Is Shorting UVXY, TVIX, or VXX the Perfect Trade?


Updated: Apr 12th, 2017 | Vance Harwood | @6_Figure_Invest

The charts for long volatility Exchange Traded Products (ETP) like Barclays’s VXX, VelocityShares’ TVIX, and PowerShares’ UVXY are astonishing.

 

vxx-uvxy

I’m not aware of any other widely available securities that have declined like these.

Two questions come to mind:

  1. Why would anyone invest in these perennial losers?
  2. Why doesn’t everyone on the planet short these funds?

It turns out that there are reasonable reasons to buy these funds, and some people make money doing it. And a lot of people short these funds; it’s a crowded trade—to the point where it’s sometimes not possible to borrow the shares to short them.

It’s not easy money either way.

 
Risks of a Short Position

  • Unlike a short position in most equities fear is not your friend when you are short a volatility fund. When the market gets scared the equity declines are scary, deep, and fast—and volatility spikes up dramatically.
  • One characteristic of a short position is that its leverage moves against you if you’re wrong. When you first open a short position a 10% gain in the stock you shorted will cost you 10% of your position’s value, but the next 10% gain in the security will increase your loss by 12.2%. This increase in leverage increases rapidly if the security moves strongly against you. See this post for more information on this phenomenon.
  • Typically (75%+ of the time) the prices of long volatility funds like VXX, UVXY, and TVIX are battered by contango , but when the market tanks they turn into beasts.  First of all the VIX futures that these funds are based on spike up, second the VIX future’s term structure goes into a configuration called backwardation—which boosts the ETP’s returns, and the 2x funds often experience a compounding effect  that boosts their returns past their 2X benchmark.
  • Long volatility funds have not existed all that long, the first one was introduced in 2009, so we don’t have actual data for the earlier bear markets, but we do have historical data for the 2011 correction, where UVXY’s value went up 550% in a few months. In my simulation of UVXY’s prices that goes back to 2004, I show that that the prices of UVXY would have gone up 15X in the 2008/2009 crash.  Now you can see why some people are interested in going long with these funds.
  • In addition to the risks of typical market corrections and bear markets, a short volatility position is also vulnerable to a Black Swan type event. A major geopolitical event, natural disaster, or terrorist attack could cause a very large, essentially instantaneous jump, in the volatility funds.  The record one-day VIX jump so far was a 59% jump in February 2007, but in this post I postulate that a 100% one-day jump in the VIX is not out of the question. The VIX futures that underlie the volatility ETPs don’t track the VIX moves directly, typically the mix of futures used moves around 45% of the VIX’s percentage move, but with the 2X leveraged funds that still gives a 90% daily jump in their prices.  If an event like this happens when the market is closed there will be no chance for protective measures like stop loss orders to execute.  Even if the event happened during market hours conditions would be chaotic, and the market would likely shut down quickly.

If I haven’t managed to scare you off by now, the next section discusses specifics of initiating a short trade.

 

The Trade

  • These securities are always in the “hard to borrow” category, so it’s very likely at least a phone call to your broker will be required to create a short position. It’s also very likely you’ll have to pay an ongoing fee to borrow the shares.  Plan on the annualized fee being at least 7%.
  • You’ll need to have margin capability setup in a taxable account. Short selling is not allowed in retirement accounts like IRAs or 401Ks.
  • You’ll need extra cash / marginable securities in your account as margin. There are two different amounts required (which can vary by broker and by security), one to initiate the trade and another to maintain your position. The initial percentage will always be greater than or equal to the maintenance position.  Leveraged funds like UVXY and TVIX require extra margin.  The chart below shows Charles Schwab’s requirements for shorting ETFs as of 2-Oct-2016.

schwab-margin

  •  If your trade moves against you to the point that you don’t meet the maintenance requirements you’ll get a margin call from your broker. Not a fun thing. You have two choices at that point, either add more money / marginable securities to your account or reduce your short position by buying back some of the security.  Don’t expect your broker to be patient.

 
Managing a Short Position

  • If you hold a short position it’s critical that you have an exit plan. A few people might have a small enough position and enough margin to weather even the darkest bear markets, but most people won’t have the capital or the temperament to hang in there.  Emotionally it is very difficult to close out a short volatility position with a large loss, not only has your timing been bad, there’s also the near certainty that eventually contango will wear these funds back to levels that would be profitable for you. On the other hand, not having an exit plan raises the very real possibility that your broker will be the one closing out your position, likely at the worst possible time.
  • I haven’t looked extensively at protective strategies, but one thing to look at would be to buy out-of-the-money calls at strike prices much higher than the current trading value of the securities. That way you can limit or mitigate your maximum loss, even during a Black Swan. Essentially you’re buying an insurance policy with a high deductible.  It wouldn’t be cheap because the options would likely be expensive and usually expire worthless, but the peace of mind might very well be worth the cost.
  • One harsh reality of a short position is that while you are exposed to potentially very large losses the best case profit you can realize from your initial position is limited to 100%. For example, if you sell short $1000 worth of VXX your maximum profit can’t be more than $1000 because VXX can’t drop below zero.  And even with the ravages of contango VXX’s split adjusted price will never get all the way to zero.  As the security you short drops in value the percentage leverage of your position drops also, eventually approaching zero.
  • If your short has been successful at some point you’ll need to short additional shares to get your leverage and additional profit potential back up.  I quantify this leverage with a simple formula: Leverage =  P/Po where P is the current price and Po is the price you initially shorted at. Let’s say you are comfortable with a leverage factor between 1 and 0.7.  If it drops to .7 then you would short enough shares to bring your leverage back to 1. So if you were initially short 100 shares at $10, you have a maximum profit potential of $1000 at that point…  If the price has dropped to $7, then your maximum additional profit has dropped to $700 and your leverage has dropped to 0.7.  To get your leverage and additional profit potential back to 1.0 and $1000 your need to short an additional $300 worth of shares (~43 shares).

Alternatives to a Short Position

There are some alternatives strategies that address some of the risks and restrictions of taking a short position.   Of course, they introduce their own limitations, risks, and restrictions.

  • ProShares’s SVXY and VelocityShares’ XIV and ZIV are inverse volatility ETPs that avoid the variable leverage and unconstrained loss aspects of a short position and are allowed in retirement accounts. In exchange for solving those problems, you pick up path dependency and volatility drag.  See these posts:  How Does XIV Work?, How Does SVXY Work?, Ten Questions About Short Selling  for more information.
  • Options are available for UVXY and VXX. Instead of going short on these ETPs you can buy or sell puts and calls. Buying puts eliminates the potential for an unconstrained loss, but the premiums are steep.  No easy money here either. One additional caveat, because of their frequent reverse splits longer term options will likely become “adjusted” options that have the number of shares they control changed and track a modified version of the security price. This happens in conjunction with the reverse splits.  Theoretically, no value is lost, but by all accounts these options become less liquid and the bid/ask prices widen. These options are American style, so with long positions you’ll always have the ability to exercise them, but caution is warranted.  For more on this see UVXY Reverse Splits.

Seller Beware

  I’ve had direct contact with people that have lost hundreds of thousands of dollars on both sides of these trades.  Honestly, I think considerably more is lost on the long side, but the blowouts on the short side tend to be quick and vicious.  Most rookies get greedy and risk being blown out by even a mild correction. If you can manage to hold (and rebalance) your short position long enough it’s a rational trade—but that’s a big if.

Related Posts



Wednesday, April 12th, 2017 | Vance Harwood
  • Mark Shinnick

    Ron, I trade volatility frequently so if you can use a hand using these instruments for their best purpose and potential let me know.

  • Mark Shinnick

    Anthony, I trade volatility frequently so if you can use a hand using these instruments for their best purpose and potential let me know.

  • Mark Shinnick

    Yes, everyone backs away and things remain thin even at natural objective levels. Fewer-and-fewer of us remember those times, and that’s aside from the recency bias. Thanks for your extremely relevant input about reality.

  • BB

    Obama kept cash hordes! Good to know. How’s that uvxy working for you?

  • Roiven Saar

    Hi Vance,

    I recall an old blog from 2012 where you mentioned a strategy that uses ZIV, medium-term inverse volatility fund combined with the VIX/VXV ratio, to switch between fully invested and cash. Is that strategy still valid?

    Thanks!

  • Don

    I see a lot of ideas involving shorting VXX. Isn’t it better to buy XIV?
    Thanks!

  • Hi Roiven, That strategy stopped working in 2014. Don’t know why. Updated simulation results below.
    https://sixfigureinvesting.com/wp-content/uploads/2017/04/ZIV-simple-ratio-3Mar17.jpg

  • hb7of9

    Is it not feasible to hold those positions long term, meaning a couple of months to mayhaps a year (who knows?). Eventually I also think that it is a near certainty that in the long term, the price will again drop lower to that level. After all the trend is for volatility to spike over short (relatively) periods of time versus a long term.

    Would love to know what you think

  • hb7of9

    I have read your articles and ideation on this topic with much interest. As far as I understand, it seems like these instruments look like a sure way to make money, but as you argued that is rarely the case, as people often lose thousands of dollars. The reasons for which are:

    1) Position sizing, the market spikes and thus goes significantly against the short trader. While there is no data for the 08 crash you mentioned that there would have been a 6x (600%) spike on VXX and 15x (1500%).

    2) If there short goes against you, it leverage increases as prices go higher whereas if the short works – the potential profit percentage drops and is limited to 100%, whereas the theoretical loss is unlimited.

    I would like to challenge a couple of these ideas, mayhaps because of my limited understanding and ultracrepidarianism. This thinking is based on just trading the shares, not options.

    a) Theoretical Profit is limited to 100% and the percentage gains decrease for the trade over time.
    Then why not create a system that targets relatively small profit targets, e.g. 4% – 20% range, before the law of diminishing returns comes into effect? I.e. short the trade with a short small predetermined profit target and close it. As you stated the contango effect will batter the instrument down over time and thus, over a long enough time period the trade will close at profit. Additionally the smaller the target trade, the sooner it will close.

    b) The leverage effect if the market spikes.
    Based on the chart the biggest from a previous low on VXX was 207% between July and November of 2011. And I think you mentioned that sims showed it would have gone up 600% during the 08 crash. OK, so then why not consider the following:
    *if the maximum possible jump is expected to be 600%, then base the trade size on that.
    * e.g. if my total investment account is $15,000 then allocate $3000 as a margin of safety, divide the rest ($12K) that by the maximum expected spike 12,000 / 6 = $2000. Thus only short $2000 of the whole account.
    * e.g. buy 200 VXX @ $10.00 = $2000. If the market spikes 600%, to $60 per share = $12,000

    Does using this methodology not address the leverage spike issue?

    Additionally:
    *Add an additional $1K to the account and recalculating trade size every 12 months.
    *If the market does spike, just keep the trade open and wait it out, volatility can only last x amount of years. I am cognizant that there has been extended volatility in the distant past, e.g. 1929-, but I assume the government has gotten better at reducing such extended volatility over long periods as our understanding of economics has improved.

  • Hi Don, It really depends on how the market is acting. If volatility is volatile, with a lot of VIX movement up and down, then shorting VXX would be better (with periodic rebalancing to keep the leverage ratios in line). If volatility is relatively tame or trending, then XIV is better.

  • hb7of9

    I have read your articles and ideation on this topic with much interest. As far as I understand, it seems like these instruments look like a sure way to make money, but as you argued that is rarely the case, as people often lose thousands of dollars. The reasons for which are:
    1) Position sizing, the market spikes and thus goes significantly against the short trader. While there is no data for the 08 crash you mentioned that there would have been a 6x (600%) spike on VXX and 15x (1500%).
    2) If there short goes against you, it leverage increases as prices go higher whereas if the short works – the potential profit percentage drops and is limited to 100%, whereas the theoretical loss is unlimited.
    I would like to challenge a couple of these ideas, mayhaps because of my limited understanding and ultracrepidarianism. This thinking is based on just trading the shares, not options.
    a) Theoretical Profit is limited to 100% and the percentage gains decrease for the trade over time.
    Then why not create a system that targets relatively small profit targets, e.g. 4% – 20% range, before the law of diminishing returns comes into effect? I.e. short the trade with a short small predetermined profit target and close it. As you stated the contango effect will batter the instrument down over time and thus, over a long enough time period the trade will close at profit. Additionally the smaller the target trade, the sooner it will close.
    b) The leverage effect if the market spikes.
    Based on the chart the biggest from a previous low on VXX was 207% between July and November of 2011. And I think you mentioned that sims showed it would have gone up 600% during the 08 crash. OK, so then why not consider the following:
    *if the maximum possible jump is expected to be 600%, then base the trade size on that.
    * e.g. if my total investment account is $15,000 then allocate $3000 as a margin of safety, divide the rest ($12K) that by the maximum expected spike 12,000 / 6 = $2000. Thus only short $2000 of the whole account.
    * e.g. buy 200 VXX @ $10.00 = $2000. If the market spikes 600%, to $60 per share = $12,000
    Does using this methodology not address the leverage spike issue?
    Additionally:
    *Add an additional $1K to the account and recalculating trade size every 12 months.
    *If the market does spike, just keep the trade open and wait it out, volatility can only last x amount of years. I am cognizant that there has been extended volatility in the distant past, e.g. 1929-, but I assume the government has gotten better at reducing such extended volatility over long periods as our understanding of economics has improved.

  • home GZ

    The thing is shorting with 10% of totally investing and paying round 8% for shorting fee, it is may be not worth to do it.
    But I agree with your rebalance strategy.

  • vadim137

    Hi Optionbets, About using options to hedge a short VXX position: My concern is that in a black swan event, when VXX shoots to the sky, SPY may not move that much, and those puts may be not a good hedge. Have you thought about buying deep out-of-money calls on VXX or UVXY instead?

  • optionbets

    But if the VIX rises the extrinsic value of the S&P will still rise. VXX and UVXY suffer badly when they are in contango which is most of the time, so unless you can time it or only trade it when VIX enters contango and protect against things getting worse and black swan not appearing from complacent markets I suppose that is one option.

    I prefer to buy options when implied volatility is cheap. The implied volatility OF instruments that track the implied volatility is usually going to rise expecting some kind of mean reversion which makes things difficult…

    I think buying VXX puts is okay if you can grab a spike and expect contango to resume…

    When VIX is cheap I want to play more options and sell my XIV shares or VXX puts to fund those trades.

    Just me personally not a recommendation to buy or sel and other standard disclaimer stuff.

  • vadim137

    This makes sense, but it is hard to estimate the delta of such hedge. The hedge is simpler with VXX calls.

    Say, I am short 1000 shares of VXX. I can buy the same amount of very cheap VXX calls with the strike of 30.
    I will still need enough equity in the account to survive if VXX goes up to 30. After that, the delta of the call option will be almost 1, and I am fully hedged.

    Both VXX and its calls will loose their value as a result of contango, but as long as the number of options matches the number of shares, I will be safe even if VIX (or VXX) goes totally through the roof.

    What do you think?