If it's vol as synthetic time, wouldn't the actual passing of time mean the vol ramp is offset by theta decay? If so, wouldn't a pre-earnings straddle (regular or synthetic) be just a long gamma trade with slower time decay?
Right. My point is that synthetic time as if time is static. A theta-less position or a vol-arb. You're essentially isolating gamma if you buy it well.
The IV skews are pretty easy to confirm because you have the current pricing and I can pull them up in the option chain. I then look at the past 2 years of earnings vol activity to see estimates for vol crush in front and model what the potential cal spread would be post earnings with vol crush at certain levels. I prefer a skew of 20% minimum. The trick is not modelling the vols as it is strike selection. This is where the art comes in a little because I have to study the history of the underlying movements and current pricing to see what the risk is of a move outside of BE points. I still have reduced risk versus short straddle outright as worst case scenerio the cal goes to $0 so I know my maximum risk in worst case scenerio. Other risk is liquidity in the calendar due to closing two legs. As for distance, it depends on the skew. For PZZA i used two weeklies with the front month always attached to the earnings date. Back month in PZZA was 2 weeks later but normally I look at one month to retain value on the back month if the underlying moves too far. FOr me the art is looking at the stock and analyzing potential stock swing to select strike for best coverage. I got beat on WING as it jumped way too far but most don't that I look it. In other words I won't do it on GOOG.
hmmm, I am not quite sure why you are recommending a synthetic. The risk parameters are the same. You will lose all the ambient vol priced into the options. The key is to find an event that is not priced properly and hope it reprices correctly sometime in the near future. Theres a massive edge in predicting the earnings date before the date is actually confirmed. But I don't play to much in that sandbox. Do you have experience with that?
I do scans based on skew initially (like at least 20% skew) and then I look at historical IVs to see where current front month and back month rate versus where IV has been over time. I do not want a front month and back month both at highest intervals where I will get crushed on both. I don't model it to the higher degree Dest does, I look at the skew and then using the analyzer study the strike selections and adjust post earning vols to see what is most likely based on historical IVs.
The synthetic to limit size. 50 shares and a put = half lot. It's a cheap entry to having skin in the game. Yeah, IV_Trader and myself did the date-prediction with calendars a lot in the 2005-2010 time frame. I would caution anyone calling the date-prediction an edge; any more so than being right on a post-report move in price.
Des I'd like to work a problem out here so we can discuss it. Any stocks you have in mind that have earnings coming up and are interested in? I'll show you how I do my analysis and maybe we can find something