Back in June of last year, we first wrote about the concept of Antifragility and a promising new exchange traded product that had the potential to deliver antifragile returns. The product we highlighted was the iPath S&P 500 Dynamic VIX ETN (XVZ) which, as we explained, was designed to tread water during periods of relative calm in the markets while capturing a good portion of the upside spike in volatility during periods of market stress. Since writing that first post, we believe certain dynamics have shifted in the volatility landscape which make a buy-and-hold position in XVZ a more expensive hedge than what we had originally anticipated.
To review, the VIX Index itself is not investable, but there are various derivative products tied to the VIX, including very liquid futures and options markets. Exchange traded products such as XVZ, with the expressed goal of providing derivative exposure to the VIX or some VIX-related strategy, are typically tied to an index which tracks a position in VIX futures contracts. The first generation of VIX products (VXX and VXZ) tracked indices that maintained a constant exposure to the VIX through a rolling position in VIX futures. The chart below shows how a buy-and-hold position in each of these indexes has fared since they were launched in late 2005.
The obvious takeaway from the chart above is that although the VIX itself is mean reverting, these two indices that are tied to VIX futures are not. Notice that the VIX (plotted on the right hand scale) actually finished above where it started, while the Mid and Short-Term Futures indices finished down 98% and 48% respectively. The reason lies in something called roll cost, which we described in our previous post.
As a practical example if the front month VIX contract is trading at $20 and the second month is trading at $21, then there is a $1 roll cost of selling each contract at $20 and buying the same number of contracts at $21. This roll cost can be significant depending on the prevailing shape of the VIX futures curve.
Even though these first generation VIX products have serious flaws, they continue to be very popular, commanding over a billion dollars in assets. The next generation of VIX exchange-traded products was launched with the roll cost issue in mind. As we wrote in a follow up post entitled The Patience of Taleb, XVZ follows a rules based strategy to try to mitigate the impact of the roll cost while still capturing a good portion of the upside in a volatility spike.
XVZ shifts its allocation to VIX futures based on the cost of maintaining the exposure. When the curve is flat or inverted, it is 100% long volatility. As the curve gets steeper and steeper it tries to offset the cost by shorting the front end of the curve which is the steepest…while still maintaining long volatility exposure to the back end of the curve where the shape of the curve is a bit flatter. During these volatility regimes, the strategy will always have more exposure long in the back end than it has short in the front end of the curve in order to maintain a net long volatility position.
Intuitively this made sense. When roll cost was high, VXZ would try to capture that cost (making it a roll yield) and shift its long volatility exposure to the flatter back end of the curve. The backtest and first several months of returns for the strategy looked very promising, but then around the middle of last year the returns began a slow bleed downward.
We unpacked two of the culprits behind the sell-off, which were a fall in the VIX price and a flattening of the VIX futures curve, in our follow up post. Both of these factors are transitory in nature and weren't enough to convince us that the strategy underlying XVZ was broken. That said, our confidence in the strategy was rattled by the fact that the present drawdown was deeper than any other experienced in the backtest. What ultimately ended up being a game changer for our thesis was the fact that the roll cost in the back-end of the curve had steepened significantly, making XVZ's roll cost mitigation strategy much less effective. The average roll cost for mid-term VIX futures going back to October 2006 through the end of June 2012 was 0.7% a month. If we strip out all the days that the VIX curve was not in contango (e.g. flat or inverted) then the average roll cost over this same time period clocks in at 1.7% a month. In contrast, since June 2012 the roll cost has averaged 4.3% a month, or roughly 2.5 times as much as the historical contango-only average.
In an email exchange I had toward the end of last year with Vance Harwood, a fellow VIX aficionado that blogs at Six Figure Investing, he stated:
I'm beginning to think the heavy contango on the medium term futures is here to stay.
If this is in fact the new normal, a strategy like the one underlying XVZ will be hampered in its ability to "tread water" during times of market calm, and instead will experience a consistent performance drag from the elevated roll cost.
After several months of monitoring the shape of the VIX curve, we're now speculating that the elevated roll cost in the back end of the curve is here to stay and is probably being caused by the increased popularity of hedging strategies (such as XVZ), and speculators seeing more opportunity in longer dated VIX futures. As such, we decided to part ways with XVZ under the pretense that the contango mitigation strategy could very well be permanently broken. We also see continued risk from more curve flattening which explained the vast majority of the sell-off in the second half of 2012. We still like the thesis underlying XVZ and would be open to revisiting an investment in the product if we see some of the current headwinds dissipate, but for the time being we are looking elsewhere for our hedging needs.
Disclaimer: Transparency is one of the defining characteristics of our firm. This information is not to be construed as an offer to sell or the solicitation of an offer to buy any securities. It represents only the opinions of Season Investments or its principals. Any views expressed are provided for informational purposes only and should not be construed as an offer, an endorsement, or inducement to invest.