Dude: I want to try this new strategy on this instrument what do ya-all think? Trader: Whats your statistical back tested edge? Dude: .............
How do you reduce something so complex down to a single question like that and expect an answer? Most times it can take an entire journal and we still wont know what a traders edge is.
Can it be reduced down to a mere statistic?
Can it be based any one of the following things?
1. superior money management?
2. Iron clad discipline?
3. Phsychology?
4. Journaling and record keeping?
5. Accomplished technical ability?
6. Profoundly deep understanding of fundamentals?
7. Gigabyes of back tested data?
8. Reams of forward tested data?
9. Thousands of hours of screen time?
I can go on...
You might think that your edge is your ability to execute this one thing repeatedly which your spreadsheet tells you statistically works out in your favour 55% of the time (or whatever). If it is that simple then why cant someone else take your strategy and trade it because we all know that doesnt work?
Your edge is the sum total of everything you have been through on your journey to become a trader. Everything I have listed is important and is an edge in itself. Backtested statistical data is just one piece of it.
What do you think?
The following 5 users say Thank You to Grantx for this post:
If you don't have positive expectancy, many of the other things on your list (iron clad discipline, superior money management, thousands of hours of screen time, journaling, etc) become irrelevant.
But, if you do have a positive expectancy edge, many of those items you listed DO become important. Example: with poor money management, you can easily get wiped out even with a positive expectancy system.
It all starts with positive expectancy - if you lack that, you will eventually be screwed.
Think of it like building a castle. You need a solid foundation (Positive expectancy) before building the castle (other items in your list). Because as Jimi Hendrix said "castles made of sand, fall in the sea, eventually."
The following 10 users say Thank You to kevinkdog for this post:
This is an interesting topic, as I have wrestled with these issues for a long time. I would pose two related questions that try to get at the heart of the matter:
Is a trading strategy only valid if it can be coded and run by a computer? (Assume the coder is highly competent)
Can a purely discretionary strategy truly have an "edge?" (If it does, see question #1).
One "out" that the discretionary traders always take is "context." i.e., "I would take THIS entry, but not THAT (identical-looking) entry b/c the "bad" entry is too close to resistance." Mack from PATS is big on context, but appears to be successful. Same with Bob Volman, Al Brooks, Sam Seiden, etc.
Assuming pure discretionary price action type methods have an "edge," is there 1,000 (or 10,000) lines of computer code that could encapsulate all of the "context" and exceptions to the rule that would verify the edge?
The following 3 users say Thank You to drm7 for this post:
Positive expectancy is the sum of the parts and fluctuates along in direct response to the input. Its a representation of how disciplined you are with all the points listed (and others).
Firstly, to ask the question 'what do you think of this or that strategy' is a silly thing to ask in the first place. But equally baffling is to ask whats your edge or do you have a positive expectancy? Edge and expectancy is only relevant right now, because these metrics can change. You cant rely on your expectancy for anything. Or can you? What do you do with it?
The following 2 users say Thank You to Grantx for this post:
There can be an 'edge', which by eye-balling the screen is very clear, why long, why short, why don't touch
once you want to quantify it (at least in classic rule-based system) and program it, it becomes very very
difficult as there is always a little 'but', which is evident to the human eye/brain, but when not programmed
it doesn't count...
Today half of my strategies are classic rule based, if / then and work OK
other half are machine learning, artificial intelligence based
I'm currently working on a small new project, and teach my kid (university student) the magic of machine
learning. as a example i hope to illustrate the result of a rules based system compared to a machine learning
system. But on the same concept... The goal of the exercise is : manual, trial/error/optimisation of features
and parameters, compared to automatic feature extraction... will be a good interesting exercise
More to follow this month...
The following 9 users say Thank You to rleplae for this post:
Like the sound of the project, have a similar plan in incubation, good hunting.
However, regarding 'the edge' I'm afraid 90% is illusion caused by the magic of whole charts and hindsight, there is only 'the right-hand edge' when we're live...
Cheers
Travel Well
The following 3 users say Thank You to ratfink for this post:
I'm sorry, but you're wrong. You can't conjure profit through positive thoughts, writing in a journal, screen time or discipline in financial markets. If what you are doing is mathematically unsound you will lose. Those things you listed are prerequisite to success and may help you adjust yourself to attaining an edge. By themselves they do not equal one. You must have a positive mathematical expectancy. Everything else is secondary.
The following 5 users say Thank You to TheShrike for this post:
