I study charts in my spare time for about 2 years, I feel like I've developed a system, but before I started to trade full-time obviously I can't be sure and rely on it fully. Also some people say it tooks up to 10 years to become consistent. So I'd like to ask you, how long it was in your case (did you do it full-time or part-time), do you have some tricks that helped you to accelerate that process? Thanks
in my case, before I started trading in the end of 2006, I had studied market for 3 years. in 2004, I bought 20 dell precision desktops, and used machine learning to identify profitable patterns. the calculation was so cpu intensive, at one point, it caused outrage in my apartment. the software dug out hundreds of patterns with statistical meaning, most of which were due to over fitting. after careful research, I located one pattern that has true edge, which i have used since then for the last 12 years. the trading system has less 2% dd over the last decade, sharpe ratio on average above 7. I made 230% in 2007, 250% in 2008, since 2007 was bull and 2008 was bear year, I figured my system had survived both bull and bear cycle, and thus I had confidence that I could quit the tenured government position.
I had a Bond trader buddy at the board of trade tell me a story about his friend who was also a Bond trader. So my friends trader buddy would take his commodity perspective charts to his nieces house and pay her a visit. He would than place her on his knee and open up his charts and ask her what direct she thought it was going go. This story should tell you that the more you look at charts the less you know and do. Its best to stop and work with clean charts and experiment with simple moving averages. Don't overthink and keep it simple and just use some simple logic. There is no universal truth when it comes to an approach. You will always lose money, you have to manage how much you are willing to lose on a trade. If you get stopped out? its because you entered to close to noise or you were greedy and the market reversed on you. They will not deliberately stop you out, markets prices move and if your stop is with there range you will get stopped out.
Developing an edge for a full automated trading Algo took me several years and in spite of an acceptable statistical consistency, I continue to get it better, cause there are many market quick and aggressive situations that makes hard to take good decisions.
Consistency in basic terms would be profitable every week, or every month or so; One of the experienced guys who designed many custom indicators and systems said he could guarantee a paycheck every week, and some one at ET here (she stopped posting for many years now) said she could almost be profitable every day if she have enough opportunities to enter trades (like 10 to 20 trades) on the day. Of course, that is a simple way to define it, and certainly there are more criteria to that.
Then I am not "consistent" nor do I ever expect to be.... Do you seriously think that is a realistic and necessary goal? GAT
Same here. During my over 10 year trading career I've found only 1 pattern which works year after year (gives a positive expectancy). I have used thousands and thousands endless hours to find an another but no luck. Fortunately 1 is enough. Mental control and money management of course are mandatory but without an edge (in the long run) mr. Bell Curve will make sure you're not going to make it. I did a traditional journey with all the bells and whistles until after about 4-5 years understood to clean my charts but price bars and started to see the market which helped me to find this pattern together with a lot of data mining and backtesting.
I think you are consistent in winning based on what I read, what I described is a very simple way/one aspect to define consistency and it is realistic as people achieved that, but not necessary the only way to define it. If you heard about Goldman Sachs' performance, they in general only have a couple losing days per year (at least recent years, unless I am misinformed). Of course they have vast resource. Again, there are many ways to define consistency, and you probably demonstrated from a different aspect /criteria to show consistency (your designed your system to function in a consistent way with regard to many parameters).