Should a sophisticated model like local vol be for pricing futures and vanillas?

A

A modest quant

Joined
7/15/14
Messages
31
Points
18
Hi quants and traders, i've been trading futures and vanillas and almost never thought of using sophisticated volatility models like local vol for pricing. Should such a model be used for pricing vanillas and futures?
 
Agreed with CasanovaJ. For vanillas local vol is fine by definition. Just don't use it for path dependent exotics.
 
financeguy said:
Agreed with CasanovaJ. For vanillas local vol is fine by definition. Just don't use it for path dependent exotics.
Click to expand...
Perhaps i should have layed down some ground before i post my first question. As a trader, when i mark prices vanillas (or vanillas on futures), i don't use models like local vol which, as i understand, quants use it to price exotics. How do i mark vanilla prices: i use the BS formula which is just a conventional tool to help viewing the general level of volatility (the implied volatility), plus i might use historical prices to get some reliable information and intuition to make prices. However, i would never know the actual volatility of underlying (futures in this case). Perhaps my first questions could have been: Where is the volatility of the underlying (futures price)? What is the implied volatility, really? What is the connection between these volatilities?
 
I'm going to consider this as a genuine question rather than an excuse to say something inflammatory. The implied volatility of an underlying is the market's forecast of the standard deviation of the returns of the decreasing maturity forward of the underlying to the expiration date as measured by any fixed or variable time interval, assuming lognormality of the asset. If you for whatever reason disagree with the market's assessment of the value of this measure of asset volatility as implied through inverting options prices using the B-S formula, then deal an option and trade its gamma how you think is most profitable. If the actual realized volatility ends up being different from what the market had implied when you dealt, then you should have either made or lost money. There's no reason why you should need a measure of volatility to price the linear underlying itself. Local volatility takes a volatility smile as an input to price european style products. That includes vanillas. To mark a vanilla smile yourself to begin with, get the prices of at the money options and two or four other points on the smile and then either just use a spline (or another numerical interpolation-extrapolation method) or sabr (or another model based smile).
 
financeguy said:
I'm going to consider this as a genuine question rather than an excuse to say something inflammatory. The implied volatility of an underlying is the market's forecast of the standard deviation of the returns of the decreasing maturity forward of the underlying to the expiration date as measured by any fixed or variable time interval, assuming lognormality of the asset. If you for whatever reason disagree with the market's assessment of the value of this measure of asset volatility as implied through inverting options prices using the B-S formula, then deal an option and trade its gamma how you think is most profitable. If the actual realized volatility ends up being different from what the market had implied when you dealt, then you should have either made or lost money. There's no reason why you should need a measure of volatility to price the linear underlying itself. Local volatility takes a volatility smile as an input to price european style products. That includes vanillas. To mark a vanilla smile yourself to begin with, get the prices of at the money options and two or four other points on the smile and then either just use a spline (or another numerical interpolation-extrapolation method) or sabr (or another model based smile).
Click to expand...

I am a trader myself and i know what implied vol is. Please donot give me a lecture on implied vol.

It was my genuine question after having thought about it for a good long while. You said "...The implied volatility of an underlying is the market's forecast of the standard deviation of the returns of the decreasing maturity forward of the underlying to the expiration date as measured by any fixed or variable time interval, assuming lognormality of the asset...", but you certainly forgot that implied vol is strike- and maturity-dependent whereas the spot volatility (0r standard deviation of returns) of the underlying is not. Therefore implied vol as forcast of standard deviation of returns is an absurd notion!

I always want to make some realistic sense of whatever i or other traders do in practice. A real good practical sense, not to some textbook interpretations which are quite often absurd and nonsense in this new real world.

No more nonsensical explanations ever again, thank you very much!!!
 
Back to question,
Where is the volatility of the underlying (futures price)? What is the implied volatility, really? What is the connection between these volatilities?
Click to expand...
It is my suggestion that look for options expire in same time period that you study vol of underlying in.
i.e you would like to know 30 days vol of underlying, look for options expire 30 days.
Then, collect only out of money options information as they reflect time value only.
The weighted average of price of these options should be closer to vol of underlying. The weight is set to liquidity.
 
StevenChen said:
Back to question,

It is my suggestion that look for options expire in same time period that you study vol of underlying in.
i.e you would like to know 30 days vol of underlying, look for options expire 30 days.
Then, collect only out of money options information as they reflect time value only.
The weighted average of price of these options should be closer to vol of underlying. The weight is set to liquidity.
Click to expand...

StevenChen,

Suppose that there are no options markets (collapsed or disappeared for some reasons), i.e. no implied volatilities, where is then the volatility of the underlying or the associated futures price?
 
A modest quant said:
StevenChen,

Suppose that there are no options markets (collapsed or disappeared for some reasons), i.e. no implied volatilities, where is then the volatility of the underlying or the associated futures price?
Click to expand...
Why do you need volatility to price futures? Futures are usually considered to be martingales, just use EV.

Conceptually, I think of implied vols as nothing more than a price of a contract where you're buying realized vols
 
euroazn said:
Why do you need volatility to price futures? Futures are usually considered to be martingales, just use EV.

Conceptually, I think of implied vols as nothing more than a price of a contract where you're buying realized vols
Click to expand...

Steven, i don't think you ever traded anything, did you?
 
Anything that moves randomly in the market has volatility. What do you use such info for shows how articulate you are in trading.
 
Yes, but the volatility is already incorporated into the price of the underlying. To price futures you need only three things: Price of the underlying (So), storage costs (r) and convenience yield (d) (to incorporate the possibility of backwardation):

Then: \(F_0=S_0*e^{(r-\delta)}\)
 
diegosanaz said:
Yes, but the volatility is already incorporated into the price of the underlying. To price futures you need only three things: Price of the underlying (So), storage costs (r) and convenience yield (d) (to incorporate the possibility of backwardation):

Then: \(F_0=S_0*e^{(r-\delta)}\)
Click to expand...

Diego, when you wrote that equation, do you know what model was assumed for the underlying?
 
This is the thing, that formula is a simple price that tries to avoid arbitrage. And, since the replication of a forward contract is trivial, there is no need to assume a model for the underlying.

Ie, even if it is GBM, Heston, or some jump process, that formula still holds.

Of course, the big issue here is r and \(\delta\). Those aren't simple numbers to obtain.
 
Back
Top Bottom