The stock market is described in many places (e.g. here) as a random walk. Among the various arguments against this hypothesis, nowhere is mentioned what struck me to be the most obvious one: every stock has a lower bound of 0. A random walk, however, is unbound.
I can think of various ways to implement such a bound in a simulation that generates artificial stock market data. But are there ways to simulate a stock market where a lower bound of 0 follows from economical considerations (i.e., a market theory) instead of practical improvisation.