Thursday, 5 August 2010

Lies, More Lies and Statistics - Part 2

Following on from Part 1, I am looking at the process of reaching CP using a structured although completely discretionary methodology. In order to know what to expect and the boundaries of the structure, I need to collect information. It has been said: "Those that ignore the past cannot see the future". While I am not in the business of predicting prices, I do need to know what the outside boundary is for me to know when I am wrong in a trade (maximum adverse excursion) as well as the level where 70% plus of my trades should go.

These two are the basic metrics I need to capture in order to trade. In addition, I need to look at the Maximum Favorable Excursion, not only because of my first logical scale out point but also because of the subsequent scaling requirements.

I go about capturing this information from both back testing and live trading. There are lots of additional tools to use to look at this information, from simple spreadsheets to great applications like Market System Analyser (MSA) from

How we go about the process will be continued in Part 3.
Euro FX was awesome again today. The ECB made their statement about interest rates - sharply UNCH - but we had a great trade before the announcement and afterward as well. See the chart below.

The Gap trade on the ES was a classic even though it didn't reach its target. So all in all, a nice green day.


  1. Hi EL, Thanks for a great blog! I'm learning a lot.
    My question is this - when considering risk/reward, do you think of this in terms of probability or money? e.g. Dalton et. al. suggest a trade should have at least 2 points reward potential to 1 point risk with at least 50/50 chance of success (eg). BUT if you risk say 2 points with a 70% chance of at least 2 points reward (implying 30% chance of 2 point loss), then to my thinking, that is pretty much the same thing - actually better with the law of large numbers playing out. Even with the odds closer to the 1:2 ratio, say 35% loss 65% win with even points, youre still on to a winner until commissions come into play :).
    Is this your line of thinking. Sorry if I'm jumping the gun.