Why Yesterday's Volatility Spike Should Be Shorted

“Volatility is a symptom that people have no clue of the underlying value.” -Jeremy Grantham

Volatility is never a thing that exists in the present. It cannot. Necessarily, you need some historical data to calculate volatility, and it is an amorphous thing constantly evolving with time and the new information being absorbed by the market.

While each volatility event is unique and should be evaluated on its own merits, it's very important to remember that volatility, particularly of equity assets, has specific stylized facts that can help orient us during volatility spikes. Some of those stylized facts are:

• Volatility cannot be and is never constant.
• Volatility is subject to mean reversion and clustering.
• Volatility and volume have a positive relationship. The highest volatility occurs during lopsided, high-volume selling that is historically infrequent in markets.
• Volatility's distribution is approximately log-normal
• When volatility curves invert, it is usually a good sign that volatility has reached a local peak.

The second property is particularly important. Most volatility models are based on the tendency of volatility to cluster. They are based on an incredibly simple assumption: current volatility is a good picture of future volatility.

But you also have to balance this with the property of mean reversion. When you consider the following facts, I think it becomes clear that the theoretical mean of volatility is drifting lower, not higher, despite yesterday's spike:

CBOE

• Realized volatility compared to 1M implied volatility has been at record lows.
• The sell-off yesterday occurred after a massive relief rally into the Russell 2000. This is not a bearish event.
• Labor markets remain very strong, and we are entering an earnings season which tends to suppress volatility at the index level.
• Inflation appears vanquished, and the Fed should start cuts imminently, according to inferred odds from the futures.

I like to take a qualitative assessment of volatility to determine which way a theoretical mean would go. The one-year mean of the $VIX is near where it closed yesterday. Still, I think a helpful approach is to use other quantitative and qualitative developments to infer future directionality. Given the strength of underlying economic catalysts, such an assessment suggests volatility can remain unseasonably subdued.

Assessing the context in which yesterday's volatility occurred is very important. Firstly, the Russell 2000 recently had an incredibly anomalous spike in price after inflation data showed that the issue appeared largely in hand. This, of course, has a great bearing on the smaller less robust businesses in the Russell. However, the event was so bullish that it caused volatility to spike on the Russell significantly, but not from excess put buying, from excess call buying.

This is not only incredibly rare, but it's hard not to see it as incredibly bullish. Essentially, the FOMO on Russell accentuated the overdone losses due to geopolitics and Trump's comments on Taiwan yesterday.

Furthermore, where we are in the Fed cutting cycle strongly suggests a likelihood of gains ahead, so long as we avoid any major negative catalysts. While summer doldrums in terms of volume can often result in large downdrops and swings of a negative variety, yesterday's drop seems to have been accelerated by an ongoing rotation that is largely positive and suggests a widening, not ending, of the ongoing bull market.