10 months fulltime and still no edge
- Thread starter estim
- Start date
All things considered, to say that you can become a great mechanical trader without first being a great discretionary trader is a fallacy. In my view, you can't be a great mechanical trader unless you've already established yourself as a great discretionary trader. Why? To become a great mechanical trader means that you have a perfect system in place to work with. But exactly who created this system? Unless you spent a fortune acquiring the system from a third-party, you had to know what you were doing to devise the system from the ground up.
My suggestion for you is to go back to the basics. You would be amazed to find what lies just below the surface. Check out my "Unholy Grail to Success" for a pointer.
My suggestion for you is to go back to the basics. You would be amazed to find what lies just below the surface. Check out my "Unholy Grail to Success" for a pointer.
Truth! I have met some very successful traders. They made millions and millions of dollars in trading. But in the end, they blew almost all of their money away. Now they are content working for $20/hour job. I guess they were too arrogant when they became successful. They thought they were the master of the universe and would never fail.
estim,
You are very close, I can feel it. You are using the right tools with matlab and Ninja. And I can see you understand using the right statistics and looking in the right area. You should definitely explore what you listed as not working.
The only thing you are missing is longer time frame. Cointegration and pairs trading all need time on the hourly candle to function properly, and trades can last several days. Make sure you look at the results and really study it, do not just look at the profit and loss column and discard theories that can become very profitable to the people who can figure it out.
good luck.
You are very close, I can feel it. You are using the right tools with matlab and Ninja. And I can see you understand using the right statistics and looking in the right area. You should definitely explore what you listed as not working.
The only thing you are missing is longer time frame. Cointegration and pairs trading all need time on the hourly candle to function properly, and trades can last several days. Make sure you look at the results and really study it, do not just look at the profit and loss column and discard theories that can become very profitable to the people who can figure it out.
good luck.
I think that the essential element in a "trading edge" is keeping it as simple as possible.
The more technical indicators you use, the more signals you get that will be in conflict with one another.
10 months trading full time is still in the beginner stage in becoming a great profitable trader.
Best of luck!
The more technical indicators you use, the more signals you get that will be in conflict with one another.
10 months trading full time is still in the beginner stage in becoming a great profitable trader.
Best of luck!
In my experience, the notable enemies when struggling for finding a trading strategy are two :
- our high and understandable desire to be successful, and therefore a sort of precocious "indulgence" in believing in some results, without carrying on serious scientific investigation about the "robustness" of the trading method
- overfitting. That is the tendency to create strategies which give promising result on recorded data, but will systematically result in losses when actually traded
The first enemy is therefore psychological" and with discipline and help of the ferocious criticism from other people (like ET users) can be kept under control.
The second one is immensely more difficult to beat and requires very hard work (massive simulations, hard programming, intelligence, etc) and skills to devise and implement the methods to discriminate between really "robust" methods and "overfitted delusions".
There seems to be no easy shortcut.
T
----------------------------------------
G-Bot (Day Trading) Project
http://www.datatime.eu/public/gbot/
Hi JB3,
I have been re-thinking my fully automated trading project the last month and your comment is pushing me in the direction I will change. So for a start I will do larger time frames 60 minutes for the signals and daily frames for the overall market condition. Secondly I will go back and look at pairstrading again under these new conditions. And third it does not have to be as profitable as I first expected. Of course my focus has been robustness over profit but I might have discarded some small profitable systems because I wanted the BIG system. Finally I have read some books on portfolio management which I will try to include when doing pairs trading on baskets of stocks.
I have been re-thinking my fully automated trading project the last month and your comment is pushing me in the direction I will change. So for a start I will do larger time frames 60 minutes for the signals and daily frames for the overall market condition. Secondly I will go back and look at pairstrading again under these new conditions. And third it does not have to be as profitable as I first expected. Of course my focus has been robustness over profit but I might have discarded some small profitable systems because I wanted the BIG system. Finally I have read some books on portfolio management which I will try to include when doing pairs trading on baskets of stocks.
Who would have guessed that it would be hard to find an edge in financial markets after 10 months?
Coming up, classes on:
1. How to win the Nobel prize in 10 months
2. How to become a blue chip CEO in 10 months
3. How to go from novice to NBA pro in 10 months
4. How to learn brain surgery in 10 months
Coming up, classes on:
1. How to win the Nobel prize in 10 months
2. How to become a blue chip CEO in 10 months
3. How to go from novice to NBA pro in 10 months
4. How to learn brain surgery in 10 months