Sunday, October 13, 2013

Inverse ETFs on VIX: are they suitable for buy-and-hold investing?

May be you already know that almost permanent contango on VIX futures makes it possible to build successful shorting VIX strategies. You can short VIX using VIX futures or shorting VXX, or going long inverse ETFs like XIV. So you may think why not use inverse VIX ETF for long-term buy-and-hold investment.

The main argument against it - there is not enough historical data for inverse VIX ETF. XIV was started in the end of 2010. The most interesting period of 2007-2009 is not present in data.

Fortunately, futures on VIX where started in 2004 and using it we can restore XIV back in time. Here is the chart of the restored XIV since 2006:


As you can see, 2007-2008 years where very hard for XIV holders, more than tenfold decline is definitely beyond any imaginable psychological limit. Year 2011 which is present in data was not easy too, the equity declined from 20.0 to 5.0.


Knowing about strong negative correlation between S&P 500 index and VIX volatility index, you may suspect that XIV is just a proxy to an inverse leverage fund on S&P 500. If you calculate standard deviation of price returns XIV and SPY for their common time period, ratio will be 3.9.


But maybe it worth holding XIV with a coefficient of 0.25? Maybe holding XIV has some edge against SPY? Let's check this theory using my favorite tool: equal-volatility pair. Here is the dynamic of a XIV/SPY pair portfolio where each leg of the pair is re-balanced daily using trailing one-month volatility:


As you can see, XIV has no edge against SPY. So there is no reason to hold XIV.


Ok, but here's the thing. Short-term XIV has an alternative: medium-term ZIV. XIV is made of a futures basket with an one-month duration, ZIV if made of a futures basket with a five-month duration. It makes ZIV less volatile, and it is for good, because it makes ZIV less prone to compounding error. Compare XIV (black) with ZIV (red):

As you can see, returns are almost identical, but ZIV has lesser volatility. For example, in 2011 ZIV declined from 18.0 to 10.0, which is unpleasant too, but not so much as with XIV.

SPY/ZIV volatility ratio for their common period is 1.9. Let's check equal-volatility pair for ZIV/SPY:

ZIV has an edge! So it's not just a wicked proxy to 2-x inverse leveraged fund on S&P 500, it is something better!

Some more calculations: from the inception ZIV has grown 2.6 times (log return = 0.96), the same time SPY shows 1.5 (log return = 0.42). Using volatility ratio 1.9 it comes down to the following comparison: ZIV 0.51 against SPY 0.42, or 17% yearly against 14% yearly.


Conclusion:


ZIV has doubled volatility against S&P 500 and definitely has some edge. Holding it the only equity in a portfolio would be too boldly, but with a coefficient of 0.5 it may become reasonable substitute for an index ETF. But planning re-balancing rate take into account ZIV's liquidity, its bid-ask spread is far from 0.01.


No comments:

Post a Comment