Despite the historically tranquil period between Thanksgiving and Christmas, stocks have been on a roller-coaster ride over the past two weeks-and not in a good way. The rampant destruction in the oil and gas sector has frazzled investors, even as many believe that low oil is a boon for the domestic and worldwide economies. The major markets collectively had their worst-or-second works weeks of the year just seven days after the indices tapped new highs and whispers of Dow 18,000 could be heard on the street. The S&P 500-tracked by the ETF SPY-slid 3.4% while DIA dropped 3.6%. QQQ was spared some of the carnage with consumer spending on electronics expected to be at record levels, sliding only 2.6%. Unsurprisingly, volatility has jumped into the sell-off, with the VIX jumping an impressive 78% to 21.08 as of Friday's close (although the index was near 23 moments before the close which will likely lead to some readjustments on Monday). The 78% jump is the 160th largest 5-day bounce in the Index over the last 25 years, putting it in the 97.5th percentile. Even still, as Figure 1 below shows, the VIX still has a ways to go to reach the October 2014 intra-day high of 31 that saw the Dow drop below 16,000, which I find unlikely given the health of the overall market outside of the oil sector.
Figure 1: VIX performance in 2013
Indeed, without even going into seasonality arguments or a longwinded rationalization of the benefits of cheap oil, I believe that volatility is approaching the level at which significant money may be made by betting against it following a spike in volatility.
At 21.08, the VIX Index is at the 65th percentile over the last 15 years and in the 84th percentile over the last five years since the bullmarket started in 2009. As I have discussed previously, there is a strong tendency for mean reversion with the VIX-like a rubber band, when VIX values distance themselves above or below an historical mean value of around 19, the probability increases that the Index will reverse course to trade closer to this mean value. At 21.08, the VIX is just above this mean value and the elastic band of mean reversion is just beginning to tighten. Over the last 15, when the VIX was within +/-5% of Friday's closing value, the VIX fell 51% of the time over the next 60 days but had a mean gain of +1.0%. In other words, flip a coin. However, the index is right at the threshold of an abrupt increase in the probability for mean reversion. Figure 2 below shows the mean magnitude 60-day performance of the VIX for different ranges of the VIX.
Figure 2: 60-Day VIX performance based on starting level blocks. Friday's VIX close was at 21.
Similarly, Figure 3 below shows the 60-day frequency at which the VIX rallies given an initial starting value.
Figure 3: 60-Day percentage of the time that the VIX rallies based on starting value.
Should the VIX trade to 24, the index drops on average 9% over the next 60 days, falling 75% of the time and should it reach 28, that jumps to 81% for an average decline of 18% over 60 days.
However, investors do not trade the VIX directly. Rather, ETFs are available that are designed to track the VIX and these tend to underperform the index, offering good shorting opportunities as the probability of mean reversion increases, even at current levels. Unfortunately, data for these products is only available for the last four years since late 2010 rather than the entire period of record for the VIX.
Over this four-year period, when the VIX is within 5% of Friday's close, the Index has declined an average of 19% over the following 60 days. On the other hand, the ETF VXX which is designed to track the front-month VIX futures has underperformed, declining an average of 36.5% over the next 60 days, with the fund dropping 87% of the time. Its worst performance was, coincidentally, a 50% slide in December of 2011, while its strongest performance was a 115% gain in July of 2011. The leveraged ETF TVIX, which is designed to track 2x the VIX futures, also underperformed, sliding an average of 59.3% over sixty days, also declining 87% of the time. Its worst performance was a 76% drop also in December 2011 and its best performance was a somewhat alarming 261% rally in July 2011. This data is reflected below in Table 1.
Table 1: VIX ETF Statistics based on current VIX level
The rational for this underperformance can be attributed to rollover losses thanks to contango that is usually present in the VIX Futures market and leverage-induced losses in TVIX. These are topics that I have touched on previously HERE and HERE, and I will not regurgitate these topics here. Regardless, the point is, these funds will make investors money on the short side if the VIX declines or even just stays flat-as history predicts it will do at current levels.
My strategy at these levels will be one of slowly easing into positions. At Friday's close I initiated a small short position in VXX at $34. While the odds are in the short sellers' favor at these levels, I am nonetheless concerned about the Black Swan events that have led to VXX doubling and TVIX nearly quadrupling leading to instant margin calls, as noted above in Table. Therefore, I will slowly ease further into VXX-or add a long position in XIV, which is a VIX inverse ETF-at 10% intervals on the way up. Should the VIX cross 28, the odds tilt overwhelmingly towards mean reversion and I will start to aggressively add TVIX short positions. The entire size of my volatility positions will not exceed 20% of my portfolio and I am prepared to dump other unrelated positions in my portfolio should volatility continue to rise in order to wait for the inevitable mean reversion.
Disclosure: The author is short VXX.
The author wrote this article themselves, and it expresses their own opinions. The author is not receiving compensation for it (other than from Seeking Alpha). The author has no business relationship with any company whose stock is mentioned in this article.