Some thoughts in light of the recent spike in the VIX

This isn’t market commentary or trading advice, but I thought recent work I’ve done and written about on this newsletter could be useful in light of recent market developments; that’s the goal of this substack anyway, so here’s what recent research has to say about developments in the VIX.

Today, the VIX opened at 23.39 (also the low), made a high of 65.73, and closed the day at 38.57. It’s the largest intra-day spike of all time.

For context, here are the 10 highest VIX closes, all either in 2008 or 2020.

But if you omit 2008-09 and 2020, here is what the 10 highest closes would be.

So today’s high of 65 is somewhere in the middle, but it pulled back sharply to close at 38.5.

1. VIX overestimates volatility in normal times and underestimates it in downturns.

This is one of the stylized facts about the VIX, I’ve mentioned it before in another posts. Today’s high of 65 implies a daily volatility of 4.09%, a weekly vol of 9.15% and a monthly vol of 18.76%. The VIX’s close of 38.57 implies a daily volatility of 2.43%, weekly vol of 5.43% and monthly vol of 11.13%. Based on this stylized fact these figures are probably lower than the volatility we will see in the coming days.

However, the problem with the VIX is that it says nothing about the timing. All we know now is volatility will likely remain elevated in the short-term future, anywhere from a week to a month.

2. A good way to check if the VIX has peaked would be to look at the CBOE Put/Call Ratio.

In a recent post, I showed this chart of the VIX and the PCR of the SPX options (background). Notice that as the VIX rose in March 2020, the PCR of the SPX’s options dropped significantly below 1(right scale). The bottom panel is the number of options contracts used to calculate the VIX.

Notice that the number of options used to calculate the VIX also rose. I’ve never figured out why the PCR drops on a VIX spike. I made a chart of two different PCRs: one for all SPX options, and one for just the options used to calculate the VIX.

If you look at March-April 2020, they both dropped! If only the PCR of the options used to calculate the VIX dropped, it would make some sense because when the SPX is dropping, more puts go in-the-money, and the VIX is based on out-of-the-money options. And since the number of options used to calculate the VIX also rises, then it would make sense that the PCR of the options used to calculate the VIX drops sharply. Quite weird that actually the SPX’s PCR drops too on a significant vol spike.

So PCR drops on a VIX spike. This can be used to find the VIX peak. If you look at the VIX and the CBOE PCR during March 2020, you will see that the PCR dropped before the VIX peaked.

That hasn’t happened yet, if anything the PCR is still going up, so maybe the VIX hasn’t peaked yet.

3. Keep an eye on VIX front-month futures

I also wrote in a previous post that that one way to tell the difference between a correction and a crisis is to look at how VIX futures behave after a VIX spike. Here’s the chart I used.

In circle 1, which was in March 2020, the VIX futures followed the VIX (crisis) but in circle 2, which was in Dec 2018, the VIX stayed quite elevated above the futures.

Right now the VIX is quite elevated compared to the VIX futures, so this is probably just a correction; but keep an eye on it.

So to conclude:

1. VIX is probably underestimating vol since it typically does that in a market downturn.

2. Expect the CBOE PCR to come down to signify the VIX peak.

3. Keep an eye on the VIX futures to gauge if the correction will turn to a crisis. i.e. if it will last.These insights are based on this post. I’ve removed the paywall for your reading pleasure.