Canadian Money Forum banner

1 - 20 of 35 Posts

·
Registered
Joined
·
17,751 Posts
Discussion Starter #1
I'm experimenting with 'put protection' on a long S&P 500 position using TAIL to soften the drops/crashes in the S&P 500 while keeping upside. In the past, I've occasionally bought index puts but it never worked out, mainly because they kept expiring and I never owned them when I really needed them. I'm hopeful that TAIL will help by outsourcing the management of the portfolio of puts.

I'm counting my combined SPY+TAIL position (or ZSP+TAIL) as effectively "100% stocks". What's described here is within the scope of just my US equity allocation. Separately, I hold Canadian stocks, bonds, gold, which make up my asset allocation. There is no change to my asset allocation.

See this thread for earlier discussion, and thanks to @Topo for the great insights and ideas: Stock investing with black swan protection

Here is the strategy I am trying, at least to start with, with ideas from @Topo

My targets are 80% SPY and 20% TAIL. It's a minimal hedge, because I don't know if this will work. I will maintain these weights by occasionally rebalancing to replenish the TAIL weight every 6 months. During "bullish" times, TAIL will seem like a waste of money, and keep dropping in value. Nobody wants insurance when things are great.

However, in a down market, TAIL should rise in price and offset the drop in SPY. I will watch for conditions such as a sharp rise in VIX and minimum 10% drop in SPY. Then I will aggressively rebalance from TAIL back to SPY to get back to targets. If there is a very serious or prolonged market drop, there may be the opportunity for more than a single rebalancing.
 

·
Registered
Joined
·
17,751 Posts
Discussion Starter #2
Now logging the trades, using my fill prices

2020-06-23
SPY=313.02 weighted 80%
TAIL=22.149 weighted 20%
 

·
Registered
Joined
·
1,082 Posts
I like the idea. TAIL is interesting because it is about 80% US treasuries so it should be able to hold up ok in market rises too. If this works as well as having 40% in bonds I may look at too. Will that weighting be enough to offset losses for the other 80%?

I am looking for something too other than bonds. Something like insurance even.

You can insure property and your car for <1%. You would think there would be something similar for investments. Would gladly pay even .5% -.75% a year to cover losses like the recent 35% draw down.
 

·
Registered
Joined
·
17,751 Posts
Discussion Starter #4 (Edited)
You can insure property and your car for <1%. You would think there would be something similar for investments. Would gladly pay even .5% -.75% a year to cover losses like the recent 35% draw down.
Well that's the big question. Over time, what is the effective cost (how much performance loss) and how much downside protection do you get as a result?

I mentioned this briefly in the black swan thread, but I also have another method (somewhat active) which tries to protect against sharp declines. It has worked over the years, protecting me against some downside. As expected, it's had a long-term performance cost.

I'd like to experiment and see if the TAIL method could work better.
 

·
Registered
Joined
·
1,082 Posts
Well that's the big question. Over time, what is the effective cost (how much performance loss) and how much downside protection do you get as a result?

I mentioned this briefly in the black swan thread, but I also have another method (somewhat active) which tries to protect against sharp declines. It has worked over the years, protecting me against some downside. As expected, it's had a long-term performance cost.

I'd like to experiment and see if the TAIL method could work better.
I am interested too. Wish there was a CDN equiv as have no US account. Will the 20% be enough? ie will the options gain enough to offset the say 35% drawdown on the other 80%?
 

·
Registered
Joined
·
17,751 Posts
Discussion Starter #6
I am interested too. Wish there was a CDN equiv as have no US account. Will the 20% be enough? ie will the options gain enough to offset the say 35% drawdown on the other 80%?
With the 20% weight, no they won't completely offset the losses. I expect they will just soften the decline.

I think Topo estimated earlier that a 50% weight (adding TAIL amount equal to existing stocks) could completely eliminate the drop.

But since this is just an experimental method, I'm starting with the lightweight 20% to see how it plays out. If it's looking good, I may increase that over time.
 

·
Registered
Joined
·
1,082 Posts
With the 20% weight, no they won't completely offset the losses. I expect they will just soften the decline.

I think Topo estimated earlier that a 50% weight (adding TAIL amount equal to existing stocks) could completely eliminate the drop.

But since this is just an experimental method, I'm starting with the lightweight 20% to see how it plays out. If it's looking good, I may increase that over time.
Interesting. Wish they just had an ETF on just put options. Question . Say in a market decline of say 35%, how much would the price of the put options go up, ~ 20x that? They seem to cost ~ 5% of the underlying asset.

Another idea is VIX options that did go up ~ 400%. So if you had 10% in VIX options you would be covered. Problem is they lose $ in good years. may look at putting just 5% in a VIX ETF (when things get back to normal and it is $15 too as protection and cashing it in in crashes ( when it went up to $80)
 

·
Registered
Joined
·
996 Posts
TAIL seems to pay a small dividend of about 1.19%. I don't know if it is ROC or interest from the treasuries it keeps.
 

·
Registered
Joined
·
1,082 Posts
Here is some food for thought too. The VIX is much more volatile than the market and -ve ly correlated . So you can get away w a smaller weighting (20%) that will offset the market drops of SPY ( w 80% weighting). You can see the VIX (blue) spikes in 2009, 2019 and recently vs SPY )black)


20276

Here is a plot of 20/80 VIX/SPY (portfolio 1) vs 100% SPY since 1994. You can see it greatly reduced the impact of 2009 and is doing really well in the current crisis.

20277


The performance is much better too P1 on top. I like the idea that P1 is -.05 correlated to the market too. Interesting and will look at further.

20278
 

·
Registered
Joined
·
12,806 Posts
VIX index is not investable. Best you can get is something like VXX or UVXY (which have enormous decay due to futures roll yield) or something like this TAIL ETF which doesn't really track VIX but might be somewhat correlated.
 

·
Registered
Joined
·
17,751 Posts
Discussion Starter #11 (Edited)
Here is some food for thought too. The VIX is much more volatile than the market and -ve ly correlated . So you can get away w a smaller weighting (20%) that will offset the market drops of SPY ( w 80% weighting). You can see the VIX (blue) spikes in 2009, 2019 and recently vs SPY )black)
. . .
Here is a plot of 20/80 VIX/SPY (portfolio 1) vs 100% SPY since 1994. You can see it greatly reduced the impact of 2009 and is doing really well in the current crisis.
You hit the nail on the head Jimmy. This is the ideal thing I'm going after, and I didn't know Portfolio Visualizer could use VIX as an input!

But I'll do you one better, using what @Topo said about aggressive rebalancing. Change to monthly rebalancing, so you can "harvest" the effect of VIX soaring in a bad market, and watch what happens. It's unbelievable. Here's a link to that simulation.

While SPY returned 9.37% CAGR with sortino=0.75, the 20/80 VIX/SPY with monthly rebalancing returned an eye-popping 13.86% CAGR with sortino=2.18 ... plus it virtually eliminates the major negative years.
20279


As andrewf says, one can't invest directly in the VIX. But if you can invest in something similar, perhaps it's possible to pull off the result that's pictured.

This is what I'm trying to do with TAIL. I don't expect it to work out nearly as well as this backtest, but I would be happy with similar-to-index performance with milder drawdowns.

Does anyone see a better vehicle to use? The only other one which looks feasible to me is VXX but I don't ever invest in ETNs due to the counterparty risk. They don't hold any assets in their funds.
 

·
Registered
Joined
·
1,082 Posts
VIX index is not investable. Best you can get is something like VXX or UVXY (which have enormous decay due to futures roll yield) or something like this TAIL ETF which doesn't really track VIX but might be somewhat correlated.
Yes you can actually . Horizon's have an ETF called HUV Betapro Short term VIX futures.
 

·
Registered
Joined
·
1,082 Posts
You hit the nail on the head Jimmy. This is the ideal thing I'm going after, and I didn't know Portfolio Visualizer could use VIX as an input!

But I'll do you one better, using what @Topo said about aggressive rebalancing. Change to monthly rebalancing, so you can "harvest" the effect of VIX soaring in a bad market, and watch what happens. It's unbelievable. Here's a link to that simulation.

While SPY returned 9.37% CAGR with sortino=0.75, the 20/80 VIX/SPY with monthly rebalancing returned an eye-popping 13.86% CAGR with sortino=2.18 ... plus it just breezes through bear markets.
View attachment 20279

As andrewf says, one can't invest directly in the VIX. But if you can invest in something similar, perhaps it's possible to pull off the result that's pictured.

This is what I'm trying to do with TAIL. I'm hoping to get a result that's kind of like this chart.
First good news. Actually Horizons have an ETF called HUV - Betapro Short term VIX futures that tracks the VIX futures ( similar to VIX).


That is great. Nice work. I had concerns about holding VIX all the time as it has ridiculous volatility - read a few articles not recommending it as a LT hold. I was thinking of getting some at the bottom then selling when it spiked (in severe downturns) but rebalancing seems to work even better.

This maybe the answer to diversify. VIX is also -.66 correlation too - much better than bonds.
 

·
Registered
Joined
·
17,751 Posts
Discussion Starter #14
Yes you can actually . Horizon's have an ETF called HUV Betapro Short term VIX futures.
These things have incredibly bad decay. The 5 year return of HUV is -86% whereas VIX is up +155% in the same period. It's even worse with something like TVIX which returned -99.92%

I don't see how HUV or any of the other similar ETFs are feasible ways to do this, given how badly their prices decay over time.
 

·
Registered
Joined
·
1,082 Posts
These things have incredibly bad decay. The 5 year return of HUV is -86% whereas VIX is up +155% in the same period. It's even worse with something like TVIX which returned -99.92%

I don't see how HUV or any of the other similar ETFs are feasible ways to do this, given how badly their prices decay over time.
That is too bad. Late here and tired but will do more research on HUV or other VIX securities
 

·
Registered
Joined
·
12,806 Posts
Yes you can actually . Horizon's have an ETF called HUV Betapro Short term VIX futures.
HUV, as you will see, does not replicate the index. Just look at the performance. VIX is mean reverting, HUV is marching to zero. It tracks a constant duration VIX futures portolfio, which is a very different animal. Just plug in HUV to your backtest and see the difference.
 

·
Registered
Joined
·
1,082 Posts
HUV, as you will see, does not replicate the index. Just look at the performance. VIX is mean reverting, HUV is marching to zero. It tracks a constant duration VIX futures portolfio, which is a very different animal. Just plug in HUV to your backtest and see the difference.
Thanks. James noted that too. Very unfortunate.
 

·
Registered
Joined
·
17,751 Posts
Discussion Starter #18
This is definitely one of those "theory" vs "practice" kinds of issues. That back test shows amazing results, if it was just possible to actually track VIX, without decay.
 

·
Registered
Joined
·
12,806 Posts
An instrument that tracked VIX perfectly would be a license to print money. Probably why it doesn't exist (or would get arbitraged away).
 

·
Registered
Joined
·
17,751 Posts
Discussion Starter #20
An instrument that tracked VIX perfectly would be a license to print money. Probably why it doesn't exist (or would get arbitraged away).
Right. Additionally, if anyone else figured out the right kind of financial engineering to pull off what we're describing, they probably would keep it secret. If someone can actually demonstrate this kind of superior risk adjusted return, with such minimal effort, and if it becomes well known, just the knowledge being out there would itself change the operation of the market.
 
1 - 20 of 35 Posts
Top