Thank you for the entry prices - Does your system pick these 10 trades independent of each other or does it use expected correlations to make it profitable? ( sorry for all the questions i've recently been using a lot of ratio spreads and have had really good luck with them but yours seem to be more complex and I'd like to understand exactly what your doing)
Each trade is independent. My system scans for positive expectancy spreads with current market prices as input. Every market seems to have some positive expectancy trades, I have to choose which ones to trade. For example, I didn't trade RIMM this week as expected profit factor for best spreads was around 1.3. It's much bigger for the trades I entered. Also, I limit my trades to around 4 legs, despite my program finding some good 8-10 legs spreads (although rarely - it favors small number of legs when searching). No problems with questions, keep them coming. I'm new to options (but not to trading) and they force me to think.
Based on historical volatility, or more precisely on a sample of historical moves of the underlying during N amount of days, where N is days to expiration at the time of the trade.
SO.... You are assuming a log normal distribution, computing the standard deviation and then what?? Your spreads have a Z shaped continuous P/L graph. How do you get expectation out of it so you can find those with positive expectations?? http://en.wikipedia.org/wiki/Expected_value Also, since options are priced by market makers using the same assumptions shouldn't most expectations equal zero or less??
Selling meat and buying otm skew... Your taking a position on the change in skew.. I hope you realize that.... HV tells you nothing..... Implies are forward looking HV is backwards looking... So your looking backwards to trade forwards literally.... Your big payout is on large jumps and skew increase... Hard to manage .... Keep digging.....
I assume that moves in the future are equally likely to the ones in the past. Nothing to do with lognormal distribution and standard deviation. If market moved up 10% in 10 days in the past, I assume I might move 10% in the next 10 days. That means about 0.957% a day. I also assume that opposite move is equally likely - 0.957% down a day for next 10 days. I look at last 10 years and build a list of possible moves. Out of those moves I calculate which spreads give me profit factor bigger than 1.0. I like spreads that give historical profit factor of at least 2.0. Out of those spreads I choose the ones with high probability, decent average win / average loss and reasonable median result. Sorry, I don't know what you mean by 'Z shaped' Well, I can't comment how market makers price their options, but I really doubt they use exactly the same method that I do. Also, I think market makers get volatility wrong all the time, but they hedge it one way or another. If it wasn't true they would either go out of business or it would be impossible to be profitable trading vol.
So... when you say ' My system scans for positive expectancy spreads ' you are not using the term 'expectancy' in it's statistical sense http://en.wikipedia.org/wiki/Expected_value which requires knowledge of distribution and probabilities. Nor are you using 'probability' in its statistical sense either. I am asking because I use expectancy in it's statistical sense to rank my spreads and am wondering if we are doing the same thing or different things. I think we are doing different things. If you look at my 'conservative options trades' thread you will see I list probability and expectation for each of my trades. I use a sampling method to get those numbers. I gather this is not what you are doing.