Non Random Walk.com

Trade-A-Stock Help: Randomness

 

What’s with the monkey?


The chimpanzee (an ape, not a monkey) serves to illustrate an important concept. The best way to evaluate whether your insight and analysis is giving you a meaningful advantage in trading is to compare yourself to a group of traders that are randomly making decisions. This technique allows you to evaluate whether you are actually trading well or if you are simply ‘just lucky’ or ‘fooled by randomness’ as the efficient market people espouse.

Comparing Your Performance To Randomness


"In a bull market, everyone thinks they are a genius", is an old stock market euphemism. This is because most people have a "long" bias (they tend only to buy stocks hoping/thinking they will appreciate in price ) and almost all stocks go up during a really raging bull market. Click here is you don’t know what a bull market is. However, were you to make money in such a market, it really doesn't mean you were a good trader. Imagine a group of chimpanzees randomly trading the same market, only allowed to trade long, entering and exiting without thought. Surely many of them would perform well. Many would outperform you, just by chance. The only way to be confident that your decision making gives you an advantage is to measure yourself against that group of monkeys. If your performance is consistently better than the majority of the monkeys, then you can be confident that you are not deluding yourself or being 'fooled by randomness'. Of course, doing this is on a single stock is not enough, you need to consistently outperform the monkeys every time you go through the Trade-A-Stock exercise.


Calculating The Chimp Simulation


Click here for a definition of long and short.


For the "Long-Only" chimpanzee simulation: The application simulates each monkey randomly trading the same stock you did for the same time period. On each day the chimp will randomly choose to be either long the stock for that day or to not be long (it cannot go short), with an equal probability of doing either on every given day. In other words, on each day there is a 50% chance of the monkey having a long position in the stock. The application uses a pseudo-random number generator to mimic the monkey's decisions. The application calculates the total return for all of the monkey's trades during the exercise in the same manner as for the user. This process is repeated for 1000 unique monkeys. Each monkey's total return is put into a bin that produces the histogram used below.

















For the "Long & Short" chimpanzee simulation: The application simulates each monkey randomly trading the same stock you did for the same time period. On each day the chimp will randomly choose to be either long or short ( it is always trading the stock ), with an equal probability of doing either on every given day. In other words, on each day there is a 50% chance of the monkey having a long position in the stock. If it is not long the stock, then it will be short. This simulation produces some monkeys that trade extremely, even shockingly well, but you must remember that there are some monkeys that will time the stock near perfectly, trading short at just the right time and long at just the right time. This will lead to fantastic returns. Consider, a stock that simply went up 1% each day and down 1% the next. Over the course of a year it would be generally quite flat, but there might be a monkey who happened to time its trading  perfectly, going long on each up day and short on each down day. Its total return would be . . .


1.01 255 = 12.65


. . . or an over 1265% return over the course of a year. ( Using approximately 255 trading days in year ). While the probability of the above event is rare, it becomes a bit more comprehensible for a monkey to get a 200% - 300% return in a volatile stock.


Click here for further instructions on how to interpret the monkey-histogram charts.

 

A chimpanzee, strictly speaking, is not a monkey. We are aware of this but use the terms interchangeably.

Next Help Page HelpReport.htmlHelpReport.htmlshapeimage_2_link_0

Questions? Comments? We’d love to hear from you. Contact Us