Serving the Quantitative Finance Community
normality
QuoteOriginally posted by: MuisCan someone plz tell me why the stock closing prices are expected to distribute normal.I do not think that the closing prices are expected to be normally distributed, but rather the returns, i.e., the relative changes from day to day in the stock price. Also, maybe it would be more accurate to say that the normal distribution is a convenient approximation for the expected distribution./TGN
normality
A stock price is not the sum of previous logreturns. If you write the stockprice as S_t = Sum_0^t delta S_i you would have that the delta S_i's have different distributions (depending on what the shareprice is at the moment), i.e., the central limit theorem does not apply./TGN
normality
Because you know nothing about which way they are going to go. Probability is subjective, and this is a good assumption for people who know nothing. Some of these same people who assume others know more than them about direction nevertheless assume they can predict volatility, because others knowing volatility does not enable them to price it in. Which makes a distribution that accommodates the sum of many smaller moves, and has no directional bias but known volatility, convenient.
Antonin Scalia Library http://antoninscalia.com
normality
i think farmer is correct, its because of lots of interaction between buyers and sellers and each contributingvery little (i mean not impacting the price), then all of these are random from market perspective and thenwe get the brownian-pollen-grain analogy of collisions add clt(for distribution) to that and you are there.thus modeling fractional changes as coming from normal distribution works decently.---------amit7ul