3
$\begingroup$

Many research papers use differing solution methods to attempt to find the 'true' value of an option whether it be Euro, American, etc.

They never mention how they do find the true option value to test against however. Which becomes my question: how does one find the true value of an option for given input variables to test against?

Appreciate any help.

$\endgroup$
  • $\begingroup$ What is your model? Is there a close form solution for options in your model ? $\endgroup$ – user16651 Dec 25 '16 at 20:46
  • $\begingroup$ Without any dynamics we can't speak correctly. $\endgroup$ – user16651 Dec 25 '16 at 20:50
  • $\begingroup$ This is specifically regarding American Puts for which there is not a closed form solution, as of the work I've looked into. With regards to dynamics, that is an input in itself surely? If I was to say I was using risk neutral dynamics would that really help considering I'm trying to approximate a true value with those assumed dynamics. $\endgroup$ – mwtmurphy Dec 25 '16 at 21:01
  • 1
    $\begingroup$ @ Will Gu I don't agree. $\endgroup$ – user16651 Dec 26 '16 at 12:05
  • 1
    $\begingroup$ Other people have noted it above too, but the "true value" of anything is what you can sell or trade it for. The moment you start talking about statistics or probabilities, you are speculating. $\endgroup$ – barrycarter Dec 26 '16 at 17:12
3
$\begingroup$

Basically it boils down to this:

You either use a descriptive or a prescriptive (normative) model, i.e. you either think that the market is always right or you think that you alone know how to determine the "true" price of an option.

The original idea of BS was to build a prescriptive model but most modern models try to take the market prices as given and calibrate their models accordingly, thereby gaining one common variable to make all prices comparable: volatility. One prerequisite is to seek consistency for all prices which is normally done via hedging arguments (which often leads to risk-neutrality assumptions). Yet in practice even in the case of minor inconsistencies they often cannot be used to make money due to friction.

At the end it all boils down to your business model (this is what I meant with my question about the true value of a stock). When you are on the buy side you try to find mispriced options with a prescriptive model (so you are saying that you know which is the true price - and all other market participants are wrong). When you are on the sell side you use a calibrated descriptive model which mainly seeks consistency so that you can live on the spread.

(This is why I gave this link in the comments: How do we use option price models (like Black-Scholes Model) to make money in practice?)

So in a way truth is relative: It either means overall consistent with the market or it means smarter than the market.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.