Forums  > Basics  > Implied Vol Model - impact on Delta  
     
Page 1 of 1
Display using:  

hilss


Total Posts: 55
Joined: Jun 2007
 
Posted: 2017-03-07 17:50
Hello all,

I have 3 scenarios:
A) Black-Scholes assumptions (constant vol)
B) Linear strike-space (skew only model): Vol = AtmVol + (Strike - FuturePrice) * Slope
C) Percent of FuturePrice skew only model: Vol = AtmVol + (Strike - FuturePrice) / FuturePrice * Slope

For simplicity, I don't want to add standard deviation model or a delta-based space model because the idea is roughly the same (they become dependent on time and vol as well).

Relative to the Black Scholes model, if I'm long options, I will get shorter delta if I use B, and even shorter if I use C.

Given your experiences, is C more favorable than B in terms of replicating the real markets more closely? More specifically, when the underlying moves, will my theo price behave closer to the market if I used C over B? I suspect that different products behave differently. When our theo delta is different from the "real" delta (which I'm not sure I know what it is), what are the consequences of that? If we are delta neutral on average (based on our theo delta), I suspect it will cause PnL Variance. What other effects should I expect?

I have one customer who is preparing to be a market-maker in the FTSE and keeps telling me that his delta is different from others in the market and is different from Bloomberg. The Bloomberg delta (I assume black scholes) is closer to the market than his delta. When his brokers ask him for a tied-up structure (K 7450C vs. 30), the delta is almost never given (unless it got traded recently and we are not too far away in time and future price, then we can have a rough idea of what the delta is). So when it comes time to trade, his delta is off compared to other market-makers. Sometimes it's in his favor, and other times it's not. Needless to say, it's bad if he keeps complaining about his delta being different (when it's not in his favor). Do you think this will average out: there will be times when the trade is in his favor vs not in his favor? Order flow and market conditions will play a role, but will this average out over time?

Sorry for the long-winded message, but simply put, is it more desirable to have a model that is percentage-based (or delta-based)? And what are the effects of having a different delta than others?

Thanks,
hilss




Patrik
Founding Member

Total Posts: 1331
Joined: Mar 2004
 
Posted: 2017-03-07 18:26
The "better model" question you can estimate an answer to for historical periods by simulation using different models and/or hedging strategies. Looking into the future you can obviously never know, you have to take a view if it's likely to behave similar to history or not and so forth.

The effects of being different is that you want to ensure you know how to model market delta even if you don't believe in it. Then you can skew your quotes depending on the diff as prices moves away from the laid up price, so that it makes sense to where you think value is given your (in your opinion) superior model.

Capital Structure Demolition LLC Radiation

hilss


Total Posts: 55
Joined: Jun 2007
 
Posted: 2017-03-07 21:35
Thank you Patrick. Understood.

But given that I have no historical market-data, could someone shed some light given their experience?

Thanks,
hilss

Patrik
Founding Member

Total Posts: 1331
Joined: Mar 2004
 
Posted: 2017-03-07 22:15
I'm not an equities or FTSE specialist so can't help you there I'm afraid.

Capital Structure Demolition LLC Radiation

Baltazar


Total Posts: 1758
Joined: Jul 2004
 
Posted: 2017-03-08 00:29
This could be a good starting point for you

http://screpey.free.fr/papers/QMF04.pdf

Qui fait le malin tombe dans le ravin

hilss


Total Posts: 55
Joined: Jun 2007
 
Posted: 2017-03-08 00:45
Thank you... indeed that is a good start... I read it quickly, I will spend more time reading it carefully...
Thank you both.
hilss
Previous Thread :: Next Thread 
Page 1 of 1