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.
Topic Author
HEDGEFUNDIE
Posts: 3973
Joined: Sun Oct 22, 2017 2:06 pm

Re: HEDGEFUNDIE's excellent adventure [3x leveraged ETF strategy]

Post by HEDGEFUNDIE » Wed Feb 06, 2019 6:06 pm

UnderWater wrote:
Wed Feb 06, 2019 4:34 pm
Jags4186 wrote:
Wed Feb 06, 2019 4:21 pm
UnderWater wrote:
Wed Feb 06, 2019 4:14 pm
HEDGEFUNDIE wrote:
Tue Feb 05, 2019 2:56 pm
Here is how 60% TMF & 40% UPRO, rebalanced quarterly, would have performed over 1987-2018, as compared to the S&P 500:
Image
[ quote formatted by admin LadyGeek]

I seem to be missing something very basic. When I run a 23.75% CAGR on a $10k initial investment, I wind up with $709k after 20 years. To hit the $10MM or so mark, you'd need a CAGR north of 40%?
1987-2018 is 32 years
Yes, that would the something very basic that I was missing. Latched onto HEDGEFUNDIE's 20-yr time horizon I guess.
I agree this was confusing, so I re-ran the backtest to start with $100k, as I am doing in real life. Original post image updated.

Topic Author
HEDGEFUNDIE
Posts: 3973
Joined: Sun Oct 22, 2017 2:06 pm

Re: HEDGEFUNDIE's excellent adventure [3x leveraged ETF strategy]

Post by HEDGEFUNDIE » Wed Feb 06, 2019 6:15 pm

Here is a fun question for everyone.

Let's assume the strategy actually works as promised, interest rates stay low, and I get ~25% CAGR.

By year 10, my $100k will have grown to $1M. Pretty sure I won't have a problem staying the course to that point.

By year 15, my account will have grown to $3M. Again, no problem staying the course.

By year 18, my account will be at $5.9M... Hmmm.... :|

By year 20, my account will be at $9.3M :shock:

You see where I am going with this. If someone was actually able to hold on for the entire 32 years of the backtest, he / she would be sitting on $100M right now.

At what point would you bail on the strategy, and why?

A big reason I started this thread is because I want the BH community to hold me accountable to my plan. So if I should change my goals, let's make those changes now!
Last edited by HEDGEFUNDIE on Sat Feb 16, 2019 3:52 pm, edited 1 time in total.

Tanelorn
Posts: 1673
Joined: Thu May 01, 2014 9:35 pm

Re: HEDGEFUNDIE's excellent adventure [3x leveraged ETF strategy]

Post by Tanelorn » Wed Feb 06, 2019 6:40 pm

HEDGEFUNDIE wrote:
Wed Feb 06, 2019 6:15 pm
You see where I am going with this. If someone was actually able to hold on for the entire 32 years of the backtest, he / she would be sitting on $100M right now.

At what point would you bail on the strategy, and why?
Maybe you’d want to preserve your profits in a way that couldn’t get hit with some sort of black swan blowup? If so, either have a % of your portfolio in a more traditional allocation of fixed income or a minimum amount relative to your total savings goals in normal investments (ie $3M or whatever). For strategies you think will do well, they’re good to invest in at least prospectively but you want to make sure your size their allocation appropriate to their volatility and downside risk. Ponzi schemes work great if you get in early, but only if you keep cashing out your proceeds - let it all ride and you may find out the XIV or Madoff lesson more dramatically.

At least for this type of long-only, no explicit margin type of investment, the worst you can lose is everything (like for a 3x long ETF if the underlying market drops >33% in a day, if they even allow that with broad marker circuit breakers these days).

jaj2276
Posts: 473
Joined: Sat Apr 16, 2011 5:13 pm

Re: HEDGEFUNDIE's excellent adventure [3x leveraged ETF strategy]

Post by jaj2276 » Wed Feb 06, 2019 8:09 pm

vineviz wrote:
Tue Feb 05, 2019 4:11 pm
HEDGEFUNDIE wrote:
Tue Feb 05, 2019 2:56 pm
What is your goal?
$10M. If past is prologue, I should be able to hit this number in 20 years when I am 53 years old, at which point I will stop "playing the game" and move into [unleveraged] Treasuries. Penalty-free Roth IRA withdrawals begin, as we all know, at 59.5.
When I've looked at this strategy (and if I were younger I'd probably be actually doing it), the strategy that made the most sense to me was to work out a "target equity exposure" for each year and then gradually deleverage in order to maintain that.

For instance, you're starting out with $120k in equity exposure (40% x 3) and let's state your goal as having $12m in equity exposure at age 53 (aka a total portfolio of $10m)

Let's say at age 40 your portfolio is worth $650,000. That's equity exposure of $780k but your equity exposure target at that age is only $600. You'd remove enough from the leveraged portfolio to reduce your equity exposure back down to your target.

A policy similar to this one is more precisely managing your risk exposure as opposed to your capital investment. A string of good/lucky years up front would mean taking some money off the table earlier so that you are maintaining a similar risk profile through time, rather than a similar capital allocation profile.
Vineviz,

How are you determining that his equity exposure target at age 40 is $600k?

My wife (her 38, me 43) has a small Roth that really won't do much to our retirement picture if it ends up at $0. She started late so her account is only $50k. Maybe it can get to $10m if we let it ride a bit longer or maybe at 63 we decide that its value of $5m is more than sufficient to not let it ride for $10m. Regardless, I'd like to try this in her Roth because its final value of $0 (a distinct possibility) wouldn't change our retirement picture. I don't want totally ape HedgeFundie's strategy so I'd like to put in a few different wrinkles (thinking maybe 50/50, rebalance every 4 months, and maybe use equity exposure as opposed to capital allocation during rebalance events).

Edit 1: Forgot to add that I've already opened an M1F account and have liquidated the wife's current holdings of 50% bonds, 50% SCV. Hopefully the transfer process is smooth.

bgf
Posts: 1170
Joined: Fri Nov 10, 2017 9:35 am

Re: HEDGEFUNDIE's excellent adventure [3x leveraged ETF strategy]

Post by bgf » Wed Feb 06, 2019 8:17 pm

i dont understand one thing. why not put all 15% in UPRO?

we're only talking about 15% of your portfolio to begin with, an amount you are willing and able to lose, potentially.

i get the hogwash about higher sharpe ratios yada yada but why do you care? who cares if the sharpe ratio is higher? again, this 15% is on "lock down," so the volatility should be irrelevant to you. your one and only focus should be maximizing return. why else invest in a 3x ETF?

just go 15% UPRO and let it rip for 20 years.
“TE OCCIDERE POSSUNT SED TE EDERE NON POSSUNT NEFAS EST"

jaj2276
Posts: 473
Joined: Sat Apr 16, 2011 5:13 pm

Re: HEDGEFUNDIE's excellent adventure [3x leveraged ETF strategy]

Post by jaj2276 » Wed Feb 06, 2019 8:25 pm

bgf wrote:
Wed Feb 06, 2019 8:17 pm
i dont understand one thing. why not put all 15% in UPRO?

we're only talking about 15% of your portfolio to begin with, an amount you are willing and able to lose, potentially.

i get the hogwash about higher sharpe ratios yada yada but why do you care? who cares if the sharpe ratio is higher? again, this 15% is on "lock down," so the volatility should be irrelevant to you. your one and only focus should be maximizing return. why else invest in a 3x ETF?

just go 15% UPRO and let it rip for 20 years.
I don't know the quantifiable reason but I suspect it's due to being able to buy more UPRO throughout the course of the strategy using the overweighted holding of TMF. If you only held UPRO then of course you'd end up with UPROs returns. But if you held both and you rebalanced, then a few of those times you'd buy "more UPRO" and eventually have more units of UPRO (with some TMF to boot) than simply starting with the entire portfolio dedicated to UPRO.

bgf
Posts: 1170
Joined: Fri Nov 10, 2017 9:35 am

Re: HEDGEFUNDIE's excellent adventure [3x leveraged ETF strategy]

Post by bgf » Wed Feb 06, 2019 8:30 pm

jaj2276 wrote:
Wed Feb 06, 2019 8:25 pm
bgf wrote:
Wed Feb 06, 2019 8:17 pm
i dont understand one thing. why not put all 15% in UPRO?

we're only talking about 15% of your portfolio to begin with, an amount you are willing and able to lose, potentially.

i get the hogwash about higher sharpe ratios yada yada but why do you care? who cares if the sharpe ratio is higher? again, this 15% is on "lock down," so the volatility should be irrelevant to you. your one and only focus should be maximizing return. why else invest in a 3x ETF?

just go 15% UPRO and let it rip for 20 years.
I don't know the quantifiable reason but I suspect it's due to being able to buy more UPRO throughout the course of the strategy using the overweighted holding of TMF. If you only held UPRO then of course you'd end up with UPROs returns. But if you held both and you rebalanced, then a few of those times you'd buy "more UPRO" and eventually have more units of UPRO (with some TMF to boot) than simply starting with the entire portfolio dedicated to UPRO.
i get what you're saying. its no different than rebalancing from bonds into stock... problem is, in both cases your expected return is lower. rebalancing does not boost returns when you are rebalancing from a lower return instrument to a higher one. you're better off just owning 100% of the higher returning asset 100% of the time.
“TE OCCIDERE POSSUNT SED TE EDERE NON POSSUNT NEFAS EST"

jaj2276
Posts: 473
Joined: Sat Apr 16, 2011 5:13 pm

Re: HEDGEFUNDIE's excellent adventure [3x leveraged ETF strategy]

Post by jaj2276 » Wed Feb 06, 2019 8:42 pm

bgf wrote:
Wed Feb 06, 2019 8:30 pm
...

i get what you're saying. its no different than rebalancing from bonds into stock... problem is, in both cases your expected return is lower. rebalancing does not boost returns when you are rebalancing from a lower return instrument to a higher one. you're better off just owning 100% of the higher returning asset 100% of the time.
I'm not questioning the validity of that statement. The way you worded it has me wondering about the following question. What if you are rebalancing from a higher return instrument to a lower return instrument (and also from lower to higher)?

Let's say the higher instrument runs up to where you have a lot more UPRO than what you started out with. You rebalance by selling some UPRO and buying more TMF. Obv you have less units of UPRO and more units of TMF but same capital allocation between the two. Then UPRO goes down while TMF goes up. You then take some of the TMF and buy UPRO. It could be that you could buy more UPRO than you sold during the first rebalance and end up with the same amt (or more) of TMF. If you then froze any additional moves in the account over the next 30 years, the account might be higher than simply 100% UPRO. Obviously if you do rebalance more often then you get even more of these opportunities.

It's no guarantee of course that the rebalance opportunities will present themselves as neatly as my description.

gtwhitegold
Posts: 436
Joined: Fri Sep 21, 2012 1:55 pm

Re: HEDGEFUNDIE's excellent adventure [3x leveraged ETF strategy]

Post by gtwhitegold » Wed Feb 06, 2019 8:45 pm

bgf wrote:
Wed Feb 06, 2019 8:30 pm
jaj2276 wrote:
Wed Feb 06, 2019 8:25 pm
bgf wrote:
Wed Feb 06, 2019 8:17 pm
i dont understand one thing. why not put all 15% in UPRO?

we're only talking about 15% of your portfolio to begin with, an amount you are willing and able to lose, potentially.

i get the hogwash about higher sharpe ratios yada yada but why do you care? who cares if the sharpe ratio is higher? again, this 15% is on "lock down," so the volatility should be irrelevant to you. your one and only focus should be maximizing return. why else invest in a 3x ETF?

just go 15% UPRO and let it rip for 20 years.
I don't know the quantifiable reason but I suspect it's due to being able to buy more UPRO throughout the course of the strategy using the overweighted holding of TMF. If you only held UPRO then of course you'd end up with UPROs returns. But if you held both and you rebalanced, then a few of those times you'd buy "more UPRO" and eventually have more units of UPRO (with some TMF to boot) than simply starting with the entire portfolio dedicated to UPRO.
i get what you're saying. its no different than rebalancing from bonds into stock... problem is, in both cases your expected return is lower. rebalancing does not boost returns when you are rebalancing from a lower return instrument to a higher one. you're better off just owning 100% of the higher returning asset 100% of the time.
One good reason is to have an uncorrelated asset to rebalance with which should reduce the decay from daily rebalancing of the ETF. If that is the sole purpose, then you won't need more than about a 10% allocation I would guess.

hohum
Posts: 26
Joined: Thu Jan 03, 2019 2:34 pm

Re: HEDGEFUNDIE's excellent adventure [3x leveraged ETF strategy]

Post by hohum » Wed Feb 06, 2019 8:50 pm

If you are in UPRO 100%, you will go down 90% in 2008. How many years of 24% returns does it take to recover from that?

You need an uncorrelated asset, and you need a lot of it.

User avatar
willthrill81
Posts: 15063
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: HEDGEFUNDIE's excellent adventure [3x leveraged ETF strategy]

Post by willthrill81 » Wed Feb 06, 2019 9:06 pm

I wish you luck, although as long as the costs of holding the funds aren't too high, I don't think that you'll need it to succeed.
“It's a dangerous business, Frodo, going out your door. You step onto the road, and if you don't keep your feet, there's no knowing where you might be swept off to.” J.R.R. Tolkien,The Lord of the Rings

AZAttorney11
Posts: 614
Joined: Wed Jan 21, 2015 12:12 pm

Re: HEDGEFUNDIE's excellent adventure [3x leveraged ETF strategy]

Post by AZAttorney11 » Wed Feb 06, 2019 9:11 pm

You've got huge... guts... HEDGEFUNDIE. I wish you nothing but the best!

EfficientInvestor
Posts: 260
Joined: Thu Nov 01, 2018 7:02 pm
Location: Alabama

Re: HEDGEFUNDIE's excellent adventure [3x leveraged ETF strategy]

Post by EfficientInvestor » Wed Feb 06, 2019 9:30 pm

HEDGEFUNDIE wrote:
Wed Feb 06, 2019 6:15 pm
A big reason I started this thread is because I want the BH community to hold me accountable to my plan. So if I should change my goals, let's make those changes now!
If you’re looking to make any changes, you could consider splitting the stocks among large, mid, and small cap. You could also consider adding some gold as an inflation hedge. See the backtest below for performance of various AAs since 1972. Based on these results, an updated AA you could consider is as follows:

13% - UPRO
14% - MIDU
13% - TNA
50% - TMF
10% - UGLD

https://www.portfoliovisualizer.com/bac ... 0&Gold3=10

bgf
Posts: 1170
Joined: Fri Nov 10, 2017 9:35 am

Re: HEDGEFUNDIE's excellent adventure [3x leveraged ETF strategy]

Post by bgf » Wed Feb 06, 2019 9:51 pm

hohum wrote:
Wed Feb 06, 2019 8:50 pm
If you are in UPRO 100%, you will go down 90% in 2008. How many years of 24% returns does it take to recover from that?

You need an uncorrelated asset, and you need a lot of it.
and then go up a whole bunch in 2009 - 2019... 100% UPRO has a CAGR of almost 30% since inception.
“TE OCCIDERE POSSUNT SED TE EDERE NON POSSUNT NEFAS EST"

staythecourse
Posts: 6993
Joined: Mon Jan 03, 2011 9:40 am

Re: HEDGEFUNDIE's excellent adventure [3x leveraged ETF strategy]

Post by staythecourse » Wed Feb 06, 2019 9:56 pm

Great thread. I'm always open to new ideas.

What I really want to see is the 3x sp500 segment in the flat 2000's. That should be a good estimate of what if any decay occurred WITH 1% expense ratio added. This flat 2000's is the most likely (not the worst) possible outcome over a 10 year period. One could always say the world could just fall apart, but that is really unlikely.

The other concern is the timing. For me, this is a bet I will keep in my back pocket if we have anything close to the 2008 crises again.

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

reformed.trader
Posts: 78
Joined: Sat Apr 15, 2017 11:14 pm

Re: HEDGEFUNDIE's excellent adventure [3x leveraged ETF strategy]

Post by reformed.trader » Wed Feb 06, 2019 9:57 pm

SIAP

But from the 80s until basically today, bonds and equities have both gone up. And when there has been a sharp, short drawdown in stocks, it has been matched with a large upswing in bonds. When this correlation pattern breaks, this trade will be dead imo, and I am of the impression that the next "big one" will be BECAUSE rates are rising. Good luck though.

User avatar
Oicuryy
Posts: 1348
Joined: Thu Feb 22, 2007 10:29 pm

Re: HEDGEFUNDIE's excellent adventure [3x leveraged ETF strategy]

Post by Oicuryy » Wed Feb 06, 2019 10:01 pm

Thanks, EfficientInvestor and vineviz.

Minus 4% seemed like a lot to lose even on an annual basis. It made me think HEDGEFUNDIE should factor that into his simulated returns. But maybe not.

Ron
Money is fungible | Abbreviations and Acronyms

Jags4186
Posts: 4073
Joined: Wed Jun 18, 2014 7:12 pm

Re: HEDGEFUNDIE's excellent adventure [3x leveraged ETF strategy]

Post by Jags4186 » Wed Feb 06, 2019 10:11 pm

reformed.trader wrote:
Wed Feb 06, 2019 9:57 pm
SIAP

But from the 80s until basically today, bonds and equities have both gone up. And when there has been a sharp, short drawdown in stocks, it has been matched with a large upswing in bonds. When this correlation pattern breaks, this trade will be dead imo, and I am of the impression that the next "big one" will be BECAUSE rates are rising. Good luck though.
Since 1928 there have only been 4 years where long term treasuries and stocks have moved down together.

1931, 1941, 1969, and 2018

http://pages.stern.nyu.edu/~adamodar/Ne ... retSP.html

staythecourse
Posts: 6993
Joined: Mon Jan 03, 2011 9:40 am

Re: HEDGEFUNDIE's excellent adventure [3x leveraged ETF strategy]

Post by staythecourse » Wed Feb 06, 2019 10:23 pm

Just curious... what happens if the fund itself goes bunk? Is there a stewardship issue? I am assuming your money is just liquidated and you find another 3x leveraged fund? If a 2008 crises occurred is there a real possibility the fund shuts its doors and by the time the recovery period starts and a new fund is formed one loses out on some of those magical returns from the market bottom?

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

random_person
Posts: 8
Joined: Wed Feb 06, 2019 10:11 pm

Re: HEDGEFUNDIE's excellent adventure [3x leveraged ETF strategy]

Post by random_person » Wed Feb 06, 2019 10:39 pm

I spent a lot of time thinking about leveraged ETFs in recent years. I came to two main conclusions:

1. "Volatility decay" is not a thing
2. Leveraged ETFs under perform their benchmarks in the data and this is because of trading costs from rebalancing

So for the first point, many people against leveraged ETFs often present examples like the following:

Imagine for simplicity we forget about transactions costs to focus only on the impact of daily rebalancing. Say you have an underlying benchmark that goes up 10% one day and then down 10% the next day. The underlying benchmark goes from 1 dollar to 1.1 and then down to .99 over the course of the two day period. However, the leveraged ETF goes from 1 dollar to 1.2 and then down to .96. Over the course of a two day period, the underlying lost 1% of its value and the leveraged ETF lost 4%. This is what people usually refer to as "volatility decay".

What is the problem with this example? It focuses on a period where the underlying asset mean reverted. Lets imagine instead that a 10% decline and 10% increase on any given day are equally likely, regardless of what happened the previous day.

Then there are four possible outcomes each with a 1/4 chance (without leverage):
Day 1 Day 2
UP, UP 1.1 1.21
UP, DOWN 1.1 0.99
DOWN, UP 0.9 0.99
DOWN, DOWN 0.9 0.81

Notice that the average of the four cases is equal to 1, which should be expected given how the process works. Also notice what "compounding" does: it takes the .01 away from the two middle examples and moves it toward the extremes, both high and low. 1.21 is .01 above what would happen from two .1 increases and .81 is also .01 above what would happen from two .1 declines.

Now lets look at the same example with 2x leverage:

Day 1 Day 2
UP, UP 1.2 1.44
UP, DOWN 1.2 0.96
DOWN, UP 0.8 0.96
DOWN, DOWN 0.8 0.64

Notice that the average of the four cases is still equal to 1! The same thing as above happens but in a more extreme way. This time 1.44 is .04 above and .64 is .04 above what they would be from two .2 increases or decreases. the .04s are taken away from the middle and moved to the extremes.

This demonstrates that the traditional examples of volatility decay are cherry picked and shouldn't be relied upon for concluding that leveraged ETFs are flawed.

finite_difference
Posts: 1542
Joined: Thu Jul 09, 2015 7:00 pm

Re: HEDGEFUNDIE's excellent adventure [3x leveraged ETF strategy]

Post by finite_difference » Wed Feb 06, 2019 10:43 pm

HEDGEFUNDIE wrote:
Wed Feb 06, 2019 6:15 pm
Here is a fun question for everyone.

Let's assume the strategy actually works as promised, and delivers ~25% CAGR through all market conditions, thick and thin.

By year 10, my $100k will have grown to $1M. Pretty sure I won't have a problem staying the course to that point.

By year 15, my account will have grown to $3M. Again, no problem staying the course.

By year 18, my account will be at $5.9M... Hmmm.... :|

By year 20, my account will be at $9.3M :shock:

You see where I am going with this. If someone was actually able to hold on for the entire 32 years of the backtest, he / she would be sitting on $100M right now.

At what point would you bail on the strategy, and why?

A big reason I started this thread is because I want the BH community to hold me accountable to my plan. So if I should change my goals, let's make those changes now!
So you are using $100k, a.k.a that is 15% of your portfolio?

So you have approx. $666k total.

If the levered portfolio will provide 23% CAGR, then the S&P 500 portfolio will provide 10% CAGR by the same estimate.

So if you are 100% stocks with $666k you’ll end up with $4.5 million after 20 years.

An interesting experiment but I think $4.5 million would be enough for me ;)

One reason I think your method could work is if you use a tax-advantaged account. Doing it using taxable seems painful. But still, I’m surprised no one offers this strategy as an ETF/fund if it supposedly works so well.
The most precious gift we can offer anyone is our attention. - Thich Nhat Hanh

ragnathor
Posts: 56
Joined: Wed Apr 22, 2015 11:12 am

Re: HEDGEFUNDIE's excellent adventure [3x leveraged ETF strategy]

Post by ragnathor » Wed Feb 06, 2019 10:51 pm

Can you post a link for the portfolio 1 backtest?

I think an important point is these leveraged funds only track the underlying index for a short period, and if they do what they're designed they will lose money in a flat market.

Say you start at 100 points, increase 10% to 110, then drop 10% to 99 points. Take a perfect 3x fund - increases 30% to 130, then drops 30% to 91 points.

These are just arbitrary numbers to illustrate the point, but the leveraged fund loses in a flat market.

Topic Author
HEDGEFUNDIE
Posts: 3973
Joined: Sun Oct 22, 2017 2:06 pm

Re: HEDGEFUNDIE's excellent adventure [3x leveraged ETF strategy]

Post by HEDGEFUNDIE » Wed Feb 06, 2019 11:02 pm

bgf wrote:
Wed Feb 06, 2019 9:51 pm
hohum wrote:
Wed Feb 06, 2019 8:50 pm
If you are in UPRO 100%, you will go down 90% in 2008. How many years of 24% returns does it take to recover from that?

You need an uncorrelated asset, and you need a lot of it.
and then go up a whole bunch in 2009 - 2019... 100% UPRO has a CAGR of almost 30% since inception.
Here is a comparison of 100% [simulated] UPRO (Portfolio 1) vs. my strategy of 40% UPRO + 60% TMF (Portfolio 2)

Image

The drawdowns really do kill you if you don't have an uncorrelated asset to balance. Max drawdown of 96%!

Day9
Posts: 825
Joined: Mon Jun 11, 2012 6:22 pm

Re: HEDGEFUNDIE's excellent adventure [3x leveraged ETF strategy]

Post by Day9 » Wed Feb 06, 2019 11:06 pm

hohum wrote:
Wed Feb 06, 2019 8:50 pm
If you are in UPRO 100%, you will go down 90% in 2008. How many years of 24% returns does it take to recover from that?

You need an uncorrelated asset, and you need a lot of it.
I did the math. 10.7 years. 1.24^10.7 = 10
I'm just a fan of the person I got my user name from

User avatar
Robert T
Posts: 2608
Joined: Tue Feb 27, 2007 9:40 pm
Location: 1, 0.2, 0.4, 0.5
Contact:

Re: HEDGEFUNDIE's excellent adventure [3x leveraged ETF strategy]

Post by Robert T » Thu Feb 07, 2019 2:12 am

.
A 3x leveraged S&P500 fund would have struggled to survive in 2008.

By my estimate a 3x daily leveraged S&P500 fund (assuming exact replication) would have declined by 95% from June 4, 2007 to March 9, 2008, and in real world implementation would have perhaps declined by even more (for example SSO a 2x daily leveraged S&P 500 fund declined by more than an exact 2x daily leverage S&P500 series).

Using the daily S&P500 data on the iShares website as far back as they have it (to inception of their S&P500 index) to May 16, 2000 – over eighteen years ago.

So if you were 33 in 2000, you would be 51 today, two years from 53 (and 2 years from the expected compounding of $100k into $10m re: the OP).

Here are the Feb 2019 values of 100k invested in each of the following strategies (assuming perfect replication):

$272k = S&P500
$375k = S&P500 x2 daily leverage
$260k = S&P500 x3 daily leverage

The S&P500 x3 daily leveraged series lagged the S&P500 (with no leverage) over this period (recovering from a 95% loss in 2008 takes a lot of return).

In addition, implementation would not be perfect replication if past experience is anything to go by – for example SSO (2x daily leveraged S&P500) has lagged its 2x daily leveraged S&P500 benchmark since inception: 10.4% annualized return for the fund, 13.2% for 2x daily leveraged S&P500.

So any additional return above the S&P500 over this period would have had to come from leveraged treasuries, and any rebalancing return. Rebalancing would have needed to be large and frequent (with high tax costs if held in a taxable account).

Will 3x leveraged 20+ yr treasuries always save the day? Yields on 20+ treasuries in May 2000 were about 6.25%, today they are 2.89%.

There will also not be perfect replication with leveraged treasuries. For example since inception, annualized returns on TMF = 7.2%, while 3x daily leveraged 20+ treasuries (the series its tries to replicate) = 10.7%. Just to note the annualized volatility on TMF = 38%.

I suggest treading carefully with such extreme volatility of the component parts (it tends to becomes a trading strategy i.e. amplifying extremes to increase the magnitude of needed rebalancing together with increasing the frequency of rebalancing in a effort to capture a larger rebalancing return - this also increases tail risk i.e. should both declining together)

At one point I did consider using 2x leveraged (intermediate) treasuries based on this earlier paper by Bridgewater https://www.bridgewater.com/resources/e ... -risks.pdf (to move from chart 1 to 2 on stock and bond side), but all things considered, did not move forward. Simulated overall portfolio results with 2x leveraged intermediate treasuries (e,g, TYD) could be fairly closely matched by simply extending duration with no leverage (e.g. EDV). Happy to stick with unleveraged intermediate treasuries (neither TYD or EDV).

Robert
.

Valuethinker
Posts: 39210
Joined: Fri May 11, 2007 11:07 am

Re: HEDGEFUNDIE's excellent adventure [3x leveraged ETF strategy]

Post by Valuethinker » Thu Feb 07, 2019 2:37 am

Robert T wrote:
Thu Feb 07, 2019 2:12 am
.
A 3x leveraged S&P500 fund would have struggled to survive in 2008.

By my estimate a 3x daily leveraged S&P500 fund (assuming exact replication) would have declined by 95% from June 4, 2007 to March 9, 2008, and in real world implementation would have perhaps declined by even more (for example SSO a 2x daily leveraged S&P 500 fund declined by more than an exact 2x daily leverage S&P500 series).

Using the daily S&P500 data on the iShares website as far back as they have it (to inception of their S&P500 index) to May 16, 2000 – over eighteen years ago.

So if you were 33 in 2000, you would be 51 today, two years from 53 (and 2 years from the expected compounding of $100k into $10m re: the OP).

Here are the Feb 2019 values of 100k invested in each of the following strategies (assuming perfect replication):

$272k = S&P500
$375k = S&P500 x2 daily leverage
$260k = S&P500 x3 daily leverage

The S&P500 x3 daily leveraged series lagged the S&P500 (with no leverage) over this period (recovering from a 95% loss in 2008 takes a lot of return).

In addition, implementation would not be perfect replication if past experience is anything to go by – for example SSO (2x daily leveraged S&P500) has lagged its 2x daily leveraged S&P500 benchmark since inception: 10.4% annualized return for the fund, 13.2% for 2x daily leveraged S&P500.

So any additional return above the S&P500 over this period would have had to come from leveraged treasuries, and any rebalancing return. Rebalancing would have needed to be large and frequent (with high tax costs if held in a taxable account).

Will 3x leveraged 20+ yr treasuries always save the day? Yields on 20+ treasuries in May 2000 were about 6.25%, today they are 2.89%.

There will also not be perfect replication with leveraged treasuries. For example since inception, annualized returns on TMF = 7.2%, while 3x daily leveraged 20+ treasuries (the series its tries to replicate) = 10.7%. Just to note the annualized volatility on TMF = 38%.

I suggest treading carefully with such extreme volatility of the component parts (it tends to becomes a trading strategy i.e. amplifying extremes to increase the magnitude of needed rebalancing together with increasing the frequency of rebalancing in a effort to capture a larger rebalancing return - this also increases tail risk i.e. should both declining together)

At one point I did consider using 2x leveraged (intermediate) treasuries based on this earlier paper by Bridgewater https://www.bridgewater.com/resources/e ... -risks.pdf (to move from chart 1 to 2 on stock and bond side), but all things considered, did not move forward. Simulated overall portfolio results with 2x leveraged intermediate treasuries (e,g, TYD) could be fairly closely matched by simply extending duration with no leverage (e.g. EDV). Happy to stick with unleveraged intermediate treasuries (neither TYD or EDV).

Robert
.
Thank you for this dose of reality.

It's fine for a corporate entity like a bank to play games with leverage - if they go bust, only the stockholders lose (and all the other stakeholders of course like employees, customers).

But for an individual the costs of personal bankruptcy, even in America, are much greater (Irish law was 12 years to discharge, people suicided after the Crash of 2008). A wipeout can wipe you out .. forever. End your investing career. I realize things are milder in the USA - you probably *can* get a mortgage again.

You are right re US Treasury bonds. The 30 year zero coupon (perhaps that is what EDV is?) has all the leverage most people will need in a portfolio.

samsdad
Posts: 751
Joined: Sat Jan 02, 2016 6:20 pm

Re: HEDGEFUNDIE's excellent adventure [3x leveraged ETF strategy]

Post by samsdad » Thu Feb 07, 2019 6:46 am

Perhaps my prior response got lost in the shuffle, but it’d be really great if HEDGEFUNDIE or EfficientInvestor would post the raw data they’re using or link to PV with the dataset.

I’m not an excel master so perhaps I’m wrong, but it appears that following the instructions that EfficientInvestor posted yesterday in response to my prior post would require some 30000 manual formula entries to come up with the data set.

In addition, as per my prior post, it be great if someone with more knowledge than me to share their thoughts on whether the formula in Fig. 3, which "multiplie(s) the base index daily % by a factor of 3 and then subtract(s) a small amount each day to account for the expense ratio of the fund. This is equal to the 1% expense ratio divided by 250 trading days in a year" reflects how these funds operate in real life? I'm trying to avoid the GIGO problem.

I think we’re all intrigued, and I certainly wish HEDGEFUNDIE success, but without the hard data and a discussion of the accuracy of the mechanics as noted above, it’s kind of hard to go much further than that.

bgf
Posts: 1170
Joined: Fri Nov 10, 2017 9:35 am

Re: HEDGEFUNDIE's excellent adventure [3x leveraged ETF strategy]

Post by bgf » Thu Feb 07, 2019 8:17 am

HEDGEFUNDIE wrote:
Wed Feb 06, 2019 11:02 pm
bgf wrote:
Wed Feb 06, 2019 9:51 pm
hohum wrote:
Wed Feb 06, 2019 8:50 pm
If you are in UPRO 100%, you will go down 90% in 2008. How many years of 24% returns does it take to recover from that?

You need an uncorrelated asset, and you need a lot of it.
and then go up a whole bunch in 2009 - 2019... 100% UPRO has a CAGR of almost 30% since inception.
Here is a comparison of 100% [simulated] UPRO (Portfolio 1) vs. my strategy of 40% UPRO + 60% TMF (Portfolio 2)

Image

The drawdowns really do kill you if you don't have an uncorrelated asset to balance. Max drawdown of 96%!
good post and the image is powerful. it'll be maybe less powerful in 10-15 years when 100% UPRO has again blown past your strategy. it really kind of encapsulates my point. you are using a large allocation to TMF to combat a problem that you have already resolved by leaving this strategy to only 15% of your portfolio... you want to limit your downside, and you've done that.

maybe we just have different mentalities, but, in my mind, when you swing for the fences you don't hold back.

if that 95% drawdown had NOT happened (and there's no no way to know whether it will or won't during your investing horizon), then 100% UPRO is likely to so far outpace your strategy that you'll reach your goal many years ahead of time. Once you've reached your goal you can take ALL the risk off the table... meaning that your funds are at risk for much less time.

so, you have have a highly risky allocation with TMF for 20 years or you can have an even more risky allocation of 100% UPRO for maybe 12-15 years. After that, you shut it down many years early. if the big drawdown comes AFTER you've already won, who cares?

Does that make sense or is there an error in that logic?
“TE OCCIDERE POSSUNT SED TE EDERE NON POSSUNT NEFAS EST"

foo.c
Posts: 62
Joined: Sat Jul 15, 2017 4:55 pm

Re: HEDGEFUNDIE's excellent adventure [3x leveraged ETF strategy]

Post by foo.c » Thu Feb 07, 2019 8:19 am

I did something similar last year. I "diversified" across leveraged Dow, S&P, QQQ, Treasuries, and long VIX as a hedge. I made a lot of return early, then the rules got me out of stocks ETFs and the long VIX bet really paid off. I did tax loss harvesting rolling from 3x to 2x to 1x ETFs to avoid wash sales.

I beat the market until I bet on 3x Oil ETF late last year, which was stupid because it blew up my well researched (by me) method. I still ended up very slightly positive (0.65%) for the year which beat the market.

Right now I'm waiting to see what kind of tax hell I have put myself in with all the trading and K1s before determining if I will resume.

bgf
Posts: 1170
Joined: Fri Nov 10, 2017 9:35 am

Re: HEDGEFUNDIE's excellent adventure [3x leveraged ETF strategy]

Post by bgf » Thu Feb 07, 2019 8:20 am

Valuethinker wrote:
Thu Feb 07, 2019 2:37 am
Robert T wrote:
Thu Feb 07, 2019 2:12 am
.
A 3x leveraged S&P500 fund would have struggled to survive in 2008.

By my estimate a 3x daily leveraged S&P500 fund (assuming exact replication) would have declined by 95% from June 4, 2007 to March 9, 2008, and in real world implementation would have perhaps declined by even more (for example SSO a 2x daily leveraged S&P 500 fund declined by more than an exact 2x daily leverage S&P500 series).

Using the daily S&P500 data on the iShares website as far back as they have it (to inception of their S&P500 index) to May 16, 2000 – over eighteen years ago.

So if you were 33 in 2000, you would be 51 today, two years from 53 (and 2 years from the expected compounding of $100k into $10m re: the OP).

Here are the Feb 2019 values of 100k invested in each of the following strategies (assuming perfect replication):

$272k = S&P500
$375k = S&P500 x2 daily leverage
$260k = S&P500 x3 daily leverage

The S&P500 x3 daily leveraged series lagged the S&P500 (with no leverage) over this period (recovering from a 95% loss in 2008 takes a lot of return).

In addition, implementation would not be perfect replication if past experience is anything to go by – for example SSO (2x daily leveraged S&P500) has lagged its 2x daily leveraged S&P500 benchmark since inception: 10.4% annualized return for the fund, 13.2% for 2x daily leveraged S&P500.

So any additional return above the S&P500 over this period would have had to come from leveraged treasuries, and any rebalancing return. Rebalancing would have needed to be large and frequent (with high tax costs if held in a taxable account).

Will 3x leveraged 20+ yr treasuries always save the day? Yields on 20+ treasuries in May 2000 were about 6.25%, today they are 2.89%.

There will also not be perfect replication with leveraged treasuries. For example since inception, annualized returns on TMF = 7.2%, while 3x daily leveraged 20+ treasuries (the series its tries to replicate) = 10.7%. Just to note the annualized volatility on TMF = 38%.

I suggest treading carefully with such extreme volatility of the component parts (it tends to becomes a trading strategy i.e. amplifying extremes to increase the magnitude of needed rebalancing together with increasing the frequency of rebalancing in a effort to capture a larger rebalancing return - this also increases tail risk i.e. should both declining together)

At one point I did consider using 2x leveraged (intermediate) treasuries based on this earlier paper by Bridgewater https://www.bridgewater.com/resources/e ... -risks.pdf (to move from chart 1 to 2 on stock and bond side), but all things considered, did not move forward. Simulated overall portfolio results with 2x leveraged intermediate treasuries (e,g, TYD) could be fairly closely matched by simply extending duration with no leverage (e.g. EDV). Happy to stick with unleveraged intermediate treasuries (neither TYD or EDV).

Robert
.
Thank you for this dose of reality.

It's fine for a corporate entity like a bank to play games with leverage - if they go bust, only the stockholders lose (and all the other stakeholders of course like employees, customers).

But for an individual the costs of personal bankruptcy, even in America, are much greater (Irish law was 12 years to discharge, people suicided after the Crash of 2008). A wipeout can wipe you out .. forever. End your investing career. I realize things are milder in the USA - you probably *can* get a mortgage again.

You are right re US Treasury bonds. The 30 year zero coupon (perhaps that is what EDV is?) has all the leverage most people will need in a portfolio.
OP is only doing this with 15% of his portfolio. it could go to zero and he's still left with 85% of his portfolio... bankruptcy is a bit much in this scenario.
“TE OCCIDERE POSSUNT SED TE EDERE NON POSSUNT NEFAS EST"

Elysium
Posts: 1887
Joined: Mon Apr 02, 2007 6:22 pm

Re: HEDGEFUNDIE's excellent adventure [3x leveraged ETF strategy]

Post by Elysium » Thu Feb 07, 2019 9:05 am

Someone had to ask this question, so I will. If this strategy is so good as promised, why hasn't anyone else tried it to become multi-millionaires using it. Where are the professional money managers using such a strategy to beat the S&P 500?

User avatar
vineviz
Posts: 5391
Joined: Tue May 15, 2018 1:55 pm

Re: HEDGEFUNDIE's excellent adventure [3x leveraged ETF strategy]

Post by vineviz » Thu Feb 07, 2019 9:09 am

random_person wrote:
Wed Feb 06, 2019 10:39 pm
1. "Volatility decay" is not a thing
"Volatility decay" refers to the circumstances in which the returns of a daily leveraged fund differ materially over some period of time from the returns of the underlying index multiplied by the leverage factor.

For instance years like 2011 (when the S&P 500 gained 1.86% but ProShares UltraPro S&P500 (UPRO) lost 11.88%) or 2015 (when the S&P 500 gained 1.30% but ProShares UltraPro S&P500 (UPRO) lost 5.24%) or 2018 (when the S&P 500 lost 4.47% but ProShares UltraPro S&P500 (UPRO) lost 25.15%).

Using simulated data, the poster child is 1987 when the S&P 500 gained 3.91% but a 3x daily leveraged ETF would have lost 31.24%.

Gains of over 90% in both 1985 and 1989 would have smoothed the pain for an investor who could stay the course (and was managing their exposure appropriately) but a naive investor checking their statement at the end of the year in 1987 might be baffled that the fund was down 31% instead of up 12% unless they understood that "volatility decay" is "a thing".

It's definitely a minority of months in which "volatility decay" tends to manifest but it's not terribly uncommon.

Since 1970, a 3x daily leveraged ETF (using simulated data) would have produced a smaller gain than the S&P 500 in eight years, and in an additional four years would have experienced a loss more than 4x the loss in the S&P 500. By my count, that's about one year in four where "volatility decay" is definitely "a thing".

That doesn't obviate the value of these products, but anyone using them needs to be aware (as the OP definitely is) that it is wise to employ risk mitigation strategies when using them long-term.
"Far more money has been lost by investors preparing for corrections than has been lost in corrections themselves." ~~ Peter Lynch

EfficientInvestor
Posts: 260
Joined: Thu Nov 01, 2018 7:02 pm
Location: Alabama

Re: HEDGEFUNDIE's excellent adventure [3x leveraged ETF strategy]

Post by EfficientInvestor » Thu Feb 07, 2019 9:20 am

samsdad wrote:
Thu Feb 07, 2019 6:46 am
Perhaps my prior response got lost in the shuffle, but it’d be really great if HEDGEFUNDIE or EfficientInvestor would post the raw data they’re using or link to PV with the dataset.

I’m not an excel master so perhaps I’m wrong, but it appears that following the instructions that EfficientInvestor posted yesterday in response to my prior post would require some 30000 manual formula entries to come up with the data set.

In addition, as per my prior post, it be great if someone with more knowledge than me to share their thoughts on whether the formula in Fig. 3, which "multiplie(s) the base index daily % by a factor of 3 and then subtract(s) a small amount each day to account for the expense ratio of the fund. This is equal to the 1% expense ratio divided by 250 trading days in a year" reflects how these funds operate in real life? I'm trying to avoid the GIGO problem.

I think we’re all intrigued, and I certainly wish HEDGEFUNDIE success, but without the hard data and a discussion of the accuracy of the mechanics as noted above, it’s kind of hard to go much further than that.
UPDATE (2/20/19) - The data that had previously been provided in this post was based on developing a best fit to the data of actual leveraged ETFs since their inception. Since inception of 3X ETFs in 2009, borrowing rates have been low. Therefore, this data did not take into account higher borrowing rates that occurred prior to 2009. Due to this, I have removed the links to the data I had previously provided. More recent posts on this thread provide an update on how to better simulate returns of these funds prior to inception. I will also update the backtesting instructions on my website at the link below in the near future in order to reflect the updated simulation strategy.

Sorry for the delayed response. If you were to follow the instructions on the website (link to article at the bottom of post), it only takes about 5 minutes per spreadsheet to set up. You need to type in the formulas on the first row, but then you can copy the formulas down to all rows automatically. If you don't have access to Excel, you can create a free Google account and use Google Sheets. It would be my strong preference that you create your own data and not rely too heavily on the data of others. That being said, I will provide the data for UPRO and TMF that I previously sent to HEDGEFUNDIE. Download the .csv files below and then upload them to PV using the instructions in the article.

UPRO - Link Removed

The UPRO data is based on daily returns of VFINX and includes a 1% per year expense ratio (I subtract .01/250 for each day). I initially used ^GSPC as the data source, but forgot that I had switched to using VFINX because it accounts for the dividend. If you only use data from ^GSPC, your proxy will underperform UPRO. For each day, I did a weighted average of the close value (20%) and adjusted close value (80%). I determined the weighting using an iterative process until I found something that closely matched the actual UPRO performance. For those that want to recreate this data, follow the instructions at the article link below that refer to developing the TMF proxy.

TMF - Link Removed

The TMF data in this file is a hybrid of TLT and VUSTX. I used daily data from TLT for as long as it was available, because it follows the same index as TMF. Prior to inception of TLT, I used data from VUSTX. If you want to see what a comparison of the proxy TMF and actual TMF would have looked like if I had only used VUSTX data (in order to validate the use of VUSTX), see Figure 10 in the article below.

https://theleveragedindexer.com/2018/11 ... -backtest/
Last edited by EfficientInvestor on Wed Feb 20, 2019 11:36 am, edited 1 time in total.

EfficientInvestor
Posts: 260
Joined: Thu Nov 01, 2018 7:02 pm
Location: Alabama

Re: HEDGEFUNDIE's excellent adventure [3x leveraged ETF strategy]

Post by EfficientInvestor » Thu Feb 07, 2019 9:34 am

Elysium wrote:
Thu Feb 07, 2019 9:05 am
Someone had to ask this question, so I will. If this strategy is so good as promised, why hasn't anyone else tried it to become multi-millionaires using it. Where are the professional money managers using such a strategy to beat the S&P 500?
I think the general public has not considered the use of this strategy with leveraged ETFs because anything you read about them says not to hold them long term. After you learn that UPRO would have had a drawdown of 95% during 2008, you don't want to come anywhere near an investment like that. However, if you consider the use of the leveraged ETFs as part of a diversified portfolio, the picture changes completely. By applying leverage to a diversified portfolio, you can achieve higher returns with less drawdown because you are taking advantage of the much more efficient use of capital. Consider the 2X backtest below starting in 2006 using S&P 500 and 7-10 year treasuries. Despite SSO having a -81.3% drawdown during the 2008 crash, the 2X 40/60 portfolio as a whole would have drastically outperformed the S&P 500 over the time period with less drawdown than an unleveraged 60/40 portfolio.

Jul 2006 - Jan 2019
2X 40/60 - CAGR = 11.8%, Max DD = -21.7%
40/60 - CAGR = 7.7%, Max DD = -26.8%
S&P 500 - CAGR = 8.3%, Max DD = -51.0%

https://www.portfoliovisualizer.com/bac ... tion4_2=40

User avatar
hdas
Posts: 1323
Joined: Thu Jun 11, 2015 8:24 am

Re: HEDGEFUNDIE's excellent adventure [3x leveraged ETF strategy]

Post by hdas » Thu Feb 07, 2019 9:34 am

"whenever there is a randomized way of doing something, then there is a nonrandomized way that delivers better performance but requires more thought" ET Jaynes

staythecourse
Posts: 6993
Joined: Mon Jan 03, 2011 9:40 am

Re: HEDGEFUNDIE's excellent adventure [3x leveraged ETF strategy]

Post by staythecourse » Thu Feb 07, 2019 9:35 am

foo.c wrote:
Thu Feb 07, 2019 8:19 am
I did something similar last year. I "diversified" across leveraged Dow, S&P, QQQ, Treasuries, and long VIX as a hedge. I made a lot of return early, then the rules got me out of stocks ETFs and the long VIX bet really paid off. I did tax loss harvesting rolling from 3x to 2x to 1x ETFs to avoid wash sales.

I beat the market until I bet on 3x Oil ETF late last year, which was stupid because it blew up my well researched (by me) method. I still ended up very slightly positive (0.65%) for the year which beat the market.

Right now I'm waiting to see what kind of tax hell I have put myself in with all the trading and K1s before determining if I will resume.
Is your time horizon 1 year? If not then why would you care if you were up 30% or down 30% in any given year? Returns should ONLY be considered at the appropriate time horizon that the investor needs said money. This simple aspect of investing is what gives most investors the biggest frame of reference risk (in my opinion).

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

staythecourse
Posts: 6993
Joined: Mon Jan 03, 2011 9:40 am

Re: HEDGEFUNDIE's excellent adventure [3x leveraged ETF strategy]

Post by staythecourse » Thu Feb 07, 2019 9:35 am

foo.c wrote:
Thu Feb 07, 2019 8:19 am
I did something similar last year. I "diversified" across leveraged Dow, S&P, QQQ, Treasuries, and long VIX as a hedge. I made a lot of return early, then the rules got me out of stocks ETFs and the long VIX bet really paid off. I did tax loss harvesting rolling from 3x to 2x to 1x ETFs to avoid wash sales.

I beat the market until I bet on 3x Oil ETF late last year, which was stupid because it blew up my well researched (by me) method. I still ended up very slightly positive (0.65%) for the year which beat the market.

Right now I'm waiting to see what kind of tax hell I have put myself in with all the trading and K1s before determining if I will resume.
Is your time horizon 1 year? If not then why would you care if you were up 30% or down 30% in any given year? Returns should ONLY be considered at the appropriate time horizon that the investor needs said money. This simple aspect of investing is what gives most investors the biggest frame of reference risk (in my opinion).

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

EfficientInvestor
Posts: 260
Joined: Thu Nov 01, 2018 7:02 pm
Location: Alabama

Re: HEDGEFUNDIE's excellent adventure [3x leveraged ETF strategy]

Post by EfficientInvestor » Thu Feb 07, 2019 9:36 am

hdas wrote:
Thu Feb 07, 2019 9:34 am
Are you Levi? :greedy
Yes, I am Levi.

User avatar
hdas
Posts: 1323
Joined: Thu Jun 11, 2015 8:24 am

Re: HEDGEFUNDIE's excellent adventure [3x leveraged ETF strategy]

Post by hdas » Thu Feb 07, 2019 9:38 am

vineviz wrote:
Thu Feb 07, 2019 9:09 am
it is wise to employ risk mitigation strategies when using them long-term.
Can you mention examples of these risk mitigation strategies that when employed systematically and properly backtested don't diminish the profits to the extent of rendering the whole enterprise unprofitable? :greedy
"whenever there is a randomized way of doing something, then there is a nonrandomized way that delivers better performance but requires more thought" ET Jaynes

User avatar
hdas
Posts: 1323
Joined: Thu Jun 11, 2015 8:24 am

Re: HEDGEFUNDIE's excellent adventure [3x leveraged ETF strategy]

Post by hdas » Thu Feb 07, 2019 9:44 am

EfficientInvestor wrote:
Thu Feb 07, 2019 9:36 am

Yes, I am Levi.
You would do your audience a favor if your presentation included more robust backtesting techniques. Use simulations of the potential paths. Given the volatilities and the volatility of the correlation. Do present success in a probabilistic fashion. Since you are an engineer, this could be trivial. This in sample presentation is very naive and misleading. :greedy
"whenever there is a randomized way of doing something, then there is a nonrandomized way that delivers better performance but requires more thought" ET Jaynes

Sola Scriptura
Posts: 19
Joined: Tue Feb 13, 2018 5:23 pm

Re: HEDGEFUNDIE's excellent adventure [3x leveraged ETF strategy]

Post by Sola Scriptura » Thu Feb 07, 2019 10:03 am

HEDGEFUNDIE,

What I still don't understand, given your data and given everything that's been said, is that if this strategy is only 15% of your total portfolio, how do you not already have enough downside protection built in, thus enabling you to go 100% UPRO regardless of the potential for a 95% drawdown in a 2008-like scenario? In a single-day 33.33% drop, you'd be screwed anyway, so why not let 100% UPRO ride and take your chances on hitting that $10MM goal much sooner?

foo.c
Posts: 62
Joined: Sat Jul 15, 2017 4:55 pm

Re: HEDGEFUNDIE's excellent adventure [3x leveraged ETF strategy]

Post by foo.c » Thu Feb 07, 2019 10:07 am

staythecourse wrote:
Thu Feb 07, 2019 9:35 am
foo.c wrote:
Thu Feb 07, 2019 8:19 am
I did something similar last year. I "diversified" across leveraged Dow, S&P, QQQ, Treasuries, and long VIX as a hedge. I made a lot of return early, then the rules got me out of stocks ETFs and the long VIX bet really paid off. I did tax loss harvesting rolling from 3x to 2x to 1x ETFs to avoid wash sales.

I beat the market until I bet on 3x Oil ETF late last year, which was stupid because it blew up my well researched (by me) method. I still ended up very slightly positive (0.65%) for the year which beat the market.

Right now I'm waiting to see what kind of tax hell I have put myself in with all the trading and K1s before determining if I will resume.
Is your time horizon 1 year? If not then why would you care if you were up 30% or down 30% in any given year? Returns should ONLY be considered at the appropriate time horizon that the investor needs said money. This simple aspect of investing is what gives most investors the biggest frame of reference risk (in my opinion).

Good luck.
I don't really consider it investing, to be honest. If anything it was more like a second job. If/when I resume I will do more to automate it.

Topic Author
HEDGEFUNDIE
Posts: 3973
Joined: Sun Oct 22, 2017 2:06 pm

Re: HEDGEFUNDIE's excellent adventure [3x leveraged ETF strategy]

Post by HEDGEFUNDIE » Thu Feb 07, 2019 10:14 am

Elysium wrote:
Thu Feb 07, 2019 9:05 am
Someone had to ask this question, so I will. If this strategy is so good as promised, why hasn't anyone else tried it to become multi-millionaires using it. Where are the professional money managers using such a strategy to beat the S&P 500?
See Bridgewater - All Weather Fund. One of the most famous funds of all time.

I don’t have any inflation protection assets but the leveraged risk parity principle is otherwise the same.

Elysium
Posts: 1887
Joined: Mon Apr 02, 2007 6:22 pm

Re: HEDGEFUNDIE's excellent adventure [3x leveraged ETF strategy]

Post by Elysium » Thu Feb 07, 2019 10:17 am

EfficientInvestor wrote:
Thu Feb 07, 2019 9:34 am
Elysium wrote:
Thu Feb 07, 2019 9:05 am
Someone had to ask this question, so I will. If this strategy is so good as promised, why hasn't anyone else tried it to become multi-millionaires using it. Where are the professional money managers using such a strategy to beat the S&P 500?
I think the general public has not considered the use of this strategy with leveraged ETFs because anything you read about them says not to hold them long term. After you learn that UPRO would have had a drawdown of 95% during 2008, you don't want to come anywhere near an investment like that. However, if you consider the use of the leveraged ETFs as part of a diversified portfolio, the picture changes completely. By applying leverage to a diversified portfolio, you can achieve higher returns with less drawdown because you are taking advantage of the much more efficient use of capital. Consider the 2X backtest below starting in 2006 using S&P 500 and 7-10 year treasuries. Despite SSO having a -81.3% drawdown during the 2008 crash, the 2X 40/60 portfolio as a whole would have drastically outperformed the S&P 500 over the time period with less drawdown than an unleveraged 60/40 portfolio.

Jul 2006 - Jan 2019
2X 40/60 - CAGR = 11.8%, Max DD = -21.7%
40/60 - CAGR = 7.7%, Max DD = -26.8%
S&P 500 - CAGR = 8.3%, Max DD = -51.0%

https://www.portfoliovisualizer.com/bac ... tion4_2=40
That still doesn't address the question, anyone can do that on a diversified portfolio if it were that simple. It sounds like a simple strategy someone can openly post on the internet with a few backtests and implement with a few low costs leveraged ETFs, yet it promise to generate 10x more than unleveraged S&P 500.

EfficientInvestor
Posts: 260
Joined: Thu Nov 01, 2018 7:02 pm
Location: Alabama

Re: HEDGEFUNDIE's excellent adventure [3x leveraged ETF strategy]

Post by EfficientInvestor » Thu Feb 07, 2019 10:21 am

hdas wrote:
Thu Feb 07, 2019 9:44 am
EfficientInvestor wrote:
Thu Feb 07, 2019 9:36 am

Yes, I am Levi.
You would do your audience a favor if your presentation included more robust backtesting techniques. Use simulations of the potential paths. Given the volatilities and the volatility of the correlation. Do present success in a probabilistic fashion. Since you are an engineer, this could be trivial. This in sample presentation is very naive and misleading. :greedy
Thanks for the feedback. I hope to eventually find the time to do more research along the lines of what you suggest. In the meantime, I hope that the information in the articles on the site will at least help people start thinking differently about the use of leverage. The intent of the site is not to prove beyond the shadow of a doubt (e.g. via simulations of potential paths) that this concept will always work perfectly. The intent is to explain the principles behind leveraged indexing and let readers decide if it makes sense at a more macro level. As for the backtesting article that I provided the link to...the main intent of the article is to teach readers how to find data and perform their own backtest in order to prove to themselves that holding leveraged ETFs long term is not a horrible idea like most people say it is, as long as it is done in a diversified manner. I believe the article accomplishes that goal.

Topic Author
HEDGEFUNDIE
Posts: 3973
Joined: Sun Oct 22, 2017 2:06 pm

Re: HEDGEFUNDIE's excellent adventure [3x leveraged ETF strategy]

Post by HEDGEFUNDIE » Thu Feb 07, 2019 10:23 am

bgf wrote:
Thu Feb 07, 2019 8:17 am
HEDGEFUNDIE wrote:
Wed Feb 06, 2019 11:02 pm
bgf wrote:
Wed Feb 06, 2019 9:51 pm
hohum wrote:
Wed Feb 06, 2019 8:50 pm
If you are in UPRO 100%, you will go down 90% in 2008. How many years of 24% returns does it take to recover from that?

You need an uncorrelated asset, and you need a lot of it.
and then go up a whole bunch in 2009 - 2019... 100% UPRO has a CAGR of almost 30% since inception.
Here is a comparison of 100% [simulated] UPRO (Portfolio 1) vs. my strategy of 40% UPRO + 60% TMF (Portfolio 2)

Image

The drawdowns really do kill you if you don't have an uncorrelated asset to balance. Max drawdown of 96%!
good post and the image is powerful. it'll be maybe less powerful in 10-15 years when 100% UPRO has again blown past your strategy. it really kind of encapsulates my point. you are using a large allocation to TMF to combat a problem that you have already resolved by leaving this strategy to only 15% of your portfolio... you want to limit your downside, and you've done that.

maybe we just have different mentalities, but, in my mind, when you swing for the fences you don't hold back.

if that 95% drawdown had NOT happened (and there's no no way to know whether it will or won't during your investing horizon), then 100% UPRO is likely to so far outpace your strategy that you'll reach your goal many years ahead of time. Once you've reached your goal you can take ALL the risk off the table... meaning that your funds are at risk for much less time.

so, you have have a highly risky allocation with TMF for 20 years or you can have an even more risky allocation of 100% UPRO for maybe 12-15 years. After that, you shut it down many years early. if the big drawdown comes AFTER you've already won, who cares?

Does that make sense or is there an error in that logic?
This “powerful” image I posted shows UPRO took 17 years to recover from the max drawdown (2000-2017). I don’t have that kind of time to recover if the crash happens tomorrow.

The 85% of my portfolio is invested according to a normal glide path, so it’s also mostly stocks and offers no protection for the 15%.

Jags4186
Posts: 4073
Joined: Wed Jun 18, 2014 7:12 pm

Re: HEDGEFUNDIE's excellent adventure [3x leveraged ETF strategy]

Post by Jags4186 » Thu Feb 07, 2019 10:51 am

Sola Scriptura wrote:
Thu Feb 07, 2019 10:03 am
HEDGEFUNDIE,

What I still don't understand, given your data and given everything that's been said, is that if this strategy is only 15% of your total portfolio, how do you not already have enough downside protection built in, thus enabling you to go 100% UPRO regardless of the potential for a 95% drawdown in a 2008-like scenario? In a single-day 33.33% drop, you'd be screwed anyway, so why not let 100% UPRO ride and take your chances on hitting that $10MM goal much sooner?
In a singlthe day 33.33% drop he wouldn’t be screwed. He’s be out 40% of this section of his portfolio, long term treasuries would likely rally, and he could rebalance back into a leveraged ETF.

EfficientInvestor
Posts: 260
Joined: Thu Nov 01, 2018 7:02 pm
Location: Alabama

Re: HEDGEFUNDIE's excellent adventure [3x leveraged ETF strategy]

Post by EfficientInvestor » Thu Feb 07, 2019 10:52 am

Elysium wrote:
Thu Feb 07, 2019 10:17 am
EfficientInvestor wrote:
Thu Feb 07, 2019 9:34 am
Elysium wrote:
Thu Feb 07, 2019 9:05 am
Someone had to ask this question, so I will. If this strategy is so good as promised, why hasn't anyone else tried it to become multi-millionaires using it. Where are the professional money managers using such a strategy to beat the S&P 500?
I think the general public has not considered the use of this strategy with leveraged ETFs because anything you read about them says not to hold them long term. After you learn that UPRO would have had a drawdown of 95% during 2008, you don't want to come anywhere near an investment like that. However, if you consider the use of the leveraged ETFs as part of a diversified portfolio, the picture changes completely. By applying leverage to a diversified portfolio, you can achieve higher returns with less drawdown because you are taking advantage of the much more efficient use of capital. Consider the 2X backtest below starting in 2006 using S&P 500 and 7-10 year treasuries. Despite SSO having a -81.3% drawdown during the 2008 crash, the 2X 40/60 portfolio as a whole would have drastically outperformed the S&P 500 over the time period with less drawdown than an unleveraged 60/40 portfolio.

Jul 2006 - Jan 2019
2X 40/60 - CAGR = 11.8%, Max DD = -21.7%
40/60 - CAGR = 7.7%, Max DD = -26.8%
S&P 500 - CAGR = 8.3%, Max DD = -51.0%

https://www.portfoliovisualizer.com/bac ... tion4_2=40
That still doesn't address the question, anyone can do that on a diversified portfolio if it were that simple. It sounds like a simple strategy someone can openly post on the internet with a few backtests and implement with a few low costs leveraged ETFs, yet it promise to generate 10x more than unleveraged S&P 500.
I'm not sure what kind of a response you are looking for. I'm sure there are hedge funds that use similar strategies, but they would lose their competitive edge and their clients if the general public knew they could execute a similar strategy in a passive manner using leveraged ETFs. As HEDGEFUNDIE referenced, this is all very closely related to some of the white papers that Bridgewater has published regarding risk parity (link below). It is just taking it a step further and applying leverage to the entire portfolio in order to amplify return (and risk) to a point that meets your risk tolerance.

https://www.bridgewater.com/research-li ... sk-parity/

Sola Scriptura
Posts: 19
Joined: Tue Feb 13, 2018 5:23 pm

Re: HEDGEFUNDIE's excellent adventure [3x leveraged ETF strategy]

Post by Sola Scriptura » Thu Feb 07, 2019 11:32 am

Jags4186 wrote:
Thu Feb 07, 2019 10:51 am
Sola Scriptura wrote:
Thu Feb 07, 2019 10:03 am
HEDGEFUNDIE,

What I still don't understand, given your data and given everything that's been said, is that if this strategy is only 15% of your total portfolio, how do you not already have enough downside protection built in, thus enabling you to go 100% UPRO regardless of the potential for a 95% drawdown in a 2008-like scenario? In a single-day 33.33% drop, you'd be screwed anyway, so why not let 100% UPRO ride and take your chances on hitting that $10MM goal much sooner?
In a singlthe day 33.33% drop he wouldn’t be screwed. He’s be out 40% of this section of his portfolio, long term treasuries would likely rally, and he could rebalance back into a leveraged ETF.
You're right; my mistake. I overstated my case and didn't provide enough detail. What I meant is that being wiped out completely on his 40% would take a while to recover from if current low (relative to the past) treasury yields persist, even with a presumed rally (which could actually end up being more modest than you suppose). HEDGEFUNDIE has stated elsewhere that he doesn't have a long period to recover, so how much time are you assuming would be available for him to slowly move back into a leveraged ETF position?

random_person
Posts: 8
Joined: Wed Feb 06, 2019 10:11 pm

Re: HEDGEFUNDIE's excellent adventure [3x leveraged ETF strategy]

Post by random_person » Thu Feb 07, 2019 11:42 am

vineviz wrote:
Thu Feb 07, 2019 9:09 am
random_person wrote:
Wed Feb 06, 2019 10:39 pm
1. "Volatility decay" is not a thing
"Volatility decay" refers to the circumstances in which the returns of a daily leveraged fund differ materially over some period of time from the returns of the underlying index multiplied by the leverage factor.

This isn't true for a couple reasons. But first you should go back and read my original post that you replied to. The point was that for any example of under performance that happens during periods of mean reversion, you can find out performance that happens either during periods of consistent gains or consistent losses.

You should backtest 2017 and 2008. In both of these years, I would bet the leveraged ETF out performs 3x or 2x the benchmark. You just have to find years where the underlying goes consistently up or consistently down. The whole point of my post was that finding a cherry picked example doesn't demonstrate a principle.

Volatility decay is not just the idea that the ETF can under perform the benchmark. The ETF rebalances daily so it WILL differ from the benchmark. When the market strongly trends in one direction for awhile, it will tend to out perform and when it mean reverts, then it will tend to under perform.

Volatility decay is the idea that variance itself causes the ETF to under perform, and it doesn't. The variance just causes the center of the distribution to be gutted, while pushing more positive results to the extremes.

Locked