Godot Finance’s Mojito 3.0 Strategy

This is a test of Mojito 3.0, a strategy from Godot Finance for trading VIX ETPs like XIV and VXX. The always entertaining John Orford briefly discussed a previous version. This latest iteration is similar to a number of other strategies that we’ve covered on this blog, in that it compares a shorter-term measure of implied volatility to a longer-term measure, going long or short the VIX when the difference between the two is sufficiently large.

I’ve made some changes to the Godot’s original test for reasons I explain in a bit. Strategy results from 07/2004 trading XIV (inverse VIX) and VXX (long VIX) follow in blue, versus buying and holding XIV in grey. Read about test assumptions, or get help following this strategy.

20150813.01

Strategy rules:

  • Near the close, calculate the 5-day median value of the “IVTS”, or implied volatility term-structure, where IVTS = VIX spot / 45-day constant maturity price of VIX futures (1).
  • Go long XIV at the close when the 5-day median value will be < 0.91, long VXX when the 5-day median will be > 1.10, or else to cash. Hold until a change in position.
  • Read about test assumptions, or get help following this strategy.

* * *

Note that we’ve made important changes to Godot’s original test:

  • We’ve extended the test back to mid-2004, and updated it up to the present, adding an additional 8+ years of data. We’re able to do this accurately using simulated data.
  • In order to make an apples-to-apples comparison with other strategies we’ve tested on this blog, we’ve (a) opted to go long XIV as opposed to short VXX when the strategy calls for a short VIX position, and (b) increased position sizes to 100% (from 60%). Note that a short VXX position would have led to slightly different results, but not dramatically so. A good strategy will still be a good strategy (and vice-versa) with either approach.

20150813.02

As mentioned in the opening paragraph, Mojito 3.0 is similar to a number of strategies we’ve tested on this blog, in that trades are based on comparing a shorter-term measure of implied vol relative to a longer-term measure.

These types of strategies have worked historically because of the tendency to overestimate future volatility that exists across the VIX complex when the VIX complex is in its “normal” contangoed state (2): the VIX spot tends to overestimate future realized vol, VIX futures to overestimate the spot, more distant months to overestimate more than nearer months, etc.

These strategies are each using different metrics to judge whether the VIX complex is in that contangoed state, or more specifically, a contangoed state that is likely to mean that VIX futures are overestimating the eventual spot.

Other examples of this type of strategy include: comparing first vs second month futures, VIX vs front month futures, VIX vs 1-month CM, VIX vs VXV, V&M’s VIX:VXV ratio, QT’s VXV:VXMT ratio, Evolution Capital’s strategy, and many more.

Is this strategy better or worse than those other variations?

That’s impossible to say. I think the broader concept has merit for sure, and it’s a broad concept that we use in our own trading, but I also think that over the long-term, taking a more holistic view that considers many of the key data points across the VIX complex together (rather than two particular data points alone) is probably the more robust solution.

20150813.03

A big thank you to Godot Finance for the thoughts and the opportunity to add our two cents here.

When the strategies that we cover on our blog (including this one) signal new trades, we include an alert on the daily report sent to subscribers. This is completely unrelated to our own strategy’s signal; it just serves to add a little color to the daily report and allows subscribers to see what other quantitative strategies are saying about the market.

Click to see Volatility Made Simple’s own elegant solution to the VIX ETP puzzle.

Good Trading,
Volatility Made Simple


Wonk notes:

  1. The “45-day constant maturity price of VIX futures” is calculated based on a weighted average of 1st and 2nd month futures when the number of calendar days to expiration for the second month is greater than 45 days, otherwise it is based on a weighted average of 2nd and 3rd month futures.
  2. I’m using the term “contangoed” loosely here to mean a more distant measure of implied volatility is priced higher than a nearer measure, rather than the stricter definition of futures vs the spot.
Posted in Strategy Backtests.