PP with a XIV Twist

General Discussion on the Permanent Portfolio Strategy

Moderator: Global Moderator

Post Reply
Kbg
Executive Member
Executive Member
Posts: 2815
Joined: Fri May 23, 2014 4:18 pm

PP with a XIV Twist

Post by Kbg » Fri May 23, 2014 5:03 pm

Thanks to this great forum and the book that introduced me to PP. For the past couple of years I have used a variation of PP which includes adding XIV and putting all the assets at 20%. One has to rebalance more often however. For those who are PP fans and have an in-depth understanding of volatility instruments might be worth your time to look into it.

To be very clear...if you are new to the markets or do not have reasonable backtesting experience just ignore this post. For those who are not in the inexperienced category I found both 2008 and 2011 to be quite persuasive. You will need simulated XIV data however for 2008.
D1984
Executive Member
Executive Member
Posts: 730
Joined: Tue Aug 16, 2011 7:23 pm

Re: PP with a XIV Twist

Post by D1984 » Fri May 23, 2014 8:41 pm

I do have the backtested XIV data (from SixFigureInvesting; I understand Kuchita sells it now as well) but looking over it again leaves me with a couple of questions/concerns regarding its use in the PP:

One, why XIV and not SVXY? Wouldn't an ETF not have the counterparty risk of an ETN?

Two, how did XIV help in 2008 and 2011? It got killed during both those years; IIRC in 2008 it was down almost 96% peak-to-trough before recovering a bit towards the end of the year. I know the LTTs in a PP portfolio would have helped offset this somewhat but clearly to gain any benefit in 2008 or 2011 (or any other year when volatility spiked) from adding this fund to a PP you would have to have far tighter bands and rebalance far more often...even 20% bands for each of the 5 components (stocks, LTTs, gold, cash, and the inverse VIX product), presumably rebalanced when any component hit 10 or 30% (or with tighter bands, when it hit 15% or 25%) might not be enough. Certainly this wouldn't be something that would've helped you in 2008 if you'd just bought it (and the other four PP components) at 20% each on 1-1-2008 and held and then rebalanced back to 20/20/20/20/20 on 12-31-2008 (it would've been good in 2009, '10 , '12, and '13, though) whereas doing a traditional 25/25/25/25 PP and then only rebalancing at the end of the year would've still worked out just fine and would've resulted in (as Craig's post on CrwlingRoad put it) "2008 - A Disaster Averted".

Three, there aren't simulated values for XIV back to the 1990s (the underlying futures themselves only go back to early 2004) but given what the VIX did in 1996, 1997, and much of 1998 it's quite possible that an inverse volatility product would actually lose money during those years even when the US TSM overall was up over 20% per year (had it existed then XIV would probably have lost money each year from 2000 to 2002 as well but since stocks themselves did poorly and volatility spiked--thanks to the tech crash, the recession, 9/11, and the fallout from the accounting scandals and Enron/Worldcom/etc--then XIV would have been doing exactly what it could have been predicted to do and merely acting as a supercharged version of whatever stocks were returning whether that be negative or positive). IOW, if anyone is thinking of adding this to the PP due to it acting like an even more leveraged bull market ETF than a 3X one, well...buyer beware...it would work most of the time but fail miserably if we ever have another period like 96-98 where volatility and market levels move in opposite directions for extended periods, or where contango and backwardation don't exactly end up happening as they would be expected to.

Four, two words: October 1987.

Finally, most of the people on this board believe (I know Craig and MT do, for starters) that you should own the assets as directly as possible when this is feasible (sometimes it isn't and then you have to make do with what ETFs or mutual funds are available). For gold, that means gold coins/bars and not ETFs like GLD. For cash and LTTs, it means actual T-bills/T-bonds at TreasuryDirect and not funds like VUSTX/TLT in place of LTTs or SHY or bank CDs in place of t-bills. Stock ownership is about the only thing that is advocated as being OK to own through indirectly through mutual funds and ETFs but that's only because it would be all but impossible and far too expensive (in trading commissions to buy and then to rebalance as the index changed) to own the total stock market (or even just the S&P 500) company-by-company unless you had many millions of $ to invest. I am not saying adding XIV is bad per se (I'm not nearly as pro "pure PP" as most of the rest of the posters here) but the idea may get a fairly chilly reception from most on this forum.
Kbg
Executive Member
Executive Member
Posts: 2815
Joined: Fri May 23, 2014 4:18 pm

Re: PP with a XIV Twist

Post by Kbg » Sun May 25, 2014 11:25 am

Great question on XIV vs SVXY. To me it comes down to tax treatment vs. Counterparty risk...and a slightly tighter spread. To each his/her own preferences and financial situation. Caveat Emptor for sure.

If you are going to throw XIV/SVXY into the mix, clearly you...maybe that is too strong...generally you will want to rebalance more. Volatility trends sort of but generally you want to siphon off volatility profits during the good times because you know the bad times will come.

With regard to poorly performing stretches of time...let's have a debate about gold in the PP. :-)

I fully agree that anyone contemplating short vol should think long and hard about 1987. We have to make a ton of assumptions about 1987 in light of theses products not even existing. Personally I don't think the VIX would go that high now because of theses products but let's assume it would. First of all let's scratch 20% of the port. It's gone. Safe assumption. Then let's assume these products still existed afterward...not so safe assumption. Let's also assume you are faithful to your rebalance...let's call this a neutral assumption. If your rebalance timing was absolutely perfect (the peak) then by the end of 1998 your short vol position is up somewhere around 10x.

The whole prior paragraph is pure historical fiction and speculation. However we do have a real life scenario.  At the 2011 XIV low the XIV was at 4.92 I believe. Friday's close was 37.46. That is about ~760% in three years. The point to the fiction paragraph and this one is that if someone gives you two bets and the first one is a guaranteed 100% loss followed by something that is a multiple of the previous bet, you take it.

I opened this thread  up with all kinds of caveats...and I agree with the response. I just happen to believe this is an interesting extension of wrapping up some crazy volatility into a package that works together. Or, just put this in your personal port with some cash and faithfully rebalance between the short vol and cash.
D1984
Executive Member
Executive Member
Posts: 730
Joined: Tue Aug 16, 2011 7:23 pm

Re: PP with a XIV Twist

Post by D1984 » Mon May 26, 2014 12:15 pm

Kbg wrote: Great question on XIV vs SVXY. To me it comes down to tax treatment vs. Counterparty risk...and a slightly tighter spread. To each his/her own preferences and financial situation. Caveat Emptor for sure.

If you are going to throw XIV/SVXY into the mix, clearly you...maybe that is too strong...generally you will want to rebalance more. Volatility trends sort of but generally you want to siphon off volatility profits during the good times because you know the bad times will come.

With regard to poorly performing stretches of time...let's have a debate about gold in the PP. :-)

I fully agree that anyone contemplating short vol should think long and hard about 1987. We have to make a ton of assumptions about 1987 in light of theses products not even existing. Personally I don't think the VIX would go that high now because of theses products but let's assume it would. First of all let's scratch 20% of the port. It's gone. Safe assumption. Then let's assume these products still existed afterward...not so safe assumption. Let's also assume you are faithful to your rebalance...let's call this a neutral assumption. If your rebalance timing was absolutely perfect (the peak) then by the end of 1998 your short vol position is up somewhere around 10x.

The whole prior paragraph is pure historical fiction and speculation. However we do have a real life scenario.  At the 2011 XIV low the XIV was at 4.92 I believe. Friday's close was 37.46. That is about ~760% in three years. The point to the fiction paragraph and this one is that if someone gives you two bets and the first one is a guaranteed 100% loss followed by something that is a multiple of the previous bet, you take it.

I opened this thread  up with all kinds of caveats...and I agree with the response. I just happen to believe this is an interesting extension of wrapping up some crazy volatility into a package that works together. Or, just put this in your personal port with some cash and faithfully rebalance between the short vol and cash.
With regards to tax treatment I wasn't considering much difference between SVXY and XIV since I would assume this would be done in a tax-sheltered account anyway (IRA, Roth IRA, 401K) since it would be likely be generating short-term capital gains often due to the frequent rebalancing.

Since you mentioned that including XIV/SVXY would result in one wanting to rebalance more, exactly which rebalancing bands (i.e. instead of 35/15) or trend following method would you recommend? Can you show backtested results since 3-26-2004 (start of VIX futures) for the 20/20/20/20/20 PP with your chosen rebalancing specifications (I would love to see how much better it did than the "standard" 4x25 PP) and show how many rebalances there were? Also, what would prevent someone from rebalancing into oblivion (i.e. keep rebalancing into a losing asset over and over and for two or three years in a row it loses money) vis-a-vis a hypothetical XIV in 2007 and 2008?

As far as gold (and its long stretches of underperformance at times) goes, there are actually several members who (moda and melveyr IIRC) who think that if one wishes to get rid of some of the gold (say down to 15% or 20% instead of 25%) and replace it with TIPS that wouldn't be the end of the world. For better or worse, gold does seem to dominate the performance of the PP at times (gold is usually its most volatile asset and the performance of the PP is slightly more correlated to gold than it is to stocks or LTTs); with that said, every asset class (save cash) has had bad years since the PP's inception so it's not just gold (heck, from 2000-2011 stocks basically returned nothing in nominal terms and actually lost money after inflation); my real worry about the "pure" PP going forward is that it will have two assets doing poorly and the one doing good won't be enough to offset it...so maybe adding XIV/SVXY could help with this by adding another asset that will (hopefully) be up more years than it will be down and which due to its greater volatility will provide more "volatility capture" gains opportunities.

Finally, you mentioned that a hypothetical XIV/SVXY would have been up ten times from immediately after the '87 crash to the end of 1998...since VIX futures didn't even exist then (much less the ETF/ETF itself), where are you getting the data from (and can you share it here)? Is it just extrapolated somehow from the VIX itself?
Kbg
Executive Member
Executive Member
Posts: 2815
Joined: Fri May 23, 2014 4:18 pm

Re: PP with a XIV Twist

Post by Kbg » Tue May 27, 2014 7:11 am

Please see a tax advisor. Just extrapolated. There is quite a bit out there with regard to selling volatility. A dose in moderation is generally beneficial. To be sure I think much homework should be done before remotely considering this.
User avatar
MachineGhost
Executive Member
Executive Member
Posts: 10054
Joined: Sat Nov 12, 2011 9:31 am

Re: PP with a XIV Twist

Post by MachineGhost » Tue May 27, 2014 9:23 am

I'm waiting for spot VIX ETF before I even consider it in a portfolio.  The roll and tracking issues just ruin everything else I've tried to profit from.
"All generous minds have a horror of what are commonly called 'Facts'. They are the brute beasts of the intellectual domain." -- Thomas Hobbes

Disclaimer: I am not a broker, dealer, investment advisor, physician, theologian or prophet.  I should not be considered as legally permitted to render such advice!
User avatar
MachineGhost
Executive Member
Executive Member
Posts: 10054
Joined: Sat Nov 12, 2011 9:31 am

Re: PP with a XIV Twist

Post by MachineGhost » Tue May 27, 2014 9:29 am

Kbg wrote: Please see a tax advisor. Just extrapolated. There is quite a bit out there with regard to selling volatility. A dose in moderation is generally beneficial. To be sure I think much homework should be done before remotely considering this.
I think homework should include learning about all the short option writers that got wiped out to the tune of tens of millions in losses owed to creditors back in 1987.
"All generous minds have a horror of what are commonly called 'Facts'. They are the brute beasts of the intellectual domain." -- Thomas Hobbes

Disclaimer: I am not a broker, dealer, investment advisor, physician, theologian or prophet.  I should not be considered as legally permitted to render such advice!
Kbg
Executive Member
Executive Member
Posts: 2815
Joined: Fri May 23, 2014 4:18 pm

Re: PP with a XIV Twist

Post by Kbg » Sat May 31, 2014 8:00 am

I agree. Selling premium is awesome until it isn't; however,  I have zero sympathy for someone who gets wiped out. That's not a short option problem that's a risk control problem. Long term repeated selling of premium is profitable if done correctly. One just has to remember it's really nothing more than selling insurance and the number one rule of an insurance company that wants to stay in business is to make sure it has adequate reserves to cover those times when it has to pay out on the policy written.

With regard to this thread specifically...on a day to day basis it is more profitable to short VXX normally. However, let's do a rerun of 1987. If I'm short VXX I will lose a large multiple of my investment unless I have a really big cash reserve. If I am long XIV I lose my investment. For me, XIV/SVXY is the only way I will do this. I think it very important to have the multiples on your side if contemplating this. 1x losses are manageable, unknown losses at some multiple are not.
Kbg
Executive Member
Executive Member
Posts: 2815
Joined: Fri May 23, 2014 4:18 pm

Re: PP with a XIV Twist

Post by Kbg » Wed Jun 18, 2014 9:55 pm

YTD performance 13.79% (backtest)
- 4.83% 1st Q
- 8.95% 2nd Q

Actual $$$ port 2nd quarter 10.56%...as often happens the XIV portion this quarter is up just shy of 50%
- Tickers DGP, SHY, UBT, UWM, XIV

I will definitely rebalance this quarter.

Full disclosure. The XIV component has also lost 49.66 and 73.11% in the worst two quarters. The best two have been nearly the same at 40.6 and 79%. Normally XIV comes in at about 30 to 34% per quarter.
ILoveMoney
Full Member
Full Member
Posts: 80
Joined: Fri Jan 24, 2014 6:26 pm

Re: PP with a XIV Twist

Post by ILoveMoney » Sat Jul 19, 2014 3:14 pm

D1984 wrote: What would prevent someone from rebalancing into oblivion (i.e. keep rebalancing into a losing asset over and over and for two or three years in a row it loses money) vis-a-vis a hypothetical XIV in 2007 and 2008?
If you are retiring in 20 or 40 years and don't need the money, would it really be a problem to rebalance in a losing asset for a couple of years? You would gain it all back plus much more wouldn't you? (VP play.)

Question to the OP. Would you keep this asset for 20 years straight without reverting it to cash at any time no matter what the economic circumstances are?
frommi
Executive Member
Executive Member
Posts: 189
Joined: Mon Mar 25, 2013 1:04 pm

Re: PP with a XIV Twist

Post by frommi » Sun Jul 20, 2014 6:32 am

I really wanted to destroy this allocation and did a backtest with monthly rebalancing for 2007 and 2008. You would have lost around 50-60%, so not a lot more than a pure equity strategy. But since 1/1/2007 you would have a CAGR of 16.9%, compared to a CAGR of 6.1% with SPY and around 14% with a 2xPP. So this is the first asset allocation approach that i start to like and that has returns that are comparable with value investing strategies. This is the first approach where rebalancing really makes sense because you harvest a lot of volatility. Of course there is counterparty risk, but you can minimize that by chosing different counterparties for each component.

But i am not sure if the current starting date is a wise idea, i probably try this after a huge spike in vol.

But nontheless, thank you Kbg for your idea!
Kbg
Executive Member
Executive Member
Posts: 2815
Joined: Fri May 23, 2014 4:18 pm

Re: PP with a XIV Twist

Post by Kbg » Sun Jul 20, 2014 10:09 pm

No problem. One can use timing on the VIX or use ZIV to mellow things out as well if the volatility of volatility is too intense. For those not wanting something high maintenance with a less bucky ride I would recommend ZIV vs. timing.  In any event...I've left lots of warnings on walking this path. Don't even think of doing it without serious due diligence.
D1984
Executive Member
Executive Member
Posts: 730
Joined: Tue Aug 16, 2011 7:23 pm

Re: PP with a XIV Twist

Post by D1984 » Mon Jul 21, 2014 8:15 pm

ILoveMoney wrote:
D1984 wrote: What would prevent someone from rebalancing into oblivion (i.e. keep rebalancing into a losing asset over and over and for two or three years in a row it loses money) vis-a-vis a hypothetical XIV in 2007 and 2008?
If you are retiring in 20 or 40 years and don't need the money, would it really be a problem to rebalance in a losing asset for a couple of years? You would gain it all back plus much more wouldn't you? (VP play.)
The issue here is that what if the time you happen to "need the money" comes at the worst possible time (i.e. right before retirement or during the first few years of retirement when you have started drawing down assets to replace earned income) for a portfolio to contain assets that can fall sharply in value? You have potentially permanently harmed your capital by withdrawing from an asset mix when one of the assets is severely down. It would be like withdrawing from a stock-heavy portfolio in 2000-2002; the "sequence-of-returns" risk factor (the risk that you have to sell assets at the worst possible time--when they are down sharply--and thus the withdrawals hurt your portfolio over the long-term far worse than would happen if the string of bad years comes late in your retirement after a long period of good years) could seriously impair the long-term survivability of your portfolio. Given that frommi has indicated that this strategy (although he used monthly rebalancing and I was actually thinking annual rebalancing instead of monthly might reduce the DD somewhat by not rebalancing into a losing asset so often when/if the asset was in a long drawn-out downturn) could be down as much as 50 or 60% I think this is a risk that shouldn't be taken lightly. I would recommend a significant cash bucket (somewhat more than the standard 25%) if someone wanted to add XIV to the PP and keep it there through his/her retirement years so they would have an asset class to draw from so as not to have to draw anything from XIV when it was sharply down.

Also, this strategy has never been tested (since VIX futures didn't even exist back then) over the years 2000-2002 when a hypothetical XIV or SVXY would have been down sharply three years in a row....the worst we've seen so far for (simulated) XIV/SVXY was from February 2007 to late 2008 or early 2009 (the simulated backtested data I have show an absolute bottom on 11-20-08 but it came down to almost that level again in late March 2009) and it recovered sharply soon after that. Or heck, look at what the VIX did from 1996 to 1998 and imagine how three bad years  (at the same time the actual stock market was going gangbusters) in a row would affect the performance of a portfolio containing XIV. People on this board panic when the "regular" PP loses 2 or 3% during a year the stock market gains over 30% (2013)...what do you think they will do when the PP's already (relatively) weak performance (vs the stock market...not in purely gain/loss terms since the PP did indeed produce positive numbers during those three years) in the mid-1990s? Or what about years like 1981, or 1969 into early 1970 (the VIX didn't exist back then but look at realized 30-day or six-month volatility and you'll see that it was higher than normal during much of those time periods)

If you're wondering where I got the hypothetical guesses for how XIV would have done in 1996 through 1998, it came from two sources. One, look at how the VIX itself did during those years...it rose. Typically XIV goes down on days the VIX rises (with a beta of maybe 0.24 to 0.8 of the VIX's rise but in the opposite direction...i.e. if the VIX rose by 1% then XIV would fall by somewhere between 0.24% and 0.8%). You can plot the simulated XIV--and the real one from when it came out to the present--vs the VIX and while you will find some outliers (and even a few times where the VIX rose slightly but XIV rose a little too) where it rose/fell more than that, the relationship above generally holds true. Two, check out the comment by "Damian" at http://investing.kuchita.com/2011/08/15 ... vix-graph/ and his screenshot of a graph at http://content.screencast.com/users/Str ... dation.png and see how simulated VXX would have done during this time (and since XIV typically does pretty close to the opposite on a daily basis, just invert the results so that "roll cost" is "roll yield" i.e so that times when the dark blue line is above the 0% range are good times for XIV when it gained value and times when it is below it are when XIV would hypothetically have lost value that month. Big spike way below the 0% line in late 97-early 98 (Asian financial crisis???) and autumn of 1998 (Russian financial crisis/default and LTCM failure).

Again, none of this is to say XIV/SVXY are "bad" per se and I think they might very well have a place in a VP, but not necessarily in one that includes mostly or exclusively the PP assets plus XIV. I was thinking maybe a "barbell" of something like anywhere from 5% to 25% XIV/SVXY and the rest in something "anti-volatility" like PHDG or VQT and rebalance it annually. I can post simulated results for that (from 3-26-04 for 2004 and then for every year until 2014 YTD) if you'd like (I have a few other ideas with XIV/SVXY as well that could potentially yield even higher returns if you--or frommi or Kbg or anyone else as well--are interested).
Last edited by D1984 on Mon Jul 21, 2014 8:20 pm, edited 1 time in total.
Kbg
Executive Member
Executive Member
Posts: 2815
Joined: Fri May 23, 2014 4:18 pm

Re: PP with a XIV Twist

Post by Kbg » Tue Jul 22, 2014 10:46 pm

I am interested, post away!
D1984
Executive Member
Executive Member
Posts: 730
Joined: Tue Aug 16, 2011 7:23 pm

Re: PP with a XIV Twist

Post by D1984 » Thu Jul 24, 2014 10:46 pm

Kbg wrote: I am interested, post away!
Ok, first of all the results from SVXY/XIV (or simulated SVXY/XIV before it actually came out as an ETP) "barbelled" with PHDG (again, simulated before it or VQT actually came out). I would recommend blends of anywhere from 5% SVXY/XIV and 95% PHDG (for conservative investors) to 20% or even 25% SVXY/XIV and 80% or 75% PHDG for aggressive ones, for those in between use either 8%, 10%, or 15% SVXY with the rest PHDG. Returns are as follows (2004 is not a full year and is only from 3-26-04 to 12-31-2004 since VIX futures themselves only started trading in late March 2004; all other years are from January 1st to December 31st except for 2014 which is from January 1st to YTD); all portfolios are rebalanced annually at the end of December 31st:

For 5% SVXY and 95% PHDG

2004 = 5.02%

2005 = 2.74%

2006 = 16.02%

2007 = 13.12%

2008 = 16.11%

2009 = 26.35%

2010 = 7.30%

2011 = 13.89%

2012 = 10.01%

2013 = 17.71%

2014 = 5.19%


For 25% SVXY and 75% PHDG

2004 = 31.88%

2005 = 16.41%

2006 = 25.08%

2007 = -1.64%

2008 = -3.06%

2009 = 39.74%

2010 = 31.46%

2011 = 1.39%

2012 = 39.56%

2013 = 36.33%

2014 = 10.69%

Obviously, portfolios with SVXY/PHDG at 10/90 or 15/85 or 20/80 will rank between the above portfolios in terms of risk vs return.

I haven't calculated the MaxDD (daily) for the ports above but IIRC SVXY would've lost about 96% (from its peak in early 2007) at its lowest in the late autumn of 2008 and hit its MaxDD a few days from when PHDG would have done so as well (simulated PHDG's MaxDD was a lot lower obviously...it hit maybe 18 or 19% down from its peak and was positive again within a month or so and finished 2008 in the black).

PHDG and SVXY seem to make a pretty good combo as far as risk/return is concerned. Almost immediately, however, I saw one potentially serious flaw....

See, PHDG (and XVZ and VSPY as well) seem to--based on backtested and actual performance--all have anywhere from a minimum of one days' time (as in one day AFTER the market fall starts and volatility spikes) to maybe four or five days at most response time to shift into "volatility-fighting mode" if the market is currently calm; this is OK (or at least tolerable) in a situation like we've had so far in years like 2008, late summer/early autumn 2011, May 2010, September 2001, summer 2002, and even October 1987 and October 1929. In all of the above market crashes/falls, stock prices fell and volatility rose somewhat for several days BEFORE any major slides/crashes in the S&P/NASDAQ/DOW so at least the aforementioned ETPs would theoretically have been at least partially long volatility (or at least in cash in SVXY's case although that depends partly on hypothetical VIX futures prices since those didn't exist before 2004) before the majority of the market slide/fall/crash really got ugly.

But what happens if we get another market fall (and concurrent volatility spike) like late February 2007 that strikes with basically no warning whatsoever? See http://sixfigureinvesting.com/2012/02/a ... ve-missed/ for an example of how XVZ would have failed to protect you at all. Or even worse, what if we get another 1987-style crash (whether just a freak flash crash or whether due to a natural disaster or terrorist attack or even a temporary large scale power outage/blackout) where the market collapses but then rallies sharply and strongly a few days later...in that case, something like XVZ could actually still be in "net long volatility mode" at the worst possible time and after getting killed (or at least not moving much either higher or lower) in the no-warning stock market crash, it would get killed again in the whipsaw volatility crash (i.e. massive fall in the VIX and in volatility) that followed a few days later. Not a particularly appealing scenario.


Of course, one could just have VXX or UVXY as a hedge for these type of situations but then if the market ISN'T crashing (which it isn't most of the time) then your "hedge" gets killed by roll losses and as such might be so little of your portfolio (thanks to losses in its value) by the time the crash rolls around that even if it doubles or triples in value it won't offset much of a fall in the rest of the portfolio....you could theoretically rebalance every month in order to "re-charge" VXX/UVXY back to its original portfolio weight but that is basically just pouring money into a black hole in any month the market doesn't fall sharply and VIX doesn't rise.

So, what is really needed is a product that:

1. Is ALWAYS net long volatility

2. Doesn't suffer from roll losses due to an unfavorable contango or backwardation environment (or at least if it does suffer from said slow day-by-day death-of-a-thousand-cuts losses, they are slight and manageable--and not horrific like VXX's or UVXY's--so that you still have some hedge left when it's needed).

3. Is designed to react DAILY (with no delay) to any volatility spikes/market crashes on the same day they happen.

4. Ideally would have at least slight positive return over the long term; with that said, even if it had flat or even slightly negative real returns over the long term it would still be an OK investment as part of a balanced portfolio and to hedge things like SVXY (or TQQQ, or UVT, or UPRO) if it spiked sharply exactly when it was needed (when everything else was crashing and correlations for every asset class besides long volatility went to nearly 1.0 with equities)...in other words, it would still be acceptable as a portfolio diversifier in much the way that gold or collateralized commodity futures are good portfolio diversifiers (because in the case of gold or CCFs they are non-correlated or negatively correlated to equities in most environments--and especially in ones where inflation is unexpected otherwise stock/bond heavy portfolio can actually raise risk-adjusted returns even if the long-term returns on gold and CCFs are basically no more than inflation itself.


Well, those are the four characteristics I believe a good "when all hell breaks loose in the markets" volatility hedge needs. It's a tall order, but I believe there may be two potential choices that can fulfill it.


The first is a mutual fund (offered by KKM and Stutland) based on their ARMOR index. The ticker symbol is RMRAX / RMRIX (for the retail and institutional versions, respectively). Info on the fund and its underlying index (including performance data) is available at:

http://www.stutland.com/brochure_page/26243789.pdf

http://www.stutland.com/brochure_page/02261994.pdf

http://kkmarmorfunds.com/wp-content/upl ... ochure.pdf

The backtested index was incepted in January 2007 and the Finanical Times website has performance data and charts for it back to that date (said data and charts are based on daily closing prices from Thompson-Reuters and Bloomberg) all the way to the present; weekly closing values are available from 2007 to mid-2011 and daily values are available for the last three years. This index would have spiked in February 2007 (unlike XVZ or SVXY) and thus did its job by going up sharply in the event that volatility rose sharply and suddenly; it would also have done roughly as well as XVZ during 2008 and 2011 but didn't decay nearly as much in 2012 and 2013.

The actual live performance of the positions underlying the index dates back to March 2013 (Stutland has been running a $5 million minimum investment hedge fund using the exact methodology in the ARMOR index since March 2013) and the actual index itself and the mutual funds date to earlier this year (2014).


The pros as I see it for this mutual fund (and its underlying index) are as follows:

1. Meets all four of my criteria above

2. Strong negative correlation on a weekly/monthly/annual basis to SVXY


The negatives are as follows:

1. Only backtested to 2007 so no history from when VIX futures started (early 2004) until January of 2007

2. Is a proprietary index so cannot be backtested on one's own to verify KKM's/Stutland's claim of performance (although it was independently verified according to them) albeit so far it seems to be working just like it's supposed to vis-a-vis risk assets like SVXY and equities....but the backtested returns are positive for every year except for 2009 and are pretty good so for all I know they "curve-fitted" the thing.

3. Is a loaded fund (5.75% load) unless bought directly from KKM and/or its transfer agent.

4. $25,000 minimum investment even on the retail shares

5. Has a 2.5% expense ratio before fee waivers and 1.5% expense ratio even after them.


The other potential choice isn't even a mutual fund or ETF or otherwise tradeable product (yet); it is merely an index so  (at least for the moment...until and unless the index provider chooses to make an ETP out of it) you'd have to do the actual daily rebalancing yourself. If you are interested, it is the Tail Risk Hedged Volatility Component Index (ticker symbol is TRSKVOL and data/charts on it is available on the Financial Times website and at http://www.velocitycapitalmgt.com/files ... 0513-1.pdf if you want to see monthly performance data)

VelocityShares came up with this index but for some reason known only to them when they decided to create an ETP out of it they chose to blend it 15/85 with the S&P 500 and rebalance monthly rather than just providing a pure volatility hedge ETF like they should have done IMO. The actual methodology of the index underlying said ETF is at http://velocityindices.com/files/smartc ... 130508.pdf (and this paper also includes the methodology underlying the TRSKVOL index itself since that index is a component of the VelocityShares Tail Risk Hedged Large Cap Index TRSKID which in turn is the index the ETF TRSK is itself based on)

With that said, since they helpfully gave us this whitepaper showing how they calculate TRSKVOL, that means the index can be independently backtested and confirmed. VH2 Solutions or Kuchita can probably do the backtested calculations going all the way back to 2004 if you want...shouldn't run more than $150 or so for a spreadsheet including the formulas needed so you can do the calculations yourself from then on in order to be able to do the needed rebalancing between UXVY and SVXY yourself (since this product is not offered as an ETP as of the moment).

The pros I see of this one are as follows:

1. It is transparent, non-proprietary, and and can be backtested (in fact, Vance Harwood at VH2 Solutions would've had to backtest it in order to create his backtested graph and data for TRSK...so he probably already has the data and will sell it to you for the usual $75 without formulas or $150 with)

2. It meets all four criteria for a good volatility hedge and crash hedge (and would be typically be strongly negatively correlated to risk assets like SVXY or equities themselves)....it spiked upwards of 170% in 2008 and went up by double digits in both August and September of 2011


The cons are as follows:

1. Have to do daily or weekly rebalancing yourself (for the long/short vol exposure) and that's in addition to the usual annual rebalancing you'd do with something like this against SVXY or XIV or UPRO; even at somewhere with low-cost trades like Interactive Brokers having to do hundreds of rebalances a year is going to eat up a significant chunk of your capital unless you already had several hundred thousand $$ to start with so that commissions aren't much percentage-wise related to your total portfolio.

2. Likely long-term returns will be flat to slightly negative, probably slightly worse overall than XVZ but nowhere near as bad as VXX. In years like 2008 and 2011 it will soar but then it can come down pretty quickly (but again, nowhere near as badly as VXX!) when volatility settles down again.


If you know of any other good ideas for hedges to use to rebalance against XIV or SVXY, I would certainly be interested in hearing about those as well.
Last edited by D1984 on Thu Jul 24, 2014 10:50 pm, edited 1 time in total.
Kbg
Executive Member
Executive Member
Posts: 2815
Joined: Fri May 23, 2014 4:18 pm

Re: PP with a XIV Twist

Post by Kbg » Sat Jul 26, 2014 8:44 pm

Great post.  I think the performance you posted speaks for itself. If one wants to hedge the SVXY/XIV for blue sky events, I suggest something very simple. Just hedge a certain percent of SVXY with VXX or UVXY and rebalance religiously. It will reduce your performance a bit, but it will protect you from a blow out to zero on your short vol position by several times.

One really needs to think about the VIX component more in terms of a betting strategy than an investment strategy. Again, if something goes to zero but can be hedged or you can rebet and the return is some multiple of what you lost, that is OK. I really do not worry about getting wiped out. I worry about the associated ETF/ETNs not staying open and/or reopening so the second bet is not possible...but one can always short the futures directly if you have to. I just like the fact that the ETFs will only go to zero whereas, who knows on the short VIX futures.
Post Reply