I've been calculating the expected payout of Iron Condors. I cant see how it could be profitable over the long term becuase there is no way that an MM will give you an IC with a positive payout expectation. Here is an example of an 80% probability 10pt IC traded from today's market. S 22NOVRUT P 1060 B 22NOVRUT P 1050 S 22NOVRUT C 1150 B 22NOVRUT C 1160 The spread price is $2.00. The probable payout will be probability x premium. .80 x $2.70 = $1.60 The probable loss is probability x (spread width- premium) .20 x ($10-$2)= $1.60 With these expected payouts, there is no way you can come out ahead over the long term. Even with changes in IV's, you just get pushed out to farther strikes. Plus, the MM is too sophisticated to give you a favorable payout. Is there any reason why he would? Please tell me if this math is wrong. Thanks.
Your math looks accurate to me. A few years ago I was looking at the expected payout on a credit spread and noticed the same thing, that the expected payout is zero before commissions. Essentially, if we're just using straight probabilities we can expect to pay commission by trading credit spreads over a long enough time period. I think about everything in terms of: Expected Outcome = P(winning)*avg winner - P(losing)*avg loser If we drop your numbers into that (as you've effectively done), we end up with: 80%*200 - 20%*-800 = 0 (ignoring commissions) The whole trading part of the equation is that people trading IC's are trying to influence some of those variables in a way that creates a positive outcome. For example, if you say that you will never take the max loss, your equation could look something like: 80%*200 - 20%*-400 = +80 The challenge becomes knowing whether or not the same probabilities apply in the second situation. My feeling is that they change, but I can't provide anything concrete (like backtesting info) to back that up.
I guess thats where edge comes into play. So something as simple as " I will close the spread, when a wing gets touched "should" result in a more positive outcome." That should be .80 x $ 2.00, $2.70 is a typo. But as I stated earlier, the spread is 2 bucks. Sle, That 80% is the probability that underlying will touch either the NOV RUT 1060P or the NOV 1150C. I got the strikes by approximating with deltas. I could be more precise by using the brokers Probability calculator for 1st std dev (68%) and 2nd std dev (95%), but i don't hear a lot of people trading those probabilities. Nonetheless, lesson learned: trade a wide spread and the MM will rape you.
Yeah. If you use the distribution implied by the options price for your probability calculations then your expected value will be the bid/offer you pay. It's like buying a red sweater and then complaining you have a sweater that's red. The way you make money is when you say the probability is different from what the market is implying, and different enough that you can pay the bid/offer.
I would think it might be smart to try to use the trend direction with tech analysis to add to your decision making with spreads. Or lack of trend (ie, flat price), in the case of iron condors.
Trading, yes spread width minus premium is max loss. New, I definitely agree. A change in iv greatly affects valuations. I guess looking at ic selling in a binary fashion is too simplistic. For example, if I always close the if when a wing gets touched then I will never realize max loss as the formula implies. Nonetheless, it does show the incredible edge that the mm has over retail folks.