top of page

Volatility ETF Trading Strategies - Part 3: Mean Reversion


* I’ll update the charts and commentary in these very old articles every now and then so that the data is stretched over longer periods of time. Last updated Nov 2015 *


In this series of five posts I’m going to be going into a little more detail about some of the most common methods of trading the various volatility ETP’s such as XIV and VXZ. I think it’s important that subscribers in the VTS community at least have a basic understanding of the complexity of these products as well as some of the potential drivers of possible strategies. The five part series will cover:




Part 3: Mean Reversion


A simple definition of mean reversion says that prices tend to eventually move back towards the average. Now individual stocks and even indexes don’t necessarily mean revert because they change and evolve over time. The S&P 500 for example was once below 100, but it's not a mean reverting index in that we have no expectation that it will ever revert back to a value below 100.


However one index that does show fairly strong mean reversion properties is the VIX Index. It has traded in a range from about 9 on the low end to about 80 on the high end and we do have an expectation to see those values plus everything else in between as we move forward in time. The VIX index is mean reverting.


And since the VIX index to a certain extent is at the heart of most volatility trading strategies, it’s natural to assume that these strategies themselves can benefit from applying mean reversion techniques.


One simple method traders use to apply mean reversion is to track a ratio of the exponential moving average compared to a simple moving average. Since an exponential moving average reacts quicker and is more weighted towards what’s currently happening, the theory goes that this crossover point can be used to predict future price movements in XIV.

  • As the VIX moves higher it increases the chances of future movements lower towards the mean and should provide a tail wind for the XIV.

  • As the VIX moves lower it increases the chances of future movements higher towards the mean and should provide a tail wind for VXZ.

So let’s see how this theory actually performed in the real world:



That’s not very good at all, what happened?  It seems that while the VIX index definitely does mean revert, the XIV and other volatility ETPs don’t seem to.

In order to understand why mean reversion strategies perform so poorly during market downturns, you have to conceptualize what’s actually happening.  As markets fall, volatility and the VIX index moves higher.  Mean reversion states that the chances of it coming back down are increasing so you keep holding the XIV through the small downturn.


However, what if a small sell off turns into a medium sized sell off?  Doesn’t that mean that the VIX keeps going higher?  If the chances of it mean reverting before were high, then now they're even higher so you keep holding the XIV through the medium sized downturn taking losses along the way.  And if that medium size sell off turns to a major one?  That’s right, you keep holding the XIV through the entire sell off waiting for mean reversion to kick in.


Mean reversion works until it doesn’t.  As John Maynard Keynes famously stated, “markets can stay irrational longer than you can stay solvent.”  You can take portfolio melting losses holding the XIV waiting for mean reversion to kick in, and when it finally does you may not have any capital left to ride the reversion back up.

At Volatility Trading Strategies we don’t use mean reversion indicators at all.  They are completely counterintuitive to what we are trying to achieve.  Our goal is to smooth out results over time and eliminate as many of the bad holding periods as we can, not increase the odds of us holding on to a falling knife.


Anybody who mentions mean reversion as part of their strategy, just smile and walk away. At some point in the future these volatility ETPs are going to suffer gut wrenching drawdowns on a buy and hold basis, and anybody relying on mean reversion is going to be in for a world of hurt.


For reference, here is our results again:



Awesome VTS Newsletter sign up - Claim your FREE 1 month trial now!


bottom of page