When my mom asked whether she should sell some of her GRMN (up 16% yesterday after announcing earnings), I dug deep into the equity options data to see what happened. Were there any clues in the options market where I could have identified this trade?

It turns out there were signs in Garmin's IV term structure and in daily options volume. The volume was six times normal on the day before the move and the term structure was seven times as heavily skewed higher in the front months than normal, even after adjusting for earnings.

This adjustment for earnings is the crux of calculating the expected implied volatility after earnings and seeing the underlying creep in implied that occurred in GRMN in the days before earnings were announced.


In the table above, the day before earnings were announced, option volume was 15,400 total contracts vs an average of 2500 per day and contango at -1.56 vs -0.2 average. Contango is the normal state of monthly implied volatilities when the front months are below the longer dated months. Here, a negative contango or backwardation, is a bright red flag reflecting GRMN's front month volatilities were well above the back months even after adjusting for the earnings effect in IV.

In order to get a good look at the implied volatility term structure you have to strip out the earnings effect. The earnings effect is the one day move that usually happens after the earnings announcement. That one day move has a huge effect on the IV of options with only a few days to expiration and less effect on ones with more days left before expiring.

To isolate the ex-earnings implied volatilities, the trick is to iterate through earnings effects until a rational term structure appears. Finding that rational term structure involves constructing a formula based on the square root of time that matches the historical term-structure relationships in the market (a subject of another post). With a formula to fit the monthly IV term structure, the expected implied volatility after earnings can be found by iterating earnings effects until the difference between the rational term structure and the ex-earnings implied volatilities of each month is minimized.

With the earnings effect stripped out, you get a way to compare IVs over time and estimate where the implied might fall after earnings are announced and volatilities settle down.