One of the goals of automated trading is to enhance performance by selecting systems with an edge (which by definition implies a higher probability than luck)

 

Coin-Flipping For Fun and Profits.

 First, let's talk about luck.  Think about a nationwide coin-flipping contest.  At the beginning, each citizen is paired-up with another for a contest.  The winner goes on to the next round.  Think how many rounds you would need to win to be City Champion, State Champion, Regional Champion, etc. 

At the end, someone would have won a large number of coin-flip contests.  Assuming they didn't cheat, they were lucky … but does the winner have an edge?  If so, what could it be?
 
Finding The Edge.

Likewise, just because a trading system is making money doesn't prove it has an edge.  Simply relying on whether something is profitable NOW means you have both the chance that the system has an edge – and also that it is just lucky.
 
If it isn't just a matter of winning, how do know which are the best systems?  We are searching for clues to help find systems that do have an edge … or at least have an edge in certain market conditions.

We have many trading systems, and a lot them test well over long periods of time. In our research, we use the term "Bot" to refer to a particular trading system operating on a particular market and timeframe.  Recently, we've been working on a project to design a validation protocol to help us determine which Bots have the most "skill" and to filter-out the majority of "lucky" Bots. 
 
It is exciting as we solve more of the bits and pieces of this puzzle.
 
The Next Step – Dynamically Selecting Trusted Bots That Are Now In-Phase.

From a practical perspective, we aren't just trying to select trading techniques that have an edge …  In other words, we don't just need trusted bots … we need to select the bots that are most likely in-phase with current market conditions.  We'll talk more about this next week.