some silly/stupid/bothering qns: y cant we use high downside puts if we think iv's too high vs rv? is it because if spot ever goes there, it wouldnt be sticky strike. the variance risk premium will likely shoot up causing iv to spike as well causing vega loss but if over lifespan rv come in below iv (i.e. only one day big move but mild moves subsequently, poor autocorr)? will pnl still be +ve at the end? what about upside calls? not advisable because higher leverage to get same amt of vega? if view is that there is not going to be any sort of huge rally, wouldnt pnl be iv - rv? similarly why dont we use the strike at the trough of the vol smile to long vol? thks.
DdeltaDvol and DvegaDvol kill you when it moves down to short puts. Realized can come in below and you can still lose money. You can use options wherever to bet on rv vs iv but you need as much theta/gamma as possible and atm offers the most of it. The problem if you use otm/itm is that once it moves ATM theta/gamma will be much larger than you anticipated increasing your vol exposure... likewise once atm moves itm/otm it decreases exposure. So you have to keep rolling options in addition to deltahedging underlyer which has serious friction.
thks for your explanation on using otm options, TskTsk Yeah i had a similar concern about trading constant maturity and atm strike as well. or using these to compare cheap/dear as passage of time will render the historical comparison less valid Here were some questions I had with using ATM strikes in the past. On entry, the strikes are fixed at atm fwds. Subsequently, even if the shape and vols levels are unchanged and the vols decay with each passing day, when fwds moves, that particular strike vol might slide up/down the skew. Do you restrike or rebalance? The more general qn is when you use constant delta/moneyness and constant maturity to compare with its past history, how do you implement such a trade during its lifetime?
Not a stupid question at all, actually. -- general thought is that if you are delta hedging, your expected p&l will be sum of Gamma(t) * (RVar(t) - IVar(t)), which means that every day you going to be realizing the moves and weighting them by gamma -- in a simple form, the above means if you are trying to play pure RV-IV, you would want to trade where your gamma is highest, which is the current level of the forward (the expected value of the stock) -- this also means that your P&L to term will be inherently stochastic, you can have the right call on RV vs IV and still lose money on the trade and vice versa -- if you are trading an OTM strike and holding it to expiration, in addition to the volatility view, you are expressing the view on the path of volatility as it will be weighted by your gamma -- here is a little statistical study you can do in your spare time using just the history of short-dated ATM vol and the history of spot for SPX. Regress RV(t0, T | T) - IV(t0, T | t0) ~ S(T)/S(t0) - 1. If you can get access to ATM vol history, you can use VIX as an appoximation - it will change the intercept of the regression, but will not change the overall relationship of volatility to spot returns.
From the perspective of the single strike, you don't care if the vols move due to overall change in level of vols or because it's a restrike of the surface (that is, "slides along the skew"). It's still vega P&L and now you got an option with new implied vol. If it's rich, you sell it. If it's cheap, you hold it. I am not sure what you mean.
thks, sle. took some time to understand you eqn and prob thats y its "stupid" qn hahah. i think that its also highly likely that ATM vol will slide up the smile when fwd moves quite a bit over the lifetime and stay there or jump suddenly and non sticky strike vol moves thks for the tip. not sure if i can attach image here: c:\temp.jpeg i always thought that there is a var risk premium that can be quite consistently harvested but it really depends on wherever mkt is up/down.
yeah so there is a mtm vega loss on an existing trade that shorts IV/RV using the previous ATM strike. And so eval whether to just hold this strike and/or add a new atm vol. research usually plots all kinds of constant maturity research. skew, vol, iv/rv, etc. sometimes they show that its mean reverting so that you should long/short this metric. but its assuming that the position that you hold is always a constant maturity which I think doesnt happen outside of rates world (ETP aside). how do you monetize this constant maturity metric mean reversion then as initial position ages?