Basically, yes. They were likely getting hurt on weekly bets from what I believe to be a lack of gamma/theta hedges. It's difficult to price to gamma. They went from 7d on index to 15d; now I have heard it's 10d? The 7d lose approx 10% from first-day decay, not 15 as mentioned earlier.
got it , thanks , B. Ones I tested binomial vs B&S gamma/theta discrepancy (not a statement , just my observations ; not sure if I did it correct) and the diff were growing sharply when time to exp become 5-7 days and less.
Absolutely, I can't blame them from trying to stay afloat. Trees will always perfrom better when gamma > vega.
that's why I was sooo surprised to see Weeklies...If only I was an institutional trader...Easy 50% an year. Oh , well, life is still good for a retailer too.
In general, any kind of contingent payoff product, be it options, digitals, barriers etc. has a parity relationship that would allow any sane person to figure out the spread in vol/probability terms. In case of digitals, Touch + No Touch = Payoff is the main one. The other calculation worth doing is Option - Digital > 0 (except for Japanese rates, where binaries would be more expensive then caps). If people care, I can post similar rules for knock-ins/outs, spread options etc. As for pricing exotic structures close to the expiry date - in general, it becomes much more of a heding approach then anything else. I'd take any price coming out of a model with a big brick of salt. For example, if I am forced to make a price that close to the expiry date, I put in a bunch of non-model issues - pin risk, event risk etc.
Spread. Unfortunately (for me) I always MUST close position(winning or losing) one day before expiration. The 2$ spread on the long combo eating most of the profits. If vix/spx vols will ever rise to 20+ again , I will trade it.
Being on the OTC side of the world, I never actually got deal with risk-based haircuts. Could you explain a bit?