I am a relative newbie to options, and am trying to accelerate my understanding of what (practically) drives vol. I already have many of the requisite texts on options (natenburg, hull, derman, bennet, etc), but there is limited coverage on REAL WORLD examples of what can cause vol to move around. Take for example, an IV CRUSH following an earnings event. Initially, I assumed it would be as simple as looking at historical data and then using that to forecast X% drop. However, I realized a number of things can mess up my "simple forecast", such as: - DTE. I've realized that different DTEs see different vol crush impact. In some cases it is weird, like the specifically one DTE gets hit WAY MORE than others (eg. the second expiry). I had a calendar that got messed up because the term structure ended up with a super weird kink where one specific DTE (the one I was long) that was much lower than the rest of the term structure (was not able to explain it) - Spot movement vs strike. This depends on shape of smile (are we moving up or down the slope). And then there's sticky delta / sticky strike, we are shifting along the curve or did the whole surface shift. How do we know what's caused by spot-vol correlation vs earnings IV crush? - Weekends. I've noticed that friday vols drop, and this is made worse if its a long weekend. Being a newbie, I can only guess that people are cutting exposure before the non trading days. - Overall index vol movement. If index dumps, there could be a "beta" effect on the particluar stock's IV as well. So there's IV CRUSH (from earnings), but IV INCREASE caused by other macro factors. - I'm sure there are many more.... As seen from the above, there are multiple things can ADD / SUBTRACT IV, which makes things complicated (for me). Some questions: 1. Any other points above I missed out on? 2. Any practical tips for a newbie like to to sharpen my intuition? Any book recommendations that cover the above specifically (i.e. real world examples) 3. Am I over thinking? Should I just stick to "average last X periods earnings vol crush"? I already tried that and my calendar got messed up. Ideally I want to improve my ability to forecasts IVs. Thank you.
The main factor for volatility movements are press releases, reports, news. For example when a press release announces that in February the FDA will decide about the drug application of a biotech company. As soon as such an announcement is made public then the IV will begin to rise. And after the event finally occurs in February, IV will fall back to normal levels. For predicting volatility changes you would need to watch & study all the company-related news, as well monitor the daily HV of the stock and ATM IV of its options... One can say that volatility is about uncertainty in the operations of a company. IV is calculated from the Bid/Ask prices: by using MidPrice and for example iteratively applying the Black-Scholes formula...
You can read: https://onlinelibrary.wiley.com/doi/book/10.1002/9781119204510 Basic Option Volatility Strategies: Understanding Popular Pricing Models Editor(s): Sheldon Natenberg https://onlinelibrary.wiley.com/doi/book/10.1002/9781119204473 Option Volatility Trading Strategies Editor(s): Sheldon Natenberg
OP wrote that he has Natenberg. Both the above books by Natenberg seem to have the same content b/c their TOCs are identical, isn't it? Also the publication date of both is identical: 2 January 2012
dte vs vol of iv and weekends can both be accounted for using var_n_dte= var_day_1+var_day_2+...+var_day_n and then taking sqrt for get std/IV. all really variations of how much weight to apply to the volatility occurring on that day with earnings weighted highly and weekends weighted at almost zero. in fact if you calculate realized vol over every weekday, I found for SPY, the day that includes the friday to monday close has weights of 1.17 an average weekday realized vol, accounting for the additional information contained in the weekend. again for #3, the above formula can be used to get exact post earnings IV but you can use 180 dte iv as a rough first approximation