Quote from dmo:
Besides, portfolio protection is not just a matter of puts or calls - much more importantly it's a matter of premium - gammas and vegas. If futures are at 1200 and you're worried about a drop, you don't want premium up at the 1300 strike - that's not going to do you any good in the event of a crash. You want premium down low, the lower the better. If there's a crash, you want that premium to be right in the heart of the action. Think of it as real estate - location location location. In the event of a crash, the best real estate to own is at the low strikes. In a crash, premium down low becomes a Park Avenue penthouse during the great NY real estate boom. Premium at high strikes might as well be a tenement in the Bronx - it's just not situated to benefit from the action.
Interesting perspective on volty as function of strike.
Here is my "theory/law" on it. If the market goes down faster than it goes up, one would expect (short term) volty to be higher at strike ATM-1, and lower at strike ATM+1. (you can replace -1 and +1 by -epsilon of + epsilon). Volty at strike ATM-2 should be greater than volty at strike ATM-1, because on the way down if underlying reaches ATM-1 strike, strike ATM-2 will be its immediate next down strike, and the at the money reasoning applies. By the process of induction, Volty should be increasing at lower strikes. The induction can alternatively be viewed as an anticipatory process in determining the right voltys. What do you think of this "theory/proof/law" of skew?