Repository of common IV patterns

[Fox]

Hi all,

The ability to predict IV will give you a head start when trading options. I thought it useful to start a repository of common implied volatility (IV) patterns. This can then be used as a reference handbook for trading decisions or for searching trading opportunities.

Please feel free to add your favourite pattern to this repository as we go along. Let's start a convention whereby the contributor has the privilege of naming the IV pattern.

I'll start with the She'll Be Fine pattern. This is a well documented pattern whereby a steadily rising underlying leads to a falling IV. One of the reasons given is that complacency sets in, fear recedes and demand for options dwindle.

Typically, when the underlying starts to fall rapidly after the steady rally, fear sets in and demand for options increase, leading to increased IV.

The attached chart of the recent dip in XJO after a long steady rally, shows this clearly. XJO plunging around the end of October led to a rise from a low of 20% to a high of more than 25%.

 

[Fox]

Another pattern I've observed is what I call the Announcement Anticipation pattern. Every earnings report season, IV will steadily rise leading to the day of announcement. After the announcement, IV will fall like a rock.

One reason for this pattern could be that option players will put on the classic long straddle strategy as recommended by many option books. The idea is that uncertainty in a company's earnings report could affect the share price either way and a straddle will help you profit in either outcomes. So these players demand options and push up IV.

Shareholders will also demand options during this period. Nervous investors would probably want to buy puts for protection in case earnings is worse than expected and the optimists will want to buy calls in anticipation of a better than expected dividend. All in all, IV rises leading up to the earnings announcement day.

After announcement, all uncertainties are removed and demand for options fall off.

 

[mazzatelli]

She'll be fine pattern, *tends* to hold with moves > 1 sigma.
It's different predicting iv as opposed to using iv to predict future realized vol

 

[mazzatelli]

Katja Ahoniemi's paper - Analyzing iv
For high level overview, pattern recognition is identifying serial correlation and exogenous regressors