VIX Spikes Were Uber Common in 2014

This is a follow up to a piece from the always excellent VIX & More: 2014 Had Third Highest Number of 20% VIX Spikes. V&M finds, as the title suggests, that 2014 exhibited the third most 1-day VIX spikes of at least 20% since 1990.

In this post I look at a different definition of VIX spikes that takes multi-day spikes into account: VIX closes of at least 20% above a 10-day moving average (*). I’ve used this criteria in the past (read more and more), and while it results in about the same number of spikes as V&M’s data, it also captures big moves where no one single day might have met V&M’s threshold.

20150108.01

The blue bars represent the number of spikes, and the grey area the median value for the VIX during that year. Note that I’ve extended the data back to 1986 using VXO prior to the launch of the VIX.

Since 2006, there has been an uptick in the number of VIX spikes. 2014 ranked second amongst all years with 5 significant spikes, despite a median VIX value near the lowest seen in the VIX’s history. Note that the market has had similar years with a high number of spikes despite a tepid VIX in 1993, 1994 and 2006.

Contrary to conventional wisdom, the VIX has actually been less “spiky” during higher VIX years. That’s partially because it’s easier for a market event to shock the volatility market when the VIX is depressed and complacent rather than already elevated and cautious, and partially due to the fact that high VIX years were more likely to have one major spike that lasted a long time (rather than a series of smaller spikes).

Note that similar results were seen in V&M’s data, and similar conclusions would have been drawn had we used other variables for calling a VIX spike other than 20% above the 10-day MA.

Was 2014 significant in terms of the number of VIX spikes? Significant yes, but not so far outside of expectations that I think it necessarily says anything about the future. What’s more concerning for me is just the general increase in the market’s tendency to spike since 2006 and whether we continue to see that increase in the future.

Good Trading,
Volatility Made Simple


Wonk note: To control for overlapping VIX spikes, I required that after the VIX rose at least 20% above its 10-day average, it had to then fall back below its 10-day average before it could then register a new spike.

VIX ETP Seasonality: Days to Expiration

20141105.01This is a follow up to a post from the always excellent Trading the Odds showing that XIV (inverse VIX) is particularly strong (and VXX particularly weak) in the two days prior to VIX futures expiration, as futures are usually forced to converge down to the spot VIX as a result of VRP. I’ve shown TTO’s results to the right (click to zoom).

I have a slightly different take on the subject. XIV strength on those days isn’t, in and of itself, necessarily an important bit of data. Given the fact that we have limited VIX ETP history to consider (just over 10 years), it could be that, just by happenstance, equities were strong on those particular days and XIV was merely along for the ride.

A different approach would be to look at the daily excess return of XIV, above and beyond what we would expect given the day’s change in equities.

20141105.02

In the graph above I’ve shown just that: the daily excess return in XIV by days to expiration.

I performed a simple linear regression between all available SPY (cash only) and XIV data, and calculated XIV’s average daily return in excess of that regression, for each day. The relationship between daily changes in SPY and XIV is not linear, but close enough for the purposes of this illustration.

Note the jump in excess return 1 and 2 days prior to expiration, confirming TTO’s original observation. To clarify, that doesn’t necessarily mean that those days were positive (or for example, that day 4 was negative), only that they tended to return more or less than one would expect given the day’s change in equities.

Just for fun, the following graph shows two hypothetical portfolios. The first (blue) only buys XIV 1 and 2 days prior to expiration, and the second (grey) buys XIV on all other days. All trades are close-to-close, and I’ve ignored transaction costs and slippage.

20141105.03

The graphs shows that, to date, it’s been mostly smooth sailing for XIV 1 and 2 days before expiration. But, likely by pure happenstance, it’s also been smooth sailing for SPY on those days as well (not shown for the sake of brevity). So that means that those positive XIV returns in TTO’s data are partially a result of the excess return on those days that we’ve shown here, but it’s also likely partially because nothing particularly bad has happened to equities on those days either.

Wonk note: The observation discussed in this post is much less pronounced with ZIV/VXZ (mid-term VIX ETPs) because they’re based on futures that are further from expiration and thus are not forced to converge to the spot VIX while ZIV/VXZ holds them.

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

Good Trading,
Volatility Made Simple

Where Will This VIX Spike Bottom Out?

In our previous post: Where Will This VIX Spike Top Out? we showed that, by one measure, this month’s VIX spike was 9th in the entire history of the VIX in terms of how quickly it rose from recent levels and then bucked mean-reversion and continued even higher. Specifically, we used Adam Warner’s criterion for calling the start of a spike as being when the VIX closed at least 20% above its 10-day average, and then measured how far the spike continued beyond that point.

Now our current spike has accomplished another historically significant feat: on Tuesday it closed more than 20% below its 10-day average. That’s only the 13th time that’s happened since 1986 (1).

To illustrate what we’re talking about here, the graph below shows the VIX index since 03/2004, when VIX futures began trading, with “plunges” of at least 15% below the 10-day average marked in orange (2). Note that I reduced the threshold to 15% to give us more observations to consider.

20141023.01

As you would expect, all of the “orange circles” occur after a sharp decline from a near-term top.

The important question is what tends to happen immediately after one of these plunges? To answer that, the following table shows the average change in the VIX in the days and week following one of our “orange circles”.

20141023.02

Historically, the day immediately following a plunge has been mixed, but in 72% of cases, the VIX has been higher a week out, with an average change of +6.2%.

The usual warnings about small sample size definitely apply here, and these stats are in no way conclusive on their own, but they do point to the tendency for VIX mean-reversion to cut both ways.

As we showed in our last post, when the VIX spikes too far too fast, there’s a strong tendency for the VIX to fall, and conversely, when the VIX plunges too far too fast, there’s an opposing (though not quite as strong) tendency for it to rise.

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

Good Trading,
Volatility Made Simple


Wonk notes:

  1. For dates prior to 1990, I used the VXO index in place of the VIX. It’s an imperfect comparison, but close enough I think for the purposes of this illustration (and it’s useful to be able to capture the 1987 crash).
  2. To control for overlapping VIX plunges, I required that after the VIX fell at least 15% below its 10-day average, it had to then rise back above its 10-day average before it could then register a new plunge.

Where Will This VIX Spike Top Out?

In previous posts we’ve broken down VIX spikes like the one we find ourselves in the middle of now. We’ve generally used Adam Warner’s criterion for calling a spike as being when the VIX closes at least 20% above its 10-day average (although note that the conclusions that follow hold for other values as well).

We’ve shown that mean-reversion tends to take over quickly following spikes of these magnitude, driving down the VIX and ETPs like VXX, and driving up inverse ETPs like XIV (see the numbers). In this post I want to look at how often that hasn’t been the case, or put another way, how often the VIX has continued to climb following one of these spikes, and where the VIX has eventually topped out.

First, to help visualize what we’re talking about, the graph below shows the VIX index since 03/2004 (when VIX futures began trading), with spikes at least 20% above the 10-day average marked in orange (1).

20141016.01

Note how the majority of the time, the orange circle marks either the top of the spike, or somewhere very close. Notable exceptions include late-2008 during the GFC, 2010/11 during the Euro debt crises, and our current spike at the far right of the graph.

Going back to 1986, there have been 75 such spikes on the VIX (2), or about 2.6 per year. In 73% of instances, that orange circle would be within 10% of where the spike eventually topped out. In 84% of instances it would be within 20%.

In other words, the impact of mean-reversion grows strong when the VIX gets so far so fast above its recent average. Very rarely does the VIX spike continue much further, except that when it does, it occasionally goes much further.

The next graph illustrates. Here we see all 75 spikes in the VIX’s entire history since 1986 (2), along with how far the spike eventually travelled “past the orange circle” before eventually topping out.

Again, the “orange circle” represents the initial VIX spike at least 20% above the 10-day average. Values of zero would mean that the orange circle represented the absolute top of the spike.

20141016.02

Our current spike, marked in orange on the graph, is now 24% above the orange circle. That puts it, by this particular measure, in 9th place in the VIX’s history.

Again, this isn’t a measure of highest VIX closes. The VIX has gone through long periods of time higher than current levels. Rather, this is a measure of when the VIX has spiked very quickly and suddenly from recent levels, and then bucked mean-reversion and continued even higher.

If the VIX continues much further from here, it enters rarefied air only seen during the largest previous crises in ’98 (LTCM), ’01 (9/11), ’10 and ’11 (Euro debt crises), ’08 (GFC), and of course, ’87 (Black Monday) (3).

Good Trading,
Volatility Made Simple


Wonk notes:

  1. To control for overlapping VIX spikes, I required that after the VIX rose at least 20% above its 10-day average, it had to then fall back below its 10-day average before it could then register a new spike.
  2. For dates prior to 1990, I used the VXO index in place of the VIX. It’s an imperfect comparison, but close enough I think for the purposes of this illustration (and it’s useful to be able to capture the 1987 crash).
  3. The last chart in this post was truncated at 150%. Black Monday (the far right bar) actually topped out at 313% above the “orange circle”.

Going Long Volatility After VIX Pops

This is a follow up to a subject often covered by one of my favorite bloggers, Adam Warner:

How do VIX ETPs perform after VIX “pops” (or spikes)?

Adam’s criteria for a VIX “pop” is a VIX close at least 20% above its 10-day moving average, which occurs a little more than three times a year (1). Many investors see these pops as a signal to buy volatility (ex. VXX or VXZ), but history shows that unless you’re already positioned for the pop before it happens, or the market goes on to crash in grand fashion (ex. 2007/08), buying volatility in the face of such pops is likely a very bad idea.

Below I’ve shown the results of holding VXX for one month (blue) or one week (grey) following a VIX pop (1).

20140910.01

20140910.02

With just a handful of exceptions, going long volatility after a VIX pop has been a very bad idea.

This is a result of mean-reversion in the underlying VIX index itself, and not because of contango or a volatility risk premium or the like. To illustrate, below I’ve shown the same two hypothetical portfolios, this time assuming we bought the VIX spot index rather than VXX, since 1990 (of course we can’t actually buy the VIX index, this is just meant to demonstrate a point).

20140910.03

Because these two portfolios are falling with even more consistency than we saw in the first graph, we know that the results in the first graph are due to mean-reversion in the spot index.

Read more about the two forces acting on VIX ETPs.

* * *

Note that the point of this post isn’t to say that it’s never the right time to go long volatility. To the contrary, we occasionally go long vol in our own strategy. The point of this post is simply to say that, contrary to popular belief, a pop in the VIX is not in and of itself a reason to go long volatility.

A big thank you to Adam Warner for the always interesting thoughts.

Good Trading,
Volatility Made Simple


Wonk note: to control for overlapping VIX pops, I required that after the VIX rose 20% above its 10-day average, it had to then fall back below its 10-day average before it could register a new pop.