I'll join the crowd and would also appreciate if someone could send me this paper to frolloos@yahoo.comthanks.

who has the paper? do you Tagoma?

@Analytical Vega, can you please share the paper ?

@Exsan Actually, I retrieved it in an oldish email account of mine.

QuoteOriginally posted by: Nimbus3000@Analytical Vega, can you please share the paper ?There is a possible legal issue with Morgan Stanley's lawyers claiming it may be IP.

As far as I understand it, the Arslan et al paper *assumes* that gamma, vanna and volga have associated theta costs. Is there a way to prove this? I mean in BS framekwork gamma has theta cost, and in say a stoch vol model you could argue theta cost is indeed gamma volga and vanna, but not sure how to abuse the BS framework correctly to allow additional volga and vanna theta costs, if this makes sense.I also found during implementation some dependency of the thetas on the choice of pillar options, anyone running into the same issue?

I've tried implementing the paper, but find that with alpha <=0.5 the far otm puts are systematically too low (not just the short dated regulatory stress puts, but say anything <10delta and <9months). However when alpha>0.5, I frequently encounter the case where there are no real roots for low strikes. Has anyone else encountered this? Any suggestions? I very much like this framework and would like to use it 'naturally' without the need to hack it into some sort of piecewise function.

QuoteOriginally posted by: frolloosAs far as I understand it, the Arslan et al paper *assumes* that gamma, vanna and volga have associated theta costs. Is there a way to prove this? I mean in BS framekwork gamma has theta cost, and in say a stoch vol model you could argue theta cost is indeed gamma volga and vanna, but not sure how to abuse the BS framework correctly to allow additional volga and vanna theta costs, if this makes sense.I also found during implementation some dependency of the thetas on the choice of pillar options, anyone running into the same issue?Implementation works. I used gamma, vomma and vanna initially instead of dollar gamma, dollar volga and dollar vanna. Now there is negligible dependency on choice of the 3 pillar options, which is good. That said, I still don't understand why this method works, and why it only works when setting r=q=0, even though the smile taken from the index market obviously doesnt assume this. The interpolation is very good, and the put extrapolation also, where I can extrapolate all the way down to K = 50% for t = 0.5 and still match IBD quotes. but the call extrapolation is sometimes off for short maturities for instance t = 0.5, K > 120%.@mcakes: which paper are you looking at? The Arslan et al paper doesn't mention an alpha I believe. Like you I also like this framework by the way.

The Giryavets paper, which is more or less the Arslan, Eid idea, but framed in a (imo) more intuitive fashion (and perhaps gives more insight into why it works).

Ah ok, I don't have the Giryavets paper so I have no idea what the alpha means there.

Could someone kindly send me the paper ? That's a topic I am really interested in.

The paper is interesting, maybe the author will pm you and send it. 

The gist of it / the basics you can derive yourself though, starting with the observation that 

[$] E^Q [dC] = rCdt = rC^{BS} dt [$] 

Then express [$] dC [$] in terms of Black-Scholes Greeks. Once you've done that you have an equation for implied volatility. If you then assume that the correlation between implied volatility and the stock/index is constant across strikes, and also that the dollar gamma, volga and vanna are constant across strikes you have basically a way to calibrate the smile using 4 pillar options (or 3 if you assume the drift of implied vol is zero). This is in essence the so-called GVV cost frameworkby Arslan et al on which the paper by Alexander Giryavets builds.

There are quite a few assumptions here and it has been shown that these assumptions actually can lead to arbitrage. That said, it is conceptually interesting and I believe still an active field of research.

@alexandergir, can you please email me the paper original paper that was on SSRN on klhzim78atgmail.com

Could @alexandergir or someone please send me a copy of the paper?  I would greatly appreciate it!  ian.patrick.reid@gmail.com  Concept sounds intriguing and the consensus of all those that do have it and have posted on it is very positive.  Thanks so much!