Just my opinion, but I think I would consider using historical data, perhaps munged by a coefficient to make it harder for a player to identify the historical b

Just my opinion, but I think I would consider using historical data, perhaps munged by a coefficient to make it harder for a player to identify the historical basis using a few trades. Market psychology is notoriously finicky and feels like it would be poorly represented by an algorithm to me. This would have the added advantage of simulating some price action based on things like news, earnings, current events etc. I think I would tend to pick gentle bull market period, or perhaps concatenate a couple of cycles to make a bull market->correction->recovery cycle. It all depends on what your goals are of course, but one bonus is that unlimited historical data is all out there and easy to get (so you could pick many periods or even have the computer pick them). A long time ago I participated in a stock market simulation that was run this way, at the end it was revealed what the actual stock basis was for the tickers and it was pretty interesting to be able to go back and look at some of the news around those stocks from the time period.