On one of the most actively traded equity options, GOOG, the market makers "incorrectly" priced what vol to sell the straddle at prior to 3Q10 earnings as GOOG ripped higher. While I do use the ATM straddle as a guide as to where the stock may move, by no means is it readily predictable of the underlying's move as market makers are *gasp* humans just like you and me and make errors too. BTW in this instance it was not just market makers, but the entire options MARKET which was wrong about GOOG's volatility post earnings. So to say that "making money is pretty pointless since the options will have that priced in" is a bit of a farce to say the least.
1) Market makers don't "price" nor "control" the market. They merely "lean on" other people's order flow. 2) The "market" does a consistently good job, not a perfectly good job, of pricing the options before major reports. According to "Bell Curve math", 68% is a great "average". 3) If GOOG were to release its earnings a few weeks before option expiration instead, the average would be higher. The "outliers" would not affect the front-month options as much.
Thanks for the clarification Nazz. But I think my point still stands with regards to buying the straddle before earnings as moves greater than 1 sigma do occur. Granted, you take on lower odds of success but I do believe that the straddle can pay off (especially in biotechs) such that it is a viable trading strategy. It's not something I would buy on GOOG prior to every earnings announcement, but in certain scenarios like the OP outlined they could very well be justified. That being said I would rarely put my money in a straddle because I'm not going to take a pure volatility bet like a straddle implies with no regard for direction. Instead I would prefer having a significant delta exposure one way or another and either swing for the fences or strike out.