It is a well know "fact" that calendars increase when volatility increases. This however is the furthest thing from the truth. The back month vol has to increase by more than root time for IV to have a positive impact on the calendar. OR the front month has to decrease by more than root time. After finishing "dynamic hedging" by taleb last night I have come to the conclusion that the calendar spread is a very complex strategy. There are many factors that influence the position such as vol of vol, skew, shadow gamma/vega. It is not an easy task and is giving me some head aches. None the less I will update on here my findings.
The term "calendar spread" is too vague without a trade example and real quotes. Once an example is posted everybody is on the same page and should be able to visualize the outcome of such a trade. Without a trade example then "calendar spreads" are very complex.
Well, RV vs IV in the front will trump the vega most of the times. What you want to do in a calendar trade (ideally) is predict the relationship between realization of gamma and dynamics of the forward vol.
realization of gamma? = IV - RV? If yes, are you recommending I delta hedge the calendar? dynamics of forward vol? = How low or high forward vol is relative to other strikes/history? Forward vol = sqrt( (V2^2*Dte2 - V1^2*Dte1) / (Dte2-Dte1)). Just to make sure I Have it right.
If you have a crazy IV of 75 or 100 just before a clinical trials release, would you be willing to pay that kind of price for a strangle or a straddle, knowing that post news release, IV might revert to a historical range of say 20-30 ?
Regardless of the vols, the further out premium will be greater. If a binary event occurs with a large move in either direction, the respective options may reach parity. In that case, do you want to be long or short the higher priced option? In addition, when vols get crazy, further week/month time premium expands more in the further term option. Vol crush will benefit the longer term more. As stated previously, all of this blows up if there is no large binary move.
What would be the percentage in drop in options prices for an IV collapse from say 80 to 40, or from 50 to 30? Would that vary across strike prices or expiry dates?
Depends on the traders view of the underlying. If he expects the move to be more than enough to make up for the debit paid then a long straddle/strangle would be a good idea.
I read @Secret Santa post as lack of movement (for long calendar) and increase in forward vola than predicted as per your equation