HEDGEFUNDIE's excellent adventure Part II: The next journey

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
occambogle
Posts: 526
Joined: Thu Dec 12, 2019 4:58 am

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by occambogle »

Perfect Uncertainty wrote: Mon Sep 07, 2020 9:54 am I use Excel on a Mac so maybe that’s an issue but the 3 queries you have error out. Maybe there is something you didn’t include in your version?
Just to say I also had the same issue with Excel on Mac.... shame, as the spreadsheet looks nice.
Mickelous
Posts: 132
Joined: Mon Apr 20, 2020 11:24 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by Mickelous »

kerstverlichting wrote: Tue Sep 08, 2020 5:12 am Alright, thanks for the explanation. I can only load 100 days of data anyway using the Yahoo connection. I will try to add the following things though:
- Option for 3-way TMF/UPRO/TQQQ, need to be able to set the UPRO/TQQQ split %
- Option for quarterly rebalancing, using 63 days of data to calculate volatility (still within the Yahoo limits), ability to easily switch between the two.
Thanks man you're doing great things. If this gives even 2 percent better returns than HFEA you will be helping accelerate retirement for many people.
runeberg
Posts: 4
Joined: Sat Jul 25, 2020 7:11 am

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by runeberg »

I must need this explained to me like a 5 year old.

If I were to use the inverse volatility strategy instead of the HFEA 55/45 split (which is in a standard configuration quarterly rebalanced):
- I would rebalance monthly
- I would use the calculated inverse volatility metric based on the last 21 days for my next months split?
User avatar
willthrill81
Posts: 20863
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by willthrill81 »

runeberg wrote: Tue Sep 08, 2020 9:46 am I must need this explained to me like a 5 year old.

If I were to use the inverse volatility strategy instead of the HFEA 55/45 split (which is in a standard configuration quarterly rebalanced):
- I would rebalance monthly
- I would use the calculated inverse volatility metric based on the last 21 days for my next months split?
Yes and yes.
“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
rchmx1
Posts: 113
Joined: Sat Oct 26, 2019 6:38 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by rchmx1 »

kerstverlichting wrote: Mon Sep 07, 2020 1:58 pm Maybe someone else can test on windows to confirm.
Works great on Windows. I wasn't sure if I'd need to update the data with each new date's prices, but just opened the file and it updated for 9/8 automatically. Thanks for putting this together!
Mickelous
Posts: 132
Joined: Mon Apr 20, 2020 11:24 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by Mickelous »

Anyone have a graph for how well regular HFEA does in backtesting vs inverse volatility / minimum variance? Or maybe a Monte Carlo sim with those would even be better.
RovenSkyfall
Posts: 82
Joined: Wed Apr 01, 2020 11:40 am

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by RovenSkyfall »

kerstverlichting wrote: Mon Sep 07, 2020 9:24 am
Btw, can anyone provide a quick explanation of the difference between minimum variance and inverse volatility? Why use one over the other?
I found this paper helpful. https://papers.ssrn.com/sol3/papers.cfm ... id=2328254
RovenSkyfall
Posts: 82
Joined: Wed Apr 01, 2020 11:40 am

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by RovenSkyfall »

perfectuncertainty wrote: Mon Sep 07, 2020 10:56 pm I personally think that including 6 months of volatility to set a portfolio today isn't the best route. We are rebalancing monthly with AAA so I prefer the volatility calculation to be shorter in duration.
I could be totally off here, but my impression was that the 6 mo historical data was actually to calculate momentum in AAA. Here
RovenSkyfall
Posts: 82
Joined: Wed Apr 01, 2020 11:40 am

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by RovenSkyfall »

Mickelous wrote: Tue Sep 08, 2020 1:36 pm Anyone have a graph for how well regular HFEA does in backtesting vs inverse volatility / minimum variance? Or maybe a Monte Carlo sim with those would even be better.
There is a graph on the page before this one with the outcomes of the different AA's
jarjarM
Posts: 276
Joined: Mon Jul 16, 2018 1:21 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by jarjarM »

RovenSkyfall wrote: Tue Sep 08, 2020 1:41 pm
perfectuncertainty wrote: Mon Sep 07, 2020 10:56 pm I personally think that including 6 months of volatility to set a portfolio today isn't the best route. We are rebalancing monthly with AAA so I prefer the volatility calculation to be shorter in duration.
I could be totally off here, but my impression was that the 6 mo historical data was actually to calculate momentum in AAA. Here
It is. That's what was presented by smartly and the resolve assets management.
jarjarM
Posts: 276
Joined: Mon Jul 16, 2018 1:21 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by jarjarM »

Mickelous wrote: Tue Sep 08, 2020 1:36 pm Anyone have a graph for how well regular HFEA does in backtesting vs inverse volatility / minimum variance? Or maybe a Monte Carlo sim with those would even be better.
Here is the backtest using the original simulated dataset.

viewtopic.php?p=5476830#p5476830

The problem with monte carlo sim is that garbage in garbage out issue, what would be your expected return for the assets, what would be the distribution, what would be the correlation/or lack there of? So that's why generally I take monte carlo sim of financial data with a grain of salt.
Mickelous
Posts: 132
Joined: Mon Apr 20, 2020 11:24 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by Mickelous »

RovenSkyfall wrote: Tue Sep 08, 2020 1:37 pm
kerstverlichting wrote: Mon Sep 07, 2020 9:24 am
Btw, can anyone provide a quick explanation of the difference between minimum variance and inverse volatility? Why use one over the other?
I found this paper helpful. https://papers.ssrn.com/sol3/papers.cfm ... id=2328254
That explains minimum variance but I didn't see much or anything about inverse volatility.
RovenSkyfall
Posts: 82
Joined: Wed Apr 01, 2020 11:40 am

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by RovenSkyfall »

Mickelous wrote: Tue Sep 08, 2020 2:05 pm
RovenSkyfall wrote: Tue Sep 08, 2020 1:37 pm
kerstverlichting wrote: Mon Sep 07, 2020 9:24 am
Btw, can anyone provide a quick explanation of the difference between minimum variance and inverse volatility? Why use one over the other?
I found this paper helpful. https://papers.ssrn.com/sol3/papers.cfm ... id=2328254
That explains minimum variance but I didn't see much or anything about inverse volatility.
I believe it is Exhibit 2 on page 20 of the paper where they discuss the inverse volatility. I think they just call it volatility weighted, but I ma be wrong.
RovenSkyfall
Posts: 82
Joined: Wed Apr 01, 2020 11:40 am

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by RovenSkyfall »

kerstverlichting wrote: Mon Sep 07, 2020 6:45 am
Image
Thank you so much for doing this, it is truly amazing. One quick question: Does the AAA Minimum variance include 6 months of momentum, or just MV in the AA?
jarjarM
Posts: 276
Joined: Mon Jul 16, 2018 1:21 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by jarjarM »

RovenSkyfall wrote: Tue Sep 08, 2020 2:23 pm
Mickelous wrote: Tue Sep 08, 2020 2:05 pm
RovenSkyfall wrote: Tue Sep 08, 2020 1:37 pm
kerstverlichting wrote: Mon Sep 07, 2020 9:24 am
Btw, can anyone provide a quick explanation of the difference between minimum variance and inverse volatility? Why use one over the other?
I found this paper helpful. https://papers.ssrn.com/sol3/papers.cfm ... id=2328254
That explains minimum variance but I didn't see much or anything about inverse volatility.
I believe it is Exhibit 2 on page 20 of the paper where they discuss the inverse volatility. I think they just call it volatility weighted, but I ma be wrong.
Yup, it's exhibit 2 of the PDF discussing the volatility weighted strategy (60 day look back). Here is another one from Resolve:
When investments have similar expected Sharpe ratios, and an investor cannot reliably estimate correlations (or we can assume correlations are homogeneous), the optimal portfolio would be weighted in proportion to the inverse of the assets’ volatilities. When investments have similar expected returns (independent of volatility) and unknown correlations, the Inverse Variance portfolio is mean-variance optimal. Note that the Inverse Volatility portfolio is consistent with the Maximum Diversification portfolio, and the Inverse Variance portfolio approximates a Minimum Variance portfolio, when all investments have identical pairwise correlations.
https://investresolve.com/portfolio-opt ... l-methods/
Hydromod
Posts: 419
Joined: Tue Mar 26, 2019 10:21 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by Hydromod »

Mickelous wrote: Tue Sep 08, 2020 1:36 pm Anyone have a graph for how well regular HFEA does in backtesting vs inverse volatility / minimum variance? Or maybe a Monte Carlo sim with those would even be better.
I have a bunch of backtesting comparisons here. You might browse back and forth; some explanation above that point, lots of figures below with different timing assumptions.
Mickelous
Posts: 132
Joined: Mon Apr 20, 2020 11:24 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by Mickelous »

Hydromod wrote: Tue Sep 08, 2020 3:20 pm
Mickelous wrote: Tue Sep 08, 2020 1:36 pm Anyone have a graph for how well regular HFEA does in backtesting vs inverse volatility / minimum variance? Or maybe a Monte Carlo sim with those would even be better.
I have a bunch of backtesting comparisons here. You might browse back and forth; some explanation above that point, lots of figures below with different timing assumptions.
So it looks like either way, inverse volatility and minimum variance give about 2% more return per year? That makes a huge difference in a 10-15 year period thanks.
User avatar
kerstverlichting
Posts: 27
Joined: Tue Feb 11, 2020 5:26 am
Contact:

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by kerstverlichting »

RovenSkyfall wrote: Tue Sep 08, 2020 2:55 pm
kerstverlichting wrote: Mon Sep 07, 2020 6:45 am
Image
Thank you so much for doing this, it is truly amazing. One quick question: Does the AAA Minimum variance include 6 months of momentum, or just MV in the AA?
Hi, no problem, it was literally 10 minutes of work. As for your question, I just took the calculations @perfectuncertainty provided. I was actually wondering the same thing. I hear this talk of 6 mos momentum, but it's not in the data. Perhaps @perfectuncertainty can shed a light on this. Is it not required? If it is, could someone please provide the calculation? I'd be happy to implement it, it's just that I'm not at all familiar with these sorts of calculations myself, so I just took what was posted on here already and made some adjustments for usability.

Btw, if anyone knows how to make it work for Apple users please also let me know as there seem to be a couple of people missing out right now.
mrtwstr
Posts: 42
Joined: Sun Aug 23, 2020 12:30 pm
Location: Connecticut

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by mrtwstr »

Still very new to this but wanted to say thanks so much for the spreadsheet! Helping me visualize what all of you are talking about.
perfectuncertainty
Posts: 124
Joined: Sun Feb 04, 2018 7:44 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by perfectuncertainty »

RovenSkyfall wrote: Tue Sep 08, 2020 1:41 pm
perfectuncertainty wrote: Mon Sep 07, 2020 10:56 pm I personally think that including 6 months of volatility to set a portfolio today isn't the best route. We are rebalancing monthly with AAA so I prefer the volatility calculation to be shorter in duration.
I could be totally off here, but my impression was that the 6 mo historical data was actually to calculate momentum in AAA. Here
It's one of the variables you can choose on PV.
perfectuncertainty
Posts: 124
Joined: Sun Feb 04, 2018 7:44 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by perfectuncertainty »

kerstverlichting wrote: Tue Sep 08, 2020 3:52 pm
RovenSkyfall wrote: Tue Sep 08, 2020 2:55 pm
kerstverlichting wrote: Mon Sep 07, 2020 6:45 am
Image
Thank you so much for doing this, it is truly amazing. One quick question: Does the AAA Minimum variance include 6 months of momentum, or just MV in the AA?
Hi, no problem, it was literally 10 minutes of work. As for your question, I just took the calculations @perfectuncertainty provided. I was actually wondering the same thing. I hear this talk of 6 mos momentum, but it's not in the data. Perhaps @perfectuncertainty can shed a light on this. Is it not required? If it is, could someone please provide the calculation? I'd be happy to implement it, it's just that I'm not at all familiar with these sorts of calculations myself, so I just took what was posted on here already and made some adjustments for usability.

Btw, if anyone knows how to make it work for Apple users please also let me know as there seem to be a couple of people missing out right now.
Excel on Mac works somewhat differently for web queries - I've opted not to use them. I actually use the builtin Excel Microsoft Stock prices tools with a toggle to enter manually or to pull data with approximately a 2 min delay.

Excel "Realtime" Stocks

For myself, I have a built quite a bit out for myself and have the data going back to 7/2010. My model uses Inverse Volatility, Minimum Variance, a multiple symbol Risk parity, Target Volatiity.

My model allows me to temper the calculation with cash overrides - for instance if I have indication that stocks are overvalued (or bonds) I have dropdowns in the model wich I can specify the a 5point scale ranging from Bullish, Tentative, Neutral, Bearish and Negative. Each one has my spreadsheet calculate AAA, maintain the ratios but pair it off. For instance, If I am Bullish on stocks, I allocate 100% of the AAA stock calculation, whereas if I am tentative I only allocate 70% etc. It graduates from 100% to 70% to 50% to 30% to 0%. Maybe it's been luck, and I like to think not :-), but I've avoided the stock rundown by pairing back very early to 0%. Since TMF and TLT weren't doing so hot to pair the risk I paired them down to 30% early last week. I use different factors for ascertaining my outlook on Stocks or Bonds - mostly technical indicators (very non-BH, I know).

Having the data and the model enables me to see how well the correlations functioned at different times.

Also, from my analysis I tend to find that QLD is a better holding than either TQQQ or UPRO - test it out for yourself. Using the asset correlation analysis tools on PV maybe you will reach the same conclusions. I do prefer it in AAA as a primary stock asset to TQQQ and UPRO - returns are almost as good and drawdowns and Sortino ratios are generally better for me.
Benosis
Posts: 68
Joined: Tue Apr 02, 2019 7:16 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by Benosis »

I am considering buying 55% UPRO / 45% TMF in a brokerage account.

Is a brokerage account fine or do people recommend using a Roth IRA? Also can I just use Robinhood or would you recommend Vanguard Brokerage Account?
chrisdds98
Posts: 142
Joined: Tue May 19, 2015 9:55 pm
Location: Austin, TX

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by chrisdds98 »

Benosis wrote: Tue Sep 08, 2020 9:32 pm I am considering buying 55% UPRO / 45% TMF in a brokerage account.

Is a brokerage account fine or do people recommend using a Roth IRA? Also can I just use Robinhood or would you recommend Vanguard Brokerage Account?
frequent selling can lead to short term cap gains. I did this in taxable as well. I'm just over 1 year in my taxable account now so I can rebalance and pay LTCG instead. Best to do it in a roth or traditional.

Can't buy leveraged funds at vanguard. seems like many here use m1 finance.
User avatar
willthrill81
Posts: 20863
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by willthrill81 »

Benosis wrote: Tue Sep 08, 2020 9:32 pm I am considering buying 55% UPRO / 45% TMF in a brokerage account.

Is a brokerage account fine or do people recommend using a Roth IRA? Also can I just use Robinhood or would you recommend Vanguard Brokerage Account?
I wouldn't recommend this. If you cannot do it with a tax-advantaged account, then don't implement it.

I use M1 Finance with a Roth IRA to implement a version of this approach.
“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
perfectuncertainty
Posts: 124
Joined: Sun Feb 04, 2018 7:44 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by perfectuncertainty »

For those of you wanting a 6-month volatility version for AAA here is an updated version.

I also added an additional AAA Inverse-Volatility Calculation that allows one to combine UPRO, TQQQ and UPRO and allocate accordingly. To duplicate this in PV here are the parameters and the backtest

AAA-6-Mo

Again, I prefer the 21-day version for reasons I've already stated.

It's showing the calculations through September 8th. PV doesn't have them updated yet, but they should match for a 6 month Volatility Period as the input, varying Inverse volatility, and Minimum Variance.
User avatar
Meaty
Posts: 796
Joined: Mon Jul 22, 2013 7:35 pm
Location: Texas

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by Meaty »

Benosis wrote: Tue Sep 08, 2020 9:32 pm I am considering buying 55% UPRO / 45% TMF in a brokerage account.

Is a brokerage account fine or do people recommend using a Roth IRA? Also can I just use Robinhood or would you recommend Vanguard Brokerage Account?
Others say not to use taxable, I think that concern is overblown. Yes, you will incur taxes when you rebalance but the benefits of the strategy should (no guarantee) outweigh that issue. If you can afford to pay the taxes then I don’t see the problem
"Discipline equals Freedom" - Jocko Willink
Alchemist
Posts: 558
Joined: Sat Aug 30, 2014 6:35 am

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by Alchemist »

I am curious how the portfolios of people who went with TQQQ instead of UPRO is holding up under the recent dip. TQQQ is down by nearly 1/3, but TMF has not gone up to compensate. Is the plan to still rebalance into TQQQ?

Just a question from an amused outside observer to your adventure.
perfectuncertainty
Posts: 124
Joined: Sun Feb 04, 2018 7:44 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by perfectuncertainty »

Alchemist wrote: Wed Sep 09, 2020 6:27 am I am curious how the portfolios of people who went with TQQQ instead of UPRO is holding up under the recent dip. TQQQ is down by nearly 1/3, but TMF has not gone up to compensate. Is the plan to still rebalance into TQQQ?

Just a question from an amused outside observer to your adventure.
The SP500 is down 4.78% in September.

TQQQ/TMF using AAA Inverse Volatility is down 9.06% in September, using Minimum Variance it is down 5.62% in September. HFEA it's down 15.7%
UPRO/TMF using AAA Inverse Volatility is down 7.24% in September, using Minimum Variance it is down 9.09% in September. HFEA it's down 7.84%

QLD/TMF using AAA Inverse Volatility is down 7.62% in September, using Minimum Variance it is down 9.08% in September. HFEA it's down 9.71%
Mickelous
Posts: 132
Joined: Mon Apr 20, 2020 11:24 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by Mickelous »

perfectuncertainty wrote: Tue Sep 08, 2020 11:01 pm For those of you wanting a 6-month volatility version for AAA here is an updated version.

I also added an additional AAA Inverse-Volatility Calculation that allows one to combine UPRO, TQQQ and UPRO and allocate accordingly. To duplicate this in PV here are the parameters and the backtest

AAA-6-Mo

Again, I prefer the 21-day version for reasons I've already stated.

It's showing the calculations through September 8th. PV doesn't have them updated yet, but they should match for a 6 month Volatility Period as the input, varying Inverse volatility, and Minimum Variance.
What about 1 month inverse volatility with three assets?
Mickelous
Posts: 132
Joined: Mon Apr 20, 2020 11:24 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by Mickelous »

perfectuncertainty wrote: Tue Sep 08, 2020 11:01 pm For those of you wanting a 6-month volatility version for AAA here is an updated version.

I also added an additional AAA Inverse-Volatility Calculation that allows one to combine UPRO, TQQQ and UPRO and allocate accordingly. To duplicate this in PV here are the parameters and the backtest

AAA-6-Mo

Again, I prefer the 21-day version for reasons I've already stated.

It's showing the calculations through September 8th. PV doesn't have them updated yet, but they should match for a 6 month Volatility Period as the input, varying Inverse volatility, and Minimum Variance.
Not allowing me to access it now.
Perfect Uncertainty
Posts: 12
Joined: Wed Jun 24, 2020 10:20 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by Perfect Uncertainty »

Mickelous wrote: Wed Sep 09, 2020 8:09 am
perfectuncertainty wrote: Tue Sep 08, 2020 11:01 pm For those of you wanting a 6-month volatility version for AAA here is an updated version.

I also added an additional AAA Inverse-Volatility Calculation that allows one to combine UPRO, TQQQ and UPRO and allocate accordingly. To duplicate this in PV here are the parameters and the backtest

AAA-6-Mo

Again, I prefer the 21-day version for reasons I've already stated.

It's showing the calculations through September 8th. PV doesn't have them updated yet, but they should match for a 6 month Volatility Period as the input, varying Inverse volatility, and Minimum Variance.
Not allowing me to access it now.
I’m out on the water this morning. I’ll update the link to shared mode later. Also, it has a 3 asset 21 day.
worthit
Posts: 263
Joined: Tue Jun 19, 2018 2:10 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by worthit »

willthrill81 wrote: Tue Sep 08, 2020 10:48 am
runeberg wrote: Tue Sep 08, 2020 9:46 am I must need this explained to me like a 5 year old.

If I were to use the inverse volatility strategy instead of the HFEA 55/45 split (which is in a standard configuration quarterly rebalanced):
- I would rebalance monthly
- I would use the calculated inverse volatility metric based on the last 21 days for my next months split?
Yes and yes.
Can some one please explain how to execute the Inverse Volatility strategy to me like you would to a 1 year old? :D. Currently using HFEA's leveraged strategy.

- What is the optimal allocation percentages, if there is one? If not, what is recommended?

- Can I still use UPRO/TMF as I currently use 55%/45% or are there any other ETFs that serves the purpose better?

- How do I calculate the inverse volatility metric in order to rebalance? In other words, what is my expected threshold to rebalance monthly?

Thanks to perfectuncertainity and kerstverlichting, I downloaded a copy of the auto-popualed SS that has the 21 day data. Do I just average the returns or is it something else I need to be looking at?

TIA.
User avatar
privatefarmer
Posts: 709
Joined: Mon Sep 08, 2014 2:45 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by privatefarmer »

I hate feeling like a free loader.... but anyone got a spreadsheet that’d calculate the allocations if using min variance on 4 assets? I Think you can also do it via “excel solver” but am too dumb to figure it out. Thanks
Mickelous
Posts: 132
Joined: Mon Apr 20, 2020 11:24 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by Mickelous »

privatefarmer wrote: Wed Sep 09, 2020 9:18 am I hate feeling like a free loader.... but anyone got a spreadsheet that’d calculate the allocations if using min variance on 4 assets? I Think you can also do it via “excel solver” but am too dumb to figure it out. Thanks
What four assets were you looking at?
User avatar
privatefarmer
Posts: 709
Joined: Mon Sep 08, 2014 2:45 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by privatefarmer »

I use qqq/TLT/VNQ/FSAGX (gold miners) and then buy the 3x ETF versions of them at same allocation. I use PV for this which I don’t mind paying for but am concerned that someday PV won’t be around anymore.... so a spreadsheet would be nice.
User avatar
willthrill81
Posts: 20863
Joined: Thu Jan 26, 2017 3:17 pm
Location: USA

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by willthrill81 »

worthit wrote: Wed Sep 09, 2020 8:47 am
willthrill81 wrote: Tue Sep 08, 2020 10:48 am
runeberg wrote: Tue Sep 08, 2020 9:46 am I must need this explained to me like a 5 year old.

If I were to use the inverse volatility strategy instead of the HFEA 55/45 split (which is in a standard configuration quarterly rebalanced):
- I would rebalance monthly
- I would use the calculated inverse volatility metric based on the last 21 days for my next months split?
Yes and yes.
Can some one please explain how to execute the Inverse Volatility strategy to me like you would to a 1 year old? :D. Currently using HFEA's leveraged strategy.

- What is the optimal allocation percentages, if there is one? If not, what is recommended?

- Can I still use UPRO/TMF as I currently use 55%/45% or are there any other ETFs that serves the purpose better?

- How do I calculate the inverse volatility metric in order to rebalance? In other words, what is my expected threshold to rebalance monthly?

Thanks to perfectuncertainity and kerstverlichting, I downloaded a copy of the auto-popualed SS that has the 21 day data. Do I just average the returns or is it something else I need to be looking at?

TIA.
I don't know all of those answers. Others here are much more versed in such calculations.
“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
User avatar
privatefarmer
Posts: 709
Joined: Mon Sep 08, 2014 2:45 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by privatefarmer »

Inverse vol is simple. You just take the inverse of the volatility of each asset, so the assets that are more volatile will have a smaller % of your portfolio. The calculation is 1-(StandDev of asset/sum of the stand devs of all assets) = % of portfolio that asset should be. Ie if you have two assets, asset A has STD dev of 15% and asset B has std dev of 10%,

Asset A allocation = 1-(15/25) = 0.4 or 40%

Asset B = 1-(10/25) = 0.6 or 60%
worthit
Posts: 263
Joined: Tue Jun 19, 2018 2:10 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by worthit »

privatefarmer wrote: Wed Sep 09, 2020 10:00 am Inverse vol is simple. You just take the inverse of the volatility of each asset, so the assets that are more volatile will have a smaller % of your portfolio. The calculation is 1-(StandDev of asset/sum of the stand devs of all assets) = % of portfolio that asset should be. Ie if you have two assets, asset A has STD dev of 15% and asset B has std dev of 10%,

Asset A allocation = 1-(15/25) = 0.4 or 40%

Asset B = 1-(10/25) = 0.6 or 60%
Thanks privatefarmer. So the asset allocation is decided :D .
AnilG wrote: Tue Sep 08, 2020 5:00 am I am using TQQQ/TMF strategy with monthly rebalancing using inverse volatility with 20 day volatility.
Now this is what I am not clear about. What does "rebalancing using inverse volatility with 20 day volatility" mean in the above response?

TIA.
User avatar
kerstverlichting
Posts: 27
Joined: Tue Feb 11, 2020 5:26 am
Contact:

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by kerstverlichting »

Hi guys, I updated the file a little to allow for some more tickers, also added a few other features, like setting a custom range. The default is 1 month, as this was suggested to be optimal, but you can go up to 100 trading days (~4.75 mos). Unfortunately that's the max of what I can load using Yahoo, so I cannot add 6 mos for people who want to use that. I am indeed aware of the built in stock feature that the newer Excel has, but it only loads the latest data, not any historical numbers, so for now I've got to stick with Yahoo unless someone knows of a better alternative. Btw, I had to mess with the calculations a little, since my model allows for setting the allocation per asset (optionally). I put the whole thing together today, so I hope I did it all correctly, no guarantees fyi.

You can add up to 3 stock etfs (UPRO, TQQQ, QLD), and 2 bond + 1 gold etf (as a bonus feature, note that it'll be treated as being bonds in the calculations) (TMF, TYD, GLD). If you are savvy, you can change the tickers yourself (Data>Queries & Connections>Right click>Edit>PowerQuery will open>Data source settings>Change source>Replace the ticker in the url with the one you want to use>Edit the query so that the Date column title is the new ticker name). It should work I suppose , but didn't try this myself jet.

The file also gives you the allocation by usd amount, percentage, and number of shares to buy for convenience. :sharebeer It should also handle stock splits correctly, but I didn't really get to test this properly.

Preview:
Image
The whole empty bit gets filled depending on the date range that is chosen.

Download: http://s000.tinyupload.com/index.php?fi ... 3284757400

EDIT: new version here: viewtopic.php?f=10&t=288192&p=5486976#p5486976

Hope this benefits people, if you have any remarks or suggestions, please let me know.
Last edited by kerstverlichting on Thu Sep 10, 2020 5:49 am, edited 1 time in total.
User avatar
privatefarmer
Posts: 709
Joined: Mon Sep 08, 2014 2:45 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by privatefarmer »

kerstverlichting wrote: Wed Sep 09, 2020 10:36 am Hi guys, I updated the file a little to allow for some more tickers, also added a few other features, like setting a custom range. The default is 1 month, as this was suggested to be optimal, but you can go up to 100 trading days (~4.75 mos). Unfortunately that's the max of what I can load using Yahoo, so I cannot add 6 mos for people who want to use that. I am indeed aware of the built in stock feature that the newer Excel has, but it only loads the latest data, not any historical numbers, so for now I've got to stick with Yahoo unless someone knows of a better alternative. Btw, I had to mess with the calculations a little, since my model allows for setting the allocation per asset (optionally). I put the whole thing together today, so I hope I did it all correctly, no guarantees fyi.

You can add up to 3 stock etfs (UPRO, TQQQ, QLD), and 2 bond + 1 gold etf (as a bonus feature, note that it'll be treated as being bonds in the calculations) (TMF, TYD, GLD). If you are savvy, you can change the tickers yourself (Data>Queries & Connections>Right click>Edit>PowerQuery will open>Data source settings>Change source>Replace the ticker in the url with the one you want to use>Edit the query so that the Date column title is the new ticker name). It should work I suppose , but didn't try this myself jet.

The file also gives you the allocation by usd amount, percentage, and number of shares to buy for convenience. :sharebeer It should also handle stock splits correctly, but I didn't really get to test this properly.

Preview:
Image
The whole empty bit gets filled depending on the date range that is chosen.

Download: http://s000.tinyupload.com/index.php?fi ... 3284757400

Hope this benefits people, if you have any remarks or suggestions, please let me know.
Bro, I just nominated you for the Nobel peace prize. This is freakin awesome thanks so much. I’ll play with it tonight.
jarjarM
Posts: 276
Joined: Mon Jul 16, 2018 1:21 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by jarjarM »

kerstverlichting wrote: Wed Sep 09, 2020 10:36 am
Hope this benefits people, if you have any remarks or suggestions, please let me know.
Just to point out a few minor details on great work from kerstverlichting, 6 month in the original AAA work was referring to 6 month look back for momentum factor (the original work looks at 10 distinct assets and choose top 5 I believe) so for a 2/3/4 component strategy, it makes no difference. Also, the current excellent excel sheet from kerstverlichting is providing the balance for 2 component portfolio, the minimum variance for 3 or more will require excel solver or some other approximation method (I personally use python CLA). :beer
RovenSkyfall
Posts: 82
Joined: Wed Apr 01, 2020 11:40 am

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by RovenSkyfall »

jarjarM wrote: Fri Sep 04, 2020 1:12 pm
danyboy7 wrote: Fri Sep 04, 2020 12:48 pm
Thanks a lot,how do I upload them on PV ? Is it required the Pro-version ?
It's being done. I did this as a pet project during the shelter at home fun.

Image

Code: Select all

		UPRO Baseline	55/45 Portfolio	VolatilityPortfolio	MomentumPortfolio	AdaptiveAssetAllocation
cagr		10.0			14.9		17.8			15.0			18.1
sharpe		0.46			0.62		0.80			0.64			0.82
volatility	0.54			0.29		0.24			0.28			0.24
max_drawdown	-98.2			-74.5		-44.2			-69.9			-39.2
60/40 vs. 55/45 didn't show much difference so I didn't simulate it separately. The data is to 2/2019 when the simulated data end. Volatility is using inverse volatility from prior month, momentum is best performing part from prior 6 months. Adaptive AA is minimum variance. Have fun.
Can I ask some clarifying questions that people might be able to answer? The Adaptive AA is minimum variance alone (i.e. doesn't include momentum), correct? Is the Combo line the inverse volatility combined with momentum? Is momentum accurate in just two funds? How does it get weighted when combined (either here or in the classic description of the Adaptive AA)? Thank you all in advance for your help!
hilink73
Posts: 476
Joined: Tue Sep 20, 2016 3:29 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by hilink73 »

How quick things change.

Asset allocation for target volatility 25% at end of last month was 80% UPRO / 20% TMF.
Today it is 43% UPRO and 57% TMF according to PV.

-11% in September. I'm feeling lucky.
perfectuncertainty
Posts: 124
Joined: Sun Feb 04, 2018 7:44 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by perfectuncertainty »

Mickelous wrote: Wed Sep 09, 2020 8:09 am
perfectuncertainty wrote: Tue Sep 08, 2020 11:01 pm For those of you wanting a 6-month volatility version for AAA here is an updated version.

I also added an additional AAA Inverse-Volatility Calculation that allows one to combine UPRO, TQQQ and UPRO and allocate accordingly. To duplicate this in PV here are the parameters and the backtest

AAA-6-Mo

Again, I prefer the 21-day version for reasons I've already stated.

It's showing the calculations through September 8th. PV doesn't have them updated yet, but they should match for a 6 month Volatility Period as the input, varying Inverse volatility, and Minimum Variance.
Not allowing me to access it now.
UPDATED Spreadsheet
jarjarM
Posts: 276
Joined: Mon Jul 16, 2018 1:21 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by jarjarM »

RovenSkyfall wrote: Wed Sep 09, 2020 12:27 pm
jarjarM wrote: Fri Sep 04, 2020 1:12 pm
danyboy7 wrote: Fri Sep 04, 2020 12:48 pm
Thanks a lot,how do I upload them on PV ? Is it required the Pro-version ?
It's being done. I did this as a pet project during the shelter at home fun.

Image

Code: Select all

		UPRO Baseline	55/45 Portfolio	VolatilityPortfolio	MomentumPortfolio	AdaptiveAssetAllocation
cagr		10.0			14.9		17.8			15.0			18.1
sharpe		0.46			0.62		0.80			0.64			0.82
volatility	0.54			0.29		0.24			0.28			0.24
max_drawdown	-98.2			-74.5		-44.2			-69.9			-39.2
60/40 vs. 55/45 didn't show much difference so I didn't simulate it separately. The data is to 2/2019 when the simulated data end. Volatility is using inverse volatility from prior month, momentum is best performing part from prior 6 months. Adaptive AA is minimum variance. Have fun.
Can I ask some clarifying questions that people might be able to answer? The Adaptive AA is minimum variance alone (i.e. doesn't include momentum), correct? Is the Combo line the inverse volatility combined with momentum? Is momentum accurate in just two funds? How does it get weighted when combined (either here or in the classic description of the Adaptive AA)? Thank you all in advance for your help!
Here is the original AAA strategy proposed by smartly based on the white paper:
Strategy rules tested:
- At the close on the last trading day of the month, calculate the 6-month (126-day) return for each of the following 10 asset classes: US equities (represented by SPY), European equities (EZU), Japanese equities (EWJ), emerging market equities (EEM), US REITs (VNQ), international REITs (RWX), US 10-year Treasuries (IEF), US 30-year Treasuries (TLT), commodities (DBC) and gold (GLD).

- Go long at the close the five assets (i.e. half of the portfolio) with the highest 6-month return. Weight the five assets according to minimum variance optimization, using a “weighted” covariance matrix calculated based on 126-day correlation and 20-day volatility.

- Hold positions until the final trading day of the following month. Rebalance the entire portfolio monthly, regardless of whether there is a change in position.
Momentum (6 months look back) is build into minimum variance but not really applicable here. The original paper uses 10 assets class and pick the top 5. However, the covariance matrix is utilizing 6month look back (6*21 trading days). For momentum, it's picking the top 2 and go 50/50 (or 33.3/33.3/33.3 for top 3). Combo is combining momentum and inverse volatility together. Hope this helps. :beer
perfectuncertainty
Posts: 124
Joined: Sun Feb 04, 2018 7:44 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by perfectuncertainty »

privatefarmer wrote: Wed Sep 09, 2020 10:59 am
kerstverlichting wrote: Wed Sep 09, 2020 10:36 am Hi guys, I updated the file a little to allow for some more tickers, also added a few other features, like setting a custom range. The default is 1 month, as this was suggested to be optimal, but you can go up to 100 trading days (~4.75 mos). Unfortunately that's the max of what I can load using Yahoo, so I cannot add 6 mos for people who want to use that. I am indeed aware of the built in stock feature that the newer Excel has, but it only loads the latest data, not any historical numbers, so for now I've got to stick with Yahoo unless someone knows of a better alternative. Btw, I had to mess with the calculations a little, since my model allows for setting the allocation per asset (optionally). I put the whole thing together today, so I hope I did it all correctly, no guarantees fyi.

You can add up to 3 stock etfs (UPRO, TQQQ, QLD), and 2 bond + 1 gold etf (as a bonus feature, note that it'll be treated as being bonds in the calculations) (TMF, TYD, GLD). If you are savvy, you can change the tickers yourself (Data>Queries & Connections>Right click>Edit>PowerQuery will open>Data source settings>Change source>Replace the ticker in the url with the one you want to use>Edit the query so that the Date column title is the new ticker name). It should work I suppose , but didn't try this myself jet.

The file also gives you the allocation by usd amount, percentage, and number of shares to buy for convenience. :sharebeer It should also handle stock splits correctly, but I didn't really get to test this properly.

Preview:
Image
The whole empty bit gets filled depending on the date range that is chosen.

Download: http://s000.tinyupload.com/index.php?fi ... 3284757400

Hope this benefits people, if you have any remarks or suggestions, please let me know.
Bro, I just nominated you for the Nobel peace prize. This is freakin awesome thanks so much. I’ll play with it tonight.
One suggestion: instead of picking the ratio of UPRO and TQQQ, why not let AAA do that for you - that's why we are using AAA
User avatar
kerstverlichting
Posts: 27
Joined: Tue Feb 11, 2020 5:26 am
Contact:

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by kerstverlichting »

perfectuncertainty wrote: Wed Sep 09, 2020 1:32 pm
privatefarmer wrote: Wed Sep 09, 2020 10:59 am
kerstverlichting wrote: Wed Sep 09, 2020 10:36 am Hi guys, I updated the file a little to allow for some more tickers, also added a few other features, like setting a custom range. The default is 1 month, as this was suggested to be optimal, but you can go up to 100 trading days (~4.75 mos). Unfortunately that's the max of what I can load using Yahoo, so I cannot add 6 mos for people who want to use that. I am indeed aware of the built in stock feature that the newer Excel has, but it only loads the latest data, not any historical numbers, so for now I've got to stick with Yahoo unless someone knows of a better alternative. Btw, I had to mess with the calculations a little, since my model allows for setting the allocation per asset (optionally). I put the whole thing together today, so I hope I did it all correctly, no guarantees fyi.

You can add up to 3 stock etfs (UPRO, TQQQ, QLD), and 2 bond + 1 gold etf (as a bonus feature, note that it'll be treated as being bonds in the calculations) (TMF, TYD, GLD). If you are savvy, you can change the tickers yourself (Data>Queries & Connections>Right click>Edit>PowerQuery will open>Data source settings>Change source>Replace the ticker in the url with the one you want to use>Edit the query so that the Date column title is the new ticker name). It should work I suppose , but didn't try this myself jet.

The file also gives you the allocation by usd amount, percentage, and number of shares to buy for convenience. :sharebeer It should also handle stock splits correctly, but I didn't really get to test this properly.

Preview:
Image
The whole empty bit gets filled depending on the date range that is chosen.

Download: http://s000.tinyupload.com/index.php?fi ... 3284757400

Hope this benefits people, if you have any remarks or suggestions, please let me know.
Bro, I just nominated you for the Nobel peace prize. This is freakin awesome thanks so much. I’ll play with it tonight.
One suggestion: instead of picking the ratio of UPRO and TQQQ, why not let AAA do that for you - that's why we are using AAA
It's a bonus feature, if you don't want to mess with it, check out the auto allocated area.

Even when setting the allocation yourself, you're only setting the allocation within a category (stocks or bonds). For example, you might want to go for a UPRO TQQQ TMF split and say the aaa would suggest to split it 30/40/30. If you're uncomfortable putting that much in TQQQ, you can use the allocation feature to lower the weight given to TQQQ. For example, you can set UPRO to 80% and TQQQ to 20%. After this the aaa will be applied and the new split should be 80% of 30+40 = 56% UPRO, 20% of 30+40 = 14% TQQQ, 30% TMF.

Now this is all an approximation. In reality, because the weights have been altered, the combined volatility will also have changed and thus it could be that the aaa for bonds has also in or decreased as a result.

But like I said, just an extra feature, if you prefer the model to do its work, use the auto allocated results.
Mickelous
Posts: 132
Joined: Mon Apr 20, 2020 11:24 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by Mickelous »

kerstverlichting wrote: Wed Sep 09, 2020 2:06 pm
perfectuncertainty wrote: Wed Sep 09, 2020 1:32 pm
privatefarmer wrote: Wed Sep 09, 2020 10:59 am
kerstverlichting wrote: Wed Sep 09, 2020 10:36 am Hi guys, I updated the file a little to allow for some more tickers, also added a few other features, like setting a custom range. The default is 1 month, as this was suggested to be optimal, but you can go up to 100 trading days (~4.75 mos). Unfortunately that's the max of what I can load using Yahoo, so I cannot add 6 mos for people who want to use that. I am indeed aware of the built in stock feature that the newer Excel has, but it only loads the latest data, not any historical numbers, so for now I've got to stick with Yahoo unless someone knows of a better alternative. Btw, I had to mess with the calculations a little, since my model allows for setting the allocation per asset (optionally). I put the whole thing together today, so I hope I did it all correctly, no guarantees fyi.

You can add up to 3 stock etfs (UPRO, TQQQ, QLD), and 2 bond + 1 gold etf (as a bonus feature, note that it'll be treated as being bonds in the calculations) (TMF, TYD, GLD). If you are savvy, you can change the tickers yourself (Data>Queries & Connections>Right click>Edit>PowerQuery will open>Data source settings>Change source>Replace the ticker in the url with the one you want to use>Edit the query so that the Date column title is the new ticker name). It should work I suppose , but didn't try this myself jet.

The file also gives you the allocation by usd amount, percentage, and number of shares to buy for convenience. :sharebeer It should also handle stock splits correctly, but I didn't really get to test this properly.

Preview:
Image
The whole empty bit gets filled depending on the date range that is chosen.

Download: http://s000.tinyupload.com/index.php?fi ... 3284757400

Hope this benefits people, if you have any remarks or suggestions, please let me know.
Bro, I just nominated you for the Nobel peace prize. This is freakin awesome thanks so much. I’ll play with it tonight.
One suggestion: instead of picking the ratio of UPRO and TQQQ, why not let AAA do that for you - that's why we are using AAA
It's a bonus feature, if you don't want to mess with it, check out the auto allocated area.

Even when setting the allocation yourself, you're only setting the allocation within a category (stocks or bonds). For example, you might want to go for a UPRO TQQQ TMF split and say the aaa would suggest to split it 30/40/30. If you're uncomfortable putting that much in TQQQ, you can use the allocation feature to lower the weight given to TQQQ. For example, you can set UPRO to 80% and TQQQ to 20%. After this the aaa will be applied and the new split should be 80% of 30+40 = 56% UPRO, 20% of 30+40 = 14% TQQQ, 30% TMF.

Now this is all an approximation. In reality, because the weights have been altered, the combined volatility will also have changed and thus it could be that the aaa for bonds has also in or decreased as a result.

But like I said, just an extra feature, if you prefer the model to do its work, use the auto allocated results.
I feel like this is the at the forefront of maximized index returns. In the future this could be the way the average Joe has his money managed, work 15 years and retire.
jarjarM
Posts: 276
Joined: Mon Jul 16, 2018 1:21 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by jarjarM »

Mickelous wrote: Wed Sep 09, 2020 2:32 pm
I feel like this is the at the forefront of maximized index returns. In the future this could be the way the average Joe has his money managed, work 15 years and retire.
Just beware that with such low interest rate, TMF won't help out the portfolio as much. The drawdown protection will be lessen (see what happens the last 3 trading days) and thus volatility of this strategy will increase. I don't think this is for the average joe nor should this be main part of anyone's portfolio as mentioned by the OP himself. It's a moonshoot and have a non zero chance of blowing up (almost did in March). Just a word of caution :beer
haranoth
Posts: 99
Joined: Wed Apr 06, 2016 12:07 pm

Re: HEDGEFUNDIE's excellent adventure Part II: The next journey

Post by haranoth »

perfectuncertainty wrote: Wed Sep 09, 2020 1:20 pm
Mickelous wrote: Wed Sep 09, 2020 8:09 am
perfectuncertainty wrote: Tue Sep 08, 2020 11:01 pm For those of you wanting a 6-month volatility version for AAA here is an updated version.

I also added an additional AAA Inverse-Volatility Calculation that allows one to combine UPRO, TQQQ and UPRO and allocate accordingly. To duplicate this in PV here are the parameters and the backtest

AAA-6-Mo

Again, I prefer the 21-day version for reasons I've already stated.

It's showing the calculations through September 8th. PV doesn't have them updated yet, but they should match for a 6 month Volatility Period as the input, varying Inverse volatility, and Minimum Variance.
Not allowing me to access it now.
UPDATED Spreadsheet
Hi,
Is this google sheets automatically updated everyday, I couldn't find the place, where you pull the data automatically for UPRO, TMF e.g.
=GOOGLEFINANCE("NYSEARCA:UPRO","price",TODAY()-21,TODAY())

Just asking, since I would like to actually take your sheet and modify it to figure out how much of a difference AAA minimum variance vs AAA inverse volatility vs Will's 25% target volatility vs. HFEA original 55/45 portfolio makes :)
Thanks for providing the google sheet, it's good to see some of the associated equations and try to learn some of this lingo.
Post Reply