Trading VIX ETP Strategies with a 1-Day Lag

Our strategy trades with a 1-day lag, meaning signals generated after the close today are not executed until the close tomorrow.

That’s partially because it makes the strategy easier for subscribers to execute, but it’s also because historically this lag has had little impact on, not just our strategy, but most VIX ETP strategies that are longer-term in nature (including most of those we’ve covered on this blog).

To illustrate, below is the simplest strategy we’ve shared previously, the First vs Second Month strategy, trading XIV and VXX since mid-2004 (read about test assumptions).

The grey line represents the strategy as originally tested, where signals are calculated and executed at the same close, while the blue line adds a 1-day lag where signals are executed at the following day’s close.

20140618.01

There has been very little difference in performance, and whatever difference has existed is likely as much random chance as anything else.

Another test, this time with DDN’s VRP strategy. Grey line = no lag, and blue line = 1-day lag.

20140618.02

Again, no significant difference.

Almost every strategy we’ve tested on this blog follows this same pattern. The only exceptions are strategies that trade much more frequently, such as TM’s RSI(2).

* * *

How can something as important as losing a full day not materially impact results?

Part of the reason lies in how infrequently these strategies are changing position, meaning that losing an individual day here or there isn’t of huge consequence.

But part of the reason lies in the short-term mean-reverting nature of the VIX and VIX products.

Most of the strategies that we’ve talked about are, in a roundabout way, akin to momentum strategies. They tend to take a position when the market is already moving in a favorable direction (as opposed to mean-reversion strategies, which take a contrarian position).

But because the VIX tends to mean-revert in the short-term, VIX products often move against you shortly after a new signal. That 1-day lag is helping, unintentionally, to counter that.

The key takeaway is that when trading the type of strategies we usually cover on this blog, it’s not critical that the trade always be executed at the close on day 0. Discretionary traders might take this opportunity to wait for the following open, close, or some other point in between.

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

Good Trading,
Volatility Made Simple


Wonk note: why didn’t I test taking positions at the following open (rather than waiting for the close)? There is simply too little data to perform sufficient analysis. Recall that data prior to the launch of each ETP in 2009/10 has been simulated (read more). We’re able to do this accurately for the close using a combination of the indices and futures data on which these ETPs are based, but we’re not able to do this with any confidence for the open.

Posted in Strategy Mechanics.