In the last few years, investors have increasingly turned to products linked to the VIX (Chicago Board Options Exchange Market Volatility Index) to manage risk in their portfolios. The VIX is calculated from the implied volatility of S&P 500 options across a range of strike prices in the front and second month weighted to achieve a constant 30-day maturity. In other words, the VIX reflects market participants' estimates of market volatility in the coming month. This metric has a very appealing quality to both retail and professional investors: it almost always goes up when the market goes down. Furthermore, in times of crisis, the VIX goes through the roof, quadrupling from about 20 in August 2008 to a high of about 80 in October 2008.
Given the VIX's statistical properties, it is no surprise that products linked to the VIX are marketed as catastrophe insurance based on Wall Street's "fear gauge." Some of the most popular instruments are ETFs and ETNs linked to VIX futures, such as iPath S&P 500 VIX Short-Term Futures ETN (VXX) and ProShares VIX Short-Term Futures ETF (VIXY). Those that really want to roll the dice or achieve the same long VIX futures position for a smaller capital outlay can buy VelocityShares Daily 2X VIX Short-Term ETN (TVIX), or ProShares Ultra VIX Short-Term Futures ETF (UVXY).
Stated simply, all of these products have been absolutely awful since inception. Since its initial issuance on 1/1/2009, VXX has lost 98% of its value, while the leveraged TVIX took only a bit more than two years to lose 99% of its value. The problem is that while these products correctly jumped when the VIX spiked, VIX futures are in steep contango most of the time. This means that maintaining a long VIX position during a period when the VIX does not move much can be quite costly. So while an occasional long VIX bet will pay off in spades, usually you are left with a quickly dissolving asset that weighs down the rest of the portfolio.
If You Must Be Long VIX
Given these features, I view long VIX products as lottery tickets for the middle and upper classes. There is a small chance that you can strike it rich if your timing is right, but a much bigger chance that you overpaid for lousy entertainment. I would never personally take on an un-hedged long VIX futures position, no matter what. In my opinion, volatility traders have only two viable options: being net short or being in cash.
All that said, the devilish appeal of long VIX products remains. And I can understand why many investors would want to take a small (2%-5%) long VIX position as a sort of hedge against a long equity portfolio. You are protected against the worst of a decline and feel safer about taking the plunge into the stock market. Or, you are convinced that you know where the market is going in the next month and want to make a small side bet. But over time, the costs of theses hedges and directional bets can really add up. And like a casino, people win just often enough to keep coming back for more. So while I'd advise against it, if you still have your heart set on a long VIX position, please do one thing: hedge.
Diversification Across VIX Positions
Depending on your time frame, the VIX and the S&P 500 have a correlation somewhere in the ballpark of -0.8. That is quite impressive, but there is an asset that has a -1.0 correlation with a long short-term VIX futures position: a short short-term VIX futures position, such as VelocityShares Daily Inverse VIX Short-Term ETN (XIV), or ProShares Short VIX Short-Term Futures ETF (SVXY). This makes sense, because long and short positions of the same underlying are exact opposites.
This basic fact has been not been adequately emphasized in commentary here on SA about VIX products. Excluding fees, investors could theoretically have a perfect hedge if they matched up long and short VIX futures ETP exposures. Where this idea gets really interesting is the possibility of rebalancing a portfolio of two assets with a -1.0 correlation. In such a portfolio, investors would be guaranteed to buy cheap and sell dear every time they rebalanced, especially given the mean-reverting nature of the VIX. This is the holy grail of Modern Portfolio Theory, one asset that always zigs while the second asset zags.
Now for those wanting that long VIX exposure, all they need is a little tweak on the same general idea: partially hedge a long VIX position with a short VIX position, and have some sort of rebalancing plan. This can reduce or eliminate your carry costs but still give you plenty of juice if the market tanks.
A Long-Short VIX Portfolio
Assuming I've sold you on the theoretical appeal of a hedged long VIX position vs. an unhedged long position, there's also empirical support for my idea. S&P Dow Jones constructs a variety of VIX futures indices whose historical performance goes back to 2008 that have been used to create exchange traded VIX products. This provides a longer period for back testing than the ETPs themselves, and includes a set of long-short indices that combine different blends of VIX futures positions.
Of most interest to VIX longs is the S&P 500 VIX Futures Tail Risk Index- Short Term, which combines a 45% 2X long short-term VIX futures position and a 55% short short-term VIX futures position. The index is composed of 13 sub-portfolios, each of which is rebalanced quarterly on a weekly rolling basis so that 1/13 of the portfolio is rebalanced every week. This index can be replicated using TVIX or UVXY on the long leg and XIV or SVXY on the short leg. Here's how the index has performed relative to an un-hedged long VIX futures position analogous to VXX or VIXY since 2008:
If I had to pick a net long VIX position as portfolio insurance, I'd definitely pick the tail risk index over an un-hedged long position because it captures the bulk of a spike while putting a significant damper on roll costs. Heck, even if I wanted a short-term speculative long VIX position, I'd still want to have a short VIX hedge to improve my odds of a win and cut down on losses in case I was wrong.
Elaborations on the General Theme
While the tail risk index seems like a decent way to construct a net long VIX futures position to me, there are plenty of ways for investors to establish and maintain a long-but-hedged VIX futures position. As the VIX ETP universe has expanded, investors can now establish positions further along the maturity curve, which have a lower "beta" relative to VIX moves but also tend to have lower roll costs. Or investors can play around with rebalancing rules, net exposure levels, or introducing a cash position into the mix. Below is a set of possible ways to flavor a hedged position to suit individual needs or in the pursuit of a better alternative than the Tail Risk Index:
|Long Mid-Term VIX Futures (VXZ)||Generally Lower Roll Costs, Less sensitivity to VIX spikes|
|Different Rebalancing Rules (% change rather than fixed time, more or less often than 1/13 every week)||Not really sure, someone should research this|
|Larger short VIX hedge||Eliminates more roll costs, less sensitivity to VIX spikes|
|Introduce cash component to create a 3-leg long, short, and cash position||Not really sure, someone should research this|
Let's Change the Conversation
The last few years have been a sort of Wild West for VIX ETPs. I've lost count of the number of articles written by gunslingers advocating a naked long VIX futures position. And sometimes those gunslingers have successfully shot the moon. I raise my glass to their successes, but I suspect that the long VIX cowboy crowd has a lot more losers than winners. And I can tell that people are wising up by the growing numbers of VIX shorts and skeptics in comments sections blasting the longs.
Personally, however, I think the discussion needs to evolve past the simple longs vs. anti-longs dynamic. The community of VIX traders is growing, as is their knowledge and experience. Regardless of how optimistic you are about your market timing skills or how much you worry about roll costs, we need a more nuanced approach towards VIX products that includes hedged positions. There are an infinite number of ways for investors to establish a hedged VIX position, and I believe that the investing public would be well-served if some of the great minds here on SA further investigate how different hedging strategies meet different investor needs.
If you ever use a hedged strategy when trading VIX products, please share your insight and leave a comment. I'm all ears!