Growing Interest in XIV
There have been a number of really interesting articles on volatility ETNs over the past couple of months. Last month, Robin Hewitt wrote an excellent piece arguing that contango does not cause price changes in the VelocityShares Daily Inverse VIX Short-Term ETN (NASDAQ:XIV) (235 comments to date). Last Thursday, David Easter published a PRO article in which he disagreed with Robin and concluded that two factors completely determine the value of the underlying index which XIV is based on: (1) changes in volatility, and (2) price difference between the first and second month futures.
Other authors, including Nathan Buehler, Toma Hentea, Harry Long, and myself, have developed trading strategies based on XIV either alone or combined with other funds. The majority, but not all, are based on contango. The prevailing idea is that contango in VIX futures favors the inverse volatility ETNs XIV and the VelocityShares Daily Inverse VIX Medium-Term ETN (ZIV), while backwardation favors the long volatility funds like the iPath S&P 500 VIX Short-Term Futures ETN (NYSEARCA:VXX) and the iPath S&P 500 VIX Mid-Term Futures ETN (NYSEARCA:VXZ).
The figure below shows that the number of SA articles with XIV listed as the primary ticker has increased considerably since 2011. (The projected number for 2016 is based on having 6 articles published to date.)
Performance of XIV
It is not too surprising that backtested strategies that use XIV appear incredibly profitable. From its inception on Nov. 30, 2010, through July 2015, XIV grew 403%. It experienced a huge drawdown in Aug. 2015, and now has a net gain since the inception of 148%.
The figure below shows growth of $10k in XIV along with the SPDR S&P 500 Trust ETF (SPY) and the ProShares UltraPro S&P 500 ETF (UPRO). (Comparing XIV to UPRO makes sense since XIV's beta is usually a bit higher than 3, so we would expect its movement to be more in line with 3x SPY than SPY.)
Performance metrics for the three funds are shown below.
Table 1. Performance metrics for XIV, SPY, and UPRO.
|Fund||Growth (%)||MDD (%)||Sharpe|
When I look at these results, I think that XIV actually doesn't look very impressive at all, since it badly underperforms a (roughly) beta-matched SPY ETF. As I argued in an article last June, XIV's impressive growth since inception is mainly due to its high beta combined with a strong bull market.
Also, UPRO is absolutely killing it.
Effects of SPY, VIX, and Contango
Before investing in XIV, you should have a basic understanding of how its price movement depends on change in SPY, change in VIX, and baseline contango. The figure below shows these dependencies.
If you take one thing away from this article, it should be this:
XIV's price movement depends heavily on SPY and VIX, and very weakly on baseline contango.
We can learn a lot from a linear regression of XIV gains vs. SPY change, VIX change, and contango at prior day's close, summarized below.
Table 2. Linear regression fit for daily XIV gains.
|Variable||Beta (Std. Error)||p|
Here are some interesting results:
- SPY change, VIX change, and contango separately explain 69.7%, 77.4%, and 0.03% of the variability in XIV gains, respectively.
- A 0.5% decline in SPY corresponds to a 3.9% rise in VIX on average. In this scenario, contango would have to be 28.2% in order for expected XIV growth to become positive. In other words, even if contango is favorable, a moderate decline in the market virtually guarantees losses.
- In fact, XIV declined on 267 of the 278 days (96.0%) when SPY lost at least 0.5%.
- Another example: Even if contango is 10%, if SPY loses 0.25%, VIX rises 2.24% on average, and you can expect a 0.6% loss in XIV.
- When SPY and VIX are flat, expected XIV growth is positive for contango >3.73%.
The main point I am trying to make is that holding XIV unhedged is essentially a high-stakes bet on the market. Even if you buy when contango is 10% or higher, you're going to lose if the market declines even moderately. Whenever you're holding XIV, you're fully exposed to its extremely high beta. You can moderate risk by investing only a small portion of your capital, but you're still trading on a weak signal.
Despite the relatively poor signal-to-noise ratio, I think contango-based XIV trading strategies are extremely promising. The key is using an appropriate hedge to minimize the dependence on SPY and VIX movement. Currently available spot VIX funds have pretty terrible tracking, so I'm working on strategies to at least remove the dependence on SPY (for example, here).
I should acknowledge that XIV's unhedged performance is actually quite good when contango is high. Its historical Sharpe ratio is 0.038 overall, but 0.086 when contango >5%, and 0.178 when contango >10%. So small unhedged allocations to XIV when contango is high isn't a bad idea. I just think a hedging strategy is much better, since it increases the signal-to-noise ratio and (ideally) completely removes dependence on market movement.
To summarize, I think that:
- Holding XIV long-term without regard to contango is a bad idea.
- Combining XIV with other funds without regard to contango is a bad idea.
- Trading XIV (and possibly VXX) based on contango is reasonable, but generally too risky.
- Trading XIV and VXX based on contango/backwardation, and using beta-matched SPY hedges, is an excellent market neutral strategy.
I obtained historical prices for various securities from Yahoo! Finance and historical contango values from The Intelligent Investor Blog. I used R to analyze data and generate figures.
Disclosure: I/we have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours.
I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.
Additional disclosure: Any opinion, findings, and conclusions or recommendations expressed in this material are those of the author and do not necessarily reflect the views of the National Science Foundation.