leveraged risk parity portfolios

[mathjak107]

after reading a lot on the work of corey hoffstein i decided to try what is called a leveraged 60/40 risk parity portfolio.

the one i am experimenting with is called the carolina reaper , after the worlds hottest pepper.

it works by using leveraged etfs .

so if we have a dollar , we put

.20 cents in upro 3x equity

.13 cents in tyd 3x bonds .

that now has the weight of a 60/40 . yet we still have .67 cents left to use to hedge

we then put .67 cents in the managed futures etf dbmf to hedge with short and long positions.


just went as far back with these funds as i could in portfolio visualizer. first time running it myself .

i was able to go back to jan 2020 , so almost 4 years , including the covid sell off

a 60/40 consisting of 60% voo (s&p fund )and 40% bnd ( total bond fund ) took 100k and turned it in to 126,130 , a cagr of 6.11%

the 60/40 carolina reaper took a 100k grew it to 136,260 ,a cagr of 8.22%

BUT GET THIS

WORST DOWN YEAR WAS MINUS 2.57% for the reaper with a worst draw down of 9.45%

the conventional 60/40 had a worst year of minus 16.16 % and a worst draw down of 20.15%

HOLY CRAP , WHAT A DIFFERENCE .

Boy this is showing a lot of promise as we went thru some awful times as well as high inflation and rising rates.

it really wants to compel one to go whole hog but i don’t have the balls yet. lol

but it certainly looked like that magical portfolio we all dream about if we don’t want 100% equities

i mean down a mere 2.57% …that’s incredible while a 60/40 was down 16.16% , yet beat it by over 2% in gains

keep in mind these returns include all expenses and these leveraged funds are not anywhere near as cheap as voo or bnd .

so like i always say , there is so much more when portfolios are involved then lowest costs of the individual components

that was vanguard marketing brainwashing investors.


expenses

dbmf .85% , upro .92 tyd 1.1%

voo .03. bnd .03

yet including all expenses the reaper not only averaged more then 2% a year more but was lower risk in every respect

yet how often here is the advice to buy low cost index funds as the primary criteria or only criteria … so marketing vs real life can be very different.

any way i threw 50k in this as a test which is enough to keep it interesting but not enough to do real damage .

i want to test drive this before making it a core holding and committing serious dollars .

so far it has been pretty docile even on days where these funds individually swung 5% . lots of yin and yang going on with the managed futures


here is some info on this model

https://www.riskparitychronicles.com/testing-three-variations-of-a-leveraged-60-40/

 

[pb4uski]

Interesting. At first it looked similar to the HFEA portfolio but the addition of DBMF makes a huge difference to the positive. Thanks for sharing.

You may want to change your PV Time Period (top box) from Year-To-Year to Month-To-Month... it adds another 5 months.

Also, I thought a combination of 20% VOO, 13% BND and 67% DBMF which removes the 3x leverging was interesting... lower return but much low volatility and worst year was flat (+0.01%)

 

[donheff]

I have no idea how a strategy like this would react if "everyone" was doing it. With index funds it doesn't seem to matter but what about these leveraged funds, do they still work if investors pile in?

 

[38Chevy454]

I can see some advantages, I think the key is how it deals with long term downward market. Is the active mgmt able to compensate for the increased losses of the 3x leveraged accounts?

 

[mathjak107]

Interesting. At first it looked similar to the HFEA portfolio but the addition of DBMF makes a huge difference to the positive. Thanks for sharing.

You may want to change your PV Time Period (top box) from Year-To-Year to Month-To-Month... it adds another 5 months.

Also, I thought a combination of 20% VOO, 13% BND and 67% DBMF which removes the 3x leverging was interesting... lower return but much low volatility and worst year was flat (+0.01%)

yes , it seems to be well thought out the way the parts mesh as a leveraged portfolio. the idea is to mimic a 60/40 with lower risk

 

[mathjak107]

I can see some advantages, I think the key is how it deals with long term downward market. Is the active mgmt able to compensate for the increased losses of the 3x leveraged accounts?

it did well over the covid down turn but there is little history on these leveraged etfs .

remember we can have strange things happen with these etfs in flash crashes and a 34% drop in a day means zero value

now before you say odds are slim of that happening .


it happened in 2010 and 2015 when we had flash crashes.

popular etfs got disjointed from their actual holdings and popular etfs like dvy plunged 36% in minutes while the actual holdings were down a mere 5% .

iShares Select Dividend ETF (DVY) dived 36%, Guggenheim S&P 500 Equal Weight (RSP) tumbled 42% and iShares Conservative Allocation Fund (AOK) dropped 50%


The 2015 flash crash wasn’t unprecedented: In 2010, another flash crash dropped some ETF prices as much as 60% in a span of just 20 minutes, according to the Securities and Exchange Commission. For a few horrific seconds, shares of the iShares Russell 1000 Growth Index ETF (IWF) traded for a penny. Vanguard Total Stock Market ETF (VTI) sold for 13 cents.



it corrected itself but not before trades actually went thru at those levels .

it shows us how imperfect etfs can be compared to mutual funds.

we have no idea what here 3x funds will do yet in that .

if the reaper works out well i can see doubling it from 50 to 100k but not the 7 figures i have in my conventional models

https://www.investmentnews.com/indus...sh-crash-75961

 

[mathjak107]

so for this past week , the reaper up 4.87% ytd .

conventional 60/40 up 2.45. vbiax


my fidelity insight growth and income model up 2.7

so reaper doing much better

 

[GaryInCO]

A couple of years ago I developed a simple set of two trading strategies, including a leveraged risk-parity basket. I didn't know about DBMF so I did a standard basket of leveraged stocks/bonds/gold: UPRO, TMF, UGL. I picked two simple strategies that only required attention once a month. (Usually it was just "Nope, nothing to do, come back in a month.")

I generated long histories of synthetic leveraged funds (based on VFINX, VUSTX, and spot gold prices) so I could backtest the strategies back to 1986. In that 35-year history, my two strategies generated 30% CAGR with 1.38 Sharpe. Worst drawdown was about 30%, in 2008 and in 1987, and a couple of 20% DDs. That is incredible performance. I had been watching the leveraged risk-parity and live-trading the other strategy for a year, and everything looked perfect.

Great, I thought. This is a simple system, only a few inputs, rock-solid for 35 years. Let's go! So I put a LARGE amount of my $$ into it.

In early 2022.

Story of my life, as soon as I risk significant money in something, I break it. Both strategies tanked bigtime in 2022. That year was the worst year for risk parity in a century. My other strategy also had by far its worst year in the 35-year history.

I gritted my teeth and kept trading it, albeit at reduced size -- drawdowns happen, and (assuming you can stop the bleeding) you have to keep trading to dig out of them. Almost 2 years later, I'm about where I was in late 2022.

I'll let you know if I give up and quit trading it. That's when risk parity will start working again ...

 

[mathjak107]

ha ha ha , sounds like something that would happen to me …

these kinds of stories is what prevents me from going whole hog in to something like this …

if i breaks it will break bad… but in this case i think the magic MAY be in the use of dbmf

 

[pb4uski]

Boy, I hear you. I did a lot of research on the HFEA (Hedgefundie's Excellent Adventure) portfolio and in Portfolio Visualizr it show great promise, so in late December I took some money in an inherited Roth IRA and invested it for DGS. The portfolio is 55% UPRO and 45% TMF and is rebalanced quarterly.

I compare it to 55% SPY and 45% TLT as an unlevered version.

So far, the HFEA portfolio has gained 1.9%... a 16.3% increase in UPRO offset by a 15.4% decline in TMF. The unlevered version has gained 1.3%... a 6.2% increase in SPY offset by a 4.6% decline in TLT.

So while I expected the real leverage to be about 200% as the 300% gets diluted, so far it is only 147%.

It's not a lot of money and is a bit of an experiment so I'll hang in there for a while and see what happens.

https://www.optimizedportfolio.com/hedgefundie-adventure/

In a hurry? Here are the highlights:

• Hedgefundie [-]is [/-]was a member of the Bogleheads forum.
• Hedgefundie created a thread in February 2019 proposing a 3x leveraged ETF investing strategy based on risk parity using the S&P 500 index (UPRO) and long-term treasury bonds (TMF) held in a 40/60 allocation. The thread later expanded into a Part 2.
• Hedgefundie later updated the strategy's asset allocation in August 2019 to 55/45 UPRO/TMF.
• Extensive backtesting, discussion, and analysis within the thread by members of the Bogleheads forum supports the validity and potential market outperformance of the strategy.
• The proposed strategy calls for quarterly rebalancing.
• Several different protocols/variations of the strategy emerged in the Excellent Adventure thread, including monthly rebalancing, rebalancing bands, and volatility targeting with various lookback periods.
• Some users have added a dash of TQQQ (3x the NASDAQ 100 index) for a minor tech tilt, as Big Tech has had a stellar run recently.
• It is recommended to implement the strategy within a Roth IRA on M1 Finance, to avoid tax implications and to make regular rebalancing seamless and easy.

 

[mathjak107]

going back to 2019 the carolina reaper did very well thru covid and the rate rise …it looks like dbmf made that difference

 

[mathjak107]

Boy, I hear you. I did a lot of research on the HFEA (Hedgefundie's Excellent Adventure) portfolio and in Portfolio Visualizr it show great promise, so in late December I took some money in an inherited Roth IRA and invested it for DGS. The portfolio is 55% UPRO and 45% TMF and is rebalanced quarterly.

I compare it to 55% SPY and 45% TLT as an unlevered version.

So far, the HFEA portfolio has gained 1.9%... a 16.3% increase in UPRO offset by a 15.4% decline in TMF. The unlevered version has gained 1.3%... a 6.2% increase in SPY offset by a 4.6% decline in TLT.

So while I expected the real leverage to be about 200% as the 300% gets diluted, so far it is only 147%.

It's not a lot of money and is a bit of an experiment so I'll hang in there for a while and see what happens.

https://www.optimizedportfolio.com/hedgefundie-adventure/

here is nothing that says that when stocks go down , bonds go up ,so i see little risk parity there ….

there is never a see saw action between assets ..there is only assets reacting to economic conditions..rising rates and inflation is kryptonite to stocks and bonds .

at least dbmf takes short positions

 

[mathjak107]

During the 2008-2009 crisis, all stocks dropped, including many bond funds.

Strategy was designed based upon false assumptions about asset class correlations.

IT IS NOT LIKE YOU CAN ASSUME BECAUSE STOCKS WENT DOWN BONDS WILL GO UP

that is a very very important concept to understand.

You can see that despite what the correlation data shows there is no seesaw link to each other that makes one go up when the other goes down .

That is the flaw in the correlation data . It is how those assets respond to what is going on in the economy that determines when they go up or down or move together not correlation data .


Simply relying upon historical asset class correlations is dangerous because many asset class correlations do not explain why the correlations exist, and what might cause them to change in the future. E.g. the correlation between stocks and bonds in the
1970s was 0.51
1980s was 0.32
1990s was 0.54
2000 to 2009 was -0.83, and
40-year period from 1972-2011 was 0.06.

Assets move for very specific reasons having to do with what is going on in the overall economy, and not what each other is doing as asset class correlations assume.

 

[pb4uski]

I agree... the HFEA portfolio presumes that when stocks zig that bonds zag and it is unclear if you can rely on that happening anymore.

 

[mathjak107]

I agree... the HFEA portfolio presumes that when stocks zig that bonds zag and it is unclear if you can rely on that happening anymore.

you never could count on it if you look at the correlations over history between stocks and bonds

 

[38Chevy454]

As with all of these financial models, there are many variables in play at any given time. I believe that is why the old saying "when stocks go down, bonds go up" does not work in reality. Especially since bonds are much more interest rate sensitive as one of the main variables that affect pricing. I do think bonds do serve a purpose on fixed income side, and are generally safe as ballast against stock variability.

I'm tempted to try the leveraged reaper portfolio for longer term money. It should do better in the long run due to the leveraging.

 

[mathjak107]

As with all of these financial models, there are many variables in play at any given time. I believe that is why the old saying "when stocks go down, bonds go up" does not work in reality. Especially since bonds are much more interest rate sensitive as one of the main variables that affect pricing. I do think bonds do serve a purpose on fixed income side, and are generally safe as ballast against stock variability.

I'm tempted to try the leveraged reaper portfolio for longer term money. It should do better in the long run due to the leveraging.

so far the reaper has been acting nicely and that’s scary because it can end up having a lot of money committed and then it breaks

 

[Quattro73]

Any updates?

 

[mathjak107]

yes , it is performing very well compared to my traditional 60/40 and i have been adding a bit more to it .

ytd

upro up 22.76% 20% of the portfolio
dbmf up 7.34. 67% of the portfolio

tyd down 8.66 13% of the portfolio

on the other hand my 60/40 up 6.23

 

[pb4uski]

I added this LRP portfolio to my tracking model for the HFEA portfolio vs a 55/45 portfolio that HFEA is based on.

Limited time period since Dec 2023, but the LPR portfolio is the leader thus far with a 17.0% increase vs 3.6% for HFEA portfolios vs 1.9% for unleveraged (55/45 SPY/TMF)

 

[mathjak107]

thought i would compare the original risk parity portfolio the permanent portfolio with the reaper

the permanent portfolio is the grand pappy of risk parity models with
25% total market
25% gold
25% long term treasuries
25% cash instruments or very short term bonds

from june 2019 to 3/1/2024

the reaper has taken 100k and is now 177,267 a cagr of 11.27% , sharpe ratio of .95

the PP is 130,239 5.72% cagr sharpe .44

pp worst year down 13.22 ,worst draw down 16.56%

for the reaper , worst year down 2.57% worst draw down 9.45%

moving to 2020

reaper 150,152 10.25% cagr sharpe .72

pp 118,484 4.15% cagr. sharpe .27

the higher the sharpe the better the risk vs reward was

moving to 2021

reaper 140,646. 11.37 cagr. sharpe .81

pp 101,788. .56% cagr sharpe minus .16


2022 the big down year

reaper 111,765 5.26% cagr sharpe .20

pp 97,665 minus 1.08% cage sharpe minus .40


2023

reaper 114,706 sharpe .69

pp. 112,548 sharpe .55


2024

reaper 106,841

pp. 100,542

 

[GaryInCO]

from june 2019 to 3/1/2024
the reaper has taken 100k and is now 177,267 a cagr of 11.27% , sharpe ratio of .95
for the reaper , worst year down 2.57% worst draw down 9.45%

moving to 2020
reaper 150,152 10.25% cagr sharpe .72

moving to 2021
reaper 140,646. 11.37 cagr. sharpe .81

2022 the big down year
reaper 111,765 5.26% cagr sharpe .20

2023
reaper 114,706 sharpe .69

2024
reaper 106,841

I'm confused. Reaper went from 150k in 2020 to 111k in 2022 and 106k in 2024, yet you show positive cagr and Sharpe every year. ??

 

[mathjak107]

the less the number of years you are in the smaller the balance. those years are starting years .

so 2024 means year to date , 2023 means starting in 2023 to 2024 , 2022 means starting in 2022 to 2024 ,etc

 

[pb4uski]

I assume that you used Portfolio Visualizer to analyze this. Is there a reason why you didn't attach a link or screenshot?

 

[mathjak107]

i did it from my desk top and am not sure how to screen shot it