weird prices but then get back to equilibrium within a half second.

I find this strategy extremely interesting. I am in on this trade for 16k, which is a tiny portion of my portfolio.

150 shares of SPXL 43.75
500 shares TMF 19.50

Good luck!

PluckyDucky
Posts: 22
Joined: Tue Jan 15, 2019 8:29 pm

Re: HEDGEFUNDIE's excellent adventure [risk parity strategy using 3x leveraged ETFs]

Post by PluckyDucky » Fri Feb 15, 2019 5:23 pm

I'm also setting up a portion of my investing in a strategy similar to this.

Is this thread the bogleheads equivalent of wall street bets?

:greedy

Post Reply