In Wilmott magazine, Mar-10, our partner and friend Brian Sentance from Xenomorph conducted an article, Not Right but Good, speaking to Yuval Milo, a researcher in accountancy, about derivatives pricing, social sciences, and how the model became the market.
As Andreas wrote in his introductory book Einführung in die Finanzmathematik : the risk-neutral measure, base of the Black-Scholes option theory, makes the option market a fair play. As long as it is applied as a kind of a game rule (my comment). And in fact in the late 1980ies the majority of the market partisipants seemed to play BS (formula), BS affected the market to follow it, as Milo stated in Brian's article.
But after the crash market participants realized the smile. It might have been an behavioral effect?
However, extensions first drove a model-method spiral with method and implementation traps.
We know that Heston and Bates extensions (stochastic volatilities and jumps) need careful treatment, especially when it comes to calibration - minimizing different error functionals (norms) for solving the inverse problems properly is challenging (Minimalism).
This might be the reason that BS has possibly lost the innocence of being a game rule.
If people have troubles to create the right price and risk profiles they forget the underlying principles and fundamentals?
We, at UnRisk, cannot influence the psychological factors, but we work really hard on the robustness of our model-method combinations.