Sunday, 8 January 2012

Will My Setup Work For The Next Trade?

This is the question that a trader needs an answer to. Without an answer to this question, you are whistling in the dark. All you have is hope and fear, and that leads to losing.

So whether you are a discretionary trader or an algo trader, testing is mandatory if you want to be CP. And the testing has to be right. What is "right"? "Right" is taking a large sample size in an objective (not subjective) manner of how you enter and exit trades, so that they can be measured.

This process involves, guess what?, a Trading Plan to start it off. Not just any kind of TP, but a very specific one. A description of how you trade that you could hand to another trader and he could make the same trades you did.

Once you have that, you have something to test. Without that, you are guesstimating and that is like that famous old (now dead) comedian, HOPE. which is what happens to most traders that don't have a TP. They're cactus as we say in Oz.

The most secure way of testing is by computer. Run hundreds of trades through it so you have a good sample size and know what to expect. Part of the output of your testing should be an equity curve so you can see at a glance how the sequence of trades looked. Having a walk forward result in your equity curve is a plus. 

Chart below shows how well these Renkos cut out noise. The fact that they don't have wicks is the price we have to pay. They work best with the most volatile markets.

No comments:

Post a Comment