Let me rephrase a question I posed in another thread.we all use the Normal dist to approximate asset returns. In fact the very foundations of quant finance theory are built on it. The Normal dist is symmetric, i.e. it has the same probabilities on the up and down sides. But if I knew something about a security that was not known by the general market, wouldn't the return distribution be assymetric (skewed) for me?By the same logic, if there are n market participants in a security, they all probably have different levels of knowledge about it, no matter how small the difference. Wouldn't that mean that using the normal dist gives them wrong results (to the extent that they haven't incorporated their own specific knowledge into it). If there was a way they could, wouldn't it make trading more accurate?

But how would one know how much more one knew than the other market participants and therefore how to quantify this additional knowledge?Also common sense suggests that people would only enter into a transaction if they felt they were at an advantage. Therefore some participants in this particular scheme must genuinely be at an advantage whilst others are simply deluded (a normal distribution of these can be assumed - a large mean of realists with the immensely stupid at one extreme and the extremely savvy at the other). So to me a Gaussian distribution sounds like a fairly sensible assumption to apply to the returns that these people should expect to receive.Finally if the information that one possesses places one at such an advantage to the rest of the market, why bother with any distributional assumption as there is no question of it not making money. Unfortunately most of the information that places one at such an advantage is generally illegal to act on.