Quote from asiaprop:
fool, the market prices options exactly according to BS and then makes adjustments for skew, fat tails, and the like. There are a few French models out there, one of which was developed by a whale at the Deutsche index options desk in NYC but I dont wanna go into depth o tn that. Fact is BS has since its introduction been extensively used, everyone knows its shortcomings and everyone works around it because there simply are very few alternatives.
You cannot model rare events nor can someone with whatever amount of work experience predict rare events. You use the models until you get to the points where you know the models dont hold and then you manage risk in "manual" mode (meaning most of the time risk reduction) until the storm is over. Simple as that. Thats how some guys struck it rich. Other, more impatient natures, frequently blow up, get fired, suck the tits of head hunters and get their new job. Thats how it works and no politician nor crisis is gonna change that...
If you use Black-Scholes to price options and then make adjustments for the prices you don't like - then you are not using Black-Scholes.
To illustrate the point, let's work on a fictional concept we shall call "FPM" - or "Fair Penis Measurement".
Let's say I produce my FPM by following the strict rules. Take a calibrated ruler, place it at the base of my penis and then take a reading at the furthest point my penis reaches along the ruler to produce a measurement of 10.5 inches. That gives me an FPM of 10.5
Then let's say another fellow, yourself for instance, does the same thing. You aren't happy with the measurement of 3.5 inches, so you decide to break off the first 5 inches of the ruler and start again. You now give yourself an FPM of 8.5.
In this case, although we followed a similar process, you did not adhere to the rules of FPM and therefore your FPM of 8.5 is invalid.
Similarly - increasing the volatility input that goes into Black Scholes on the outlying options to produce a more acceptable output is not using Black Scholes at all which requires the same volatility input to calculate all strike prices.
Now - call me a fool again...
