TVIX/TQQQ combo

Discuss all general (i.e. non-personal) investing questions and issues, investing news, and theory.
Post Reply
Topic Author
differentidea
Posts: 7
Joined: Fri Jun 19, 2020 3:19 pm

TVIX/TQQQ combo

Post by differentidea »

For the foreseeable period of time I think there will be volatility even with the fed stepping up and trying to prop everything up.

So...

I read up alot on the UPRO/TMF 401k strategy which has been discussed at length here and I am trying to put a different spin on it while the world is the way it is, until a vaccine and/or the world economies actually start growing naturally.

20% TVIX

80% TQQQ (considering 55 TQQQ/25 UPRO)

Re balance weekly or with big spikes.

Yesterday and today this worked quiet well.

1yr Backtest here:
https://www.portfoliovisualizer.com/bac ... ion6_3=100

Thoughts?
CardioMD
Posts: 169
Joined: Thu Feb 06, 2020 11:40 am

Re: TVIX/TQQQ combo

Post by CardioMD »

differentidea wrote: Fri Jun 19, 2020 3:26 pm For the foreseeable period of time I think

Yesterday and today this worked quiet well.
How long is the foreseeable future? When the foreseeable future comes to pass what will you do then?

Edit to add: volatility is your friend not your enemy.
“The stock market is a giant distraction from the business of investing.” -Jack Bogle
Semantics
Posts: 163
Joined: Tue Mar 10, 2020 1:42 pm

Re: TVIX/TQQQ combo

Post by Semantics »

TVIX loses over 90% of its value most years and having 20% of your portfolio allocated to it will kill the TQQQ component. So the question is do you have an exit criteria that doesn't amount to market timing? The market will price in a vaccine or economies growing before you hear about it. The way I see this going is when volatility starts to decline maintaining this allocation wipes out all the gains from TQQQ and then some by the time you realize it's time to adjust the allocation you've missed out on a lot. I would probably set up a purely objective adaptive allocation strategy to decrease the TVIX allocation (e.g. reduce it to 10% once volatility falls below x).

I'm experimenting with a similar portfolio, but I did 85% TQQQ / 15% VIXY, because I think VIX is already volatile enough without the 2x leverage and associated volatility decay (the mean reversion property of volatility means decay due to leverage is more of a factor than with leveraging assets that grow). This wouldn't have prevented a large drawdown in March, but it mitigates it enough for my liking while not being a huge drag on the portfolio the rest of the time. This seems to backtest okay as a static allocation, but I'll probably move to an adaptive one to reduce the volatility, however I'm not sure how best to do that yet since typical AA strategies don't seem to work that well with volatility ETFs.

Also, I think that using call options on VIX the way ^VXTH does is probably a better approach, since that avoids decay due to the futures roll and expense ratio, gives more flexibility with the amount of leverage, and is much more friendly in a taxable account (more leverage = less buying and selling, plus 1256 tax treatment).
Topic Author
differentidea
Posts: 7
Joined: Fri Jun 19, 2020 3:19 pm

Re: TVIX/TQQQ combo

Post by differentidea »

TVIX loses over 90% of its value most years and having 20% of your portfolio allocated to it will kill the TQQQ component.
Yes it does, but this is a strategy I consider for the pandemic and the potential economic fallout from it.

I am not sure it kills the TQQQ component in this environment with a 20% allocation. Take for instance Mon, Jun 15 9:30 AM-Tue to Jun 16 9:30 AM, TVIX lost 29.3% of it's value TQQQ gained 15.5%. With 80/20 allocation, making it a gain of 6.2% total vs 4.8% QQQ or 5.8% SPY

Thing I have found so far is that this absolutely protects during large drawdowns:
June 10-11 TVIX up 65.8% TQQQ down 14.7%. With 80/20 allocation gain 1.3% vs 4.9% loss QQQ or 5.8%loss spy

Now a backtest for the bullrun from June 20th 2019 - Dec 31 2019:
TVIX lost 72.9% TQQQ up 36%. With 80/20 allocation gain 14.2% vs 12.5% QQQ or 8.8% SPY

Where this really under performs is a flat market with few moves. But it is my belief that will not happen for the next few months to a year.
June 20-Oct 14. TVIX lost 37.4% TQQQ 0%. With 80/20 allocation 7.4% loss vs 1.1% QQQ gain or 0% SPY
I would probably set up a purely objective adaptive allocation strategy to decrease the TVIX allocation (e.g. reduce it to 10% once volatility falls below x).
Which is exactly what I would do, evaluate consistently and when it falls below a certain performance level, switch back to more traditional investments.
Also, I think that using call options on VIX the way ^VXTH does is probably a better approach, since that avoids decay due to the futures roll and expense ratio, gives more flexibility with the amount of leverage, and is much more friendly in a taxable account (more leverage = less buying and selling, plus 1256 tax treatment).
Great idea, I'll look into it.
Topic Author
differentidea
Posts: 7
Joined: Fri Jun 19, 2020 3:19 pm

Re: TVIX/TQQQ combo

Post by differentidea »

CardioMD wrote: Fri Jun 19, 2020 4:09 pm
differentidea wrote: Fri Jun 19, 2020 3:26 pm For the foreseeable period of time I think

Yesterday and today this worked quiet well.
How long is the foreseeable future? When the foreseeable future comes to pass what will you do then?

Edit to add: volatility is your friend not your enemy.
Evaluate the performance of it on a weekly level.

Of course it is. That's what I like about this.
Fortune Seeker
Posts: 107
Joined: Mon Apr 13, 2020 9:52 am
Location: UK

Re: TVIX/TQQQ combo

Post by Fortune Seeker »

How did this strategy perform in 2000-2003? Did you model it?
Topic Author
differentidea
Posts: 7
Joined: Fri Jun 19, 2020 3:19 pm

Re: TVIX/TQQQ combo

Post by differentidea »

Fortune Seeker wrote: Fri Jun 19, 2020 6:52 pm How did this strategy perform in 2000-2003? Did you model it?
Not applicable, sars didn't affect the us or the entire world like the pandemic now. Tvix only goes back to 2010.
Fortune Seeker
Posts: 107
Joined: Mon Apr 13, 2020 9:52 am
Location: UK

Re: TVIX/TQQQ combo

Post by Fortune Seeker »

I'm not talking about SARS epidemic. I am talking about situation where NASDAQ 100 would drop sharpy while VIX would only increase moderately instead of skyrocketing to 80.

That's one of weak spots in this strategy - VIX hedges different index.
Fortune Seeker
Posts: 107
Joined: Mon Apr 13, 2020 9:52 am
Location: UK

Re: TVIX/TQQQ combo

Post by Fortune Seeker »

You can roughly model TVIX using VIX which has historic data from about 90.
Semantics
Posts: 163
Joined: Tue Mar 10, 2020 1:42 pm

Re: TVIX/TQQQ combo

Post by Semantics »

differentidea wrote: Fri Jun 19, 2020 6:23 pm
TVIX loses over 90% of its value most years and having 20% of your portfolio allocated to it will kill the TQQQ component.
Yes it does, but this is a strategy I consider for the pandemic and the potential economic fallout from it.

I am not sure it kills the TQQQ component in this environment with a 20% allocation. Take for instance Mon, Jun 15 9:30 AM-Tue to Jun 16 9:30 AM, TVIX lost 29.3% of it's value TQQQ gained 15.5%. With 80/20 allocation, making it a gain of 6.2% total vs 4.8% QQQ or 5.8% SPY

Thing I have found so far is that this absolutely protects during large drawdowns:
June 10-11 TVIX up 65.8% TQQQ down 14.7%. With 80/20 allocation gain 1.3% vs 4.9% loss QQQ or 5.8%loss spy

Now a backtest for the bullrun from June 20th 2019 - Dec 31 2019:
TVIX lost 72.9% TQQQ up 36%. With 80/20 allocation gain 14.2% vs 12.5% QQQ or 8.8% SPY

Where this really under performs is a flat market with few moves. But it is my belief that will not happen for the next few months to a year.
June 20-Oct 14. TVIX lost 37.4% TQQQ 0%. With 80/20 allocation 7.4% loss vs 1.1% QQQ gain or 0% SPY
Makes sense, seeing those numbers is helpful. Looks like I'd sleep a bit easier at night with your allocation in this environment. It was looking at the PV chart in the 2014-2019 timeframe that scared me, but you're right that simply adjusting the allocation when we hit a more stable volatility level would take care of that.
Fortune Seeker wrote: Fri Jun 19, 2020 6:52 pm How did this strategy perform in 2000-2003? Did you model it?
Modeling it with ^VIX it gets wiped out and takes 10 years to recover to the starting point. I've found it pretty hard to come up with stuff that works well with TQQQ in 2000-2003, since QQQ got destroyed over such a long period of time that other components of the portfolio would just keep getting rebalanced into it and decimated. I think it makes sense to just move into less leveraged positions when volatility is super high, to make it easier to find suitable hedges.
Fortune Seeker wrote: Fri Jun 19, 2020 9:24 pm That's one of weak spots in this strategy - VIX hedges different index.
Good point. There is VXN, which has futures (no options), so in theory one could use that.
Topic Author
differentidea
Posts: 7
Joined: Fri Jun 19, 2020 3:19 pm

Re: TVIX/TQQQ combo

Post by differentidea »

Fortune Seeker wrote: Fri Jun 19, 2020 9:24 pm I'm not talking about SARS epidemic. I am talking about situation where NASDAQ 100 would drop sharpy while VIX would only increase moderately instead of skyrocketing to 80.

That's one of weak spots in this strategy - VIX hedges different index.
Understood, that's where I would want to have exposure to the SP500 with the 25/55/20 mix. But looking at the backtest the sp500 wasn't doing that well as during that time. Any thoughts on that?
Topic Author
differentidea
Posts: 7
Joined: Fri Jun 19, 2020 3:19 pm

Re: TVIX/TQQQ combo

Post by differentidea »

Semantics wrote: Fri Jun 19, 2020 10:08 pm
differentidea wrote: Fri Jun 19, 2020 6:23 pm
TVIX loses over 90% of its value most years and having 20% of your portfolio allocated to it will kill the TQQQ component.
Yes it does, but this is a strategy I consider for the pandemic and the potential economic fallout from it.

I am not sure it kills the TQQQ component in this environment with a 20% allocation. Take for instance Mon, Jun 15 9:30 AM-Tue to Jun 16 9:30 AM, TVIX lost 29.3% of it's value TQQQ gained 15.5%. With 80/20 allocation, making it a gain of 6.2% total vs 4.8% QQQ or 5.8% SPY

Thing I have found so far is that this absolutely protects during large drawdowns:
June 10-11 TVIX up 65.8% TQQQ down 14.7%. With 80/20 allocation gain 1.3% vs 4.9% loss QQQ or 5.8%loss spy

Now a backtest for the bullrun from June 20th 2019 - Dec 31 2019:
TVIX lost 72.9% TQQQ up 36%. With 80/20 allocation gain 14.2% vs 12.5% QQQ or 8.8% SPY

Where this really under performs is a flat market with few moves. But it is my belief that will not happen for the next few months to a year.
June 20-Oct 14. TVIX lost 37.4% TQQQ 0%. With 80/20 allocation 7.4% loss vs 1.1% QQQ gain or 0% SPY
Makes sense, seeing those numbers is helpful. Looks like I'd sleep a bit easier at night with your allocation in this environment. It was looking at the PV chart in the 2014-2019 timeframe that scared me, but you're right that simply adjusting the allocation when we hit a more stable volatility level would take care of that.
Fortune Seeker wrote: Fri Jun 19, 2020 6:52 pm How did this strategy perform in 2000-2003? Did you model it?
Modeling it with ^VIX it gets wiped out and takes 10 years to recover to the starting point. I've found it pretty hard to come up with stuff that works well with TQQQ in 2000-2003, since QQQ got destroyed over such a long period of time that other components of the portfolio would just keep getting rebalanced into it and decimated. I think it makes sense to just move into less leveraged positions when volatility is super high, to make it easier to find suitable hedges.
Fortune Seeker wrote: Fri Jun 19, 2020 9:24 pm That's one of weak spots in this strategy - VIX hedges different index.
Good point. There is VXN, which has futures (no options), so in theory one could use that.
What would be the advantage of using the VXN?

I'm trying to see if I can just use this strategy on a rolling basis for the next few months, if it starts to under-perform for more than lets say 2-3 weeks in a row. For now there it appears that it has performed well over the past few months in this environment.
Semantics
Posts: 163
Joined: Tue Mar 10, 2020 1:42 pm

Re: TVIX/TQQQ combo

Post by Semantics »

differentidea wrote: Fri Jun 19, 2020 11:03 pm
Semantics wrote: Fri Jun 19, 2020 10:08 pm
differentidea wrote: Fri Jun 19, 2020 6:23 pm
TVIX loses over 90% of its value most years and having 20% of your portfolio allocated to it will kill the TQQQ component.
Yes it does, but this is a strategy I consider for the pandemic and the potential economic fallout from it.

I am not sure it kills the TQQQ component in this environment with a 20% allocation. Take for instance Mon, Jun 15 9:30 AM-Tue to Jun 16 9:30 AM, TVIX lost 29.3% of it's value TQQQ gained 15.5%. With 80/20 allocation, making it a gain of 6.2% total vs 4.8% QQQ or 5.8% SPY

Thing I have found so far is that this absolutely protects during large drawdowns:
June 10-11 TVIX up 65.8% TQQQ down 14.7%. With 80/20 allocation gain 1.3% vs 4.9% loss QQQ or 5.8%loss spy

Now a backtest for the bullrun from June 20th 2019 - Dec 31 2019:
TVIX lost 72.9% TQQQ up 36%. With 80/20 allocation gain 14.2% vs 12.5% QQQ or 8.8% SPY

Where this really under performs is a flat market with few moves. But it is my belief that will not happen for the next few months to a year.
June 20-Oct 14. TVIX lost 37.4% TQQQ 0%. With 80/20 allocation 7.4% loss vs 1.1% QQQ gain or 0% SPY
Makes sense, seeing those numbers is helpful. Looks like I'd sleep a bit easier at night with your allocation in this environment. It was looking at the PV chart in the 2014-2019 timeframe that scared me, but you're right that simply adjusting the allocation when we hit a more stable volatility level would take care of that.
Fortune Seeker wrote: Fri Jun 19, 2020 6:52 pm How did this strategy perform in 2000-2003? Did you model it?
Modeling it with ^VIX it gets wiped out and takes 10 years to recover to the starting point. I've found it pretty hard to come up with stuff that works well with TQQQ in 2000-2003, since QQQ got destroyed over such a long period of time that other components of the portfolio would just keep getting rebalanced into it and decimated. I think it makes sense to just move into less leveraged positions when volatility is super high, to make it easier to find suitable hedges.
Fortune Seeker wrote: Fri Jun 19, 2020 9:24 pm That's one of weak spots in this strategy - VIX hedges different index.
Good point. There is VXN, which has futures (no options), so in theory one could use that.
What would be the advantage of using the VXN?
VXN is the volatility of the companies in the Nasdaq-100 (QQQ), so it should have a stronger (negative) correlation and therefore be a better hedge. But since most of the top companies are the same in both indices I doubt this makes any difference in practice. That was just a theoretical comment, personally I'd still use VIX.
occambogle
Posts: 580
Joined: Thu Dec 12, 2019 4:58 am

Re: TVIX/TQQQ combo

Post by occambogle »

Why only use VIX? Seems better to use TMF and some VIX-based "protection":

1 Year, see portfolio 3

10 Year, see portfolio 3
User avatar
coingaroo
Posts: 93
Joined: Fri Apr 26, 2019 11:31 am

Re: TVIX/TQQQ combo

Post by coingaroo »

I love TQQQ and I love volatility plays. I would not TVIX/TQQQ.

As Semantics mentioned, an approach similar to VXTH is likely the way to go. Buy deep OTM VIX calls, depending on the forward value of VIX.

VXTH for example:

* spends 1% of the portfolio buying deep OTM VIX calls when forward VIX is between 15 and 30
* spends 0.5% of the portfolio buying deep OTM VIX calls when forward VIX is between 30 and 50
* has no VIX calls otherwise

This has the benefit of actually CAPTURING VIX spikes, i.e. when VIX spikes to 70, doesn't waste money buying OTM VIX calls and instead sells the VIX calls for an incredibly juicy profit ;)

Since you have a QQQ bias and want to apply leverage, I would recommend constructing your own VXTH portfolio however with a mix of QQQ and TQQQ as the underlying, depending on your level of exposure (e.g. 50% TQQQ, 50% QQQ for 2x effective QQQ exposure).

Yes, you gotta buy options yourself, but your performance is probably going to be... insane.
Topic Author
differentidea
Posts: 7
Joined: Fri Jun 19, 2020 3:19 pm

Re: TVIX/TQQQ combo

Post by differentidea »

coingaroo wrote: Sat Jun 20, 2020 8:07 am I love TQQQ and I love volatility plays. I would not TVIX/TQQQ.

As Semantics mentioned, an approach similar to VXTH is likely the way to go. Buy deep OTM VIX calls, depending on the forward value of VIX.

VXTH for example:

* spends 1% of the portfolio buying deep OTM VIX calls when forward VIX is between 15 and 30
* spends 0.5% of the portfolio buying deep OTM VIX calls when forward VIX is between 30 and 50
* has no VIX calls otherwise

This has the benefit of actually CAPTURING VIX spikes, i.e. when VIX spikes to 70, doesn't waste money buying OTM VIX calls and instead sells the VIX calls for an incredibly juicy profit ;)

Since you have a QQQ bias and want to apply leverage, I would recommend constructing your own VXTH portfolio however with a mix of QQQ and TQQQ as the underlying, depending on your level of exposure (e.g. 50% TQQQ, 50% QQQ for 2x effective QQQ exposure).

Yes, you gotta buy options yourself, but your performance is probably going to be... insane.
Thank you for the response. So I can do this with my non 401k portfolio, but my 401k does not allow options trading. Any idea on how to approach this?
occambogle
Posts: 580
Joined: Thu Dec 12, 2019 4:58 am

Re: TVIX/TQQQ combo

Post by occambogle »

FYI TVIX seemingly is being delisted:

https://www.reuters.com/article/us-cred ... SKBN23T2BI
Post Reply