Hey guys, some of you may know me, I have been on this forum for 2 years or so off and on. A little history before I ask what I need help with. Been working on algorithms to trade the market since April of 2013. Been at it for months, and during the summer I had a small team of traders who traded with me (Liquid Genius I think was his forum name? just to name one…) I did really well my first couple of weeks in the summer… then brief periods on and off between my first couple of weeks and now.
I traded pretty stupid considering the low amount of start up capital. At first I traded micro trends. got in at 00:30 oanda time with no stops only taking 20 pips each trade. I took $140 up to $1170 at my peak. 14 trades in a row wins. then slowly over the summer I changed systems… leaving an old system once it lost a trade and I had attempted to adapt the system to correct the errors unsuccessfully. Essentially once I got down to breakeven, I pulled the money out. Put some money back in a little later on traded again, lost some, won some but lately I haven’t been winning like I used to or even should. I know several things need to change about my trading style. IE stops and take profits would be a great start and definitely reducing my risk. I have built over 15 different systems all based off mathematics using Microsoft excel. I have been working on system 14 as I call it for the past two or three months, but I would like to restart and go back to the drawing board. I intend on keeping certain aspects of previous systems to help build a new one which leads me to this post here.
Before I can do anything of value mathematically to achieve my goals, I must first conceptualize what is important in the past to use to predict what is to come in future I am trying to predict.
I will start this post off slow with a few questions and go deeper as I get more input from the community. For starters, my original goal right now is to predict market direction for the next day. I want to predict what open to close will be whether it be long or short. I’m not worried about wicks yet, I will get into that soon. My basic rules I intend on following, at least for now are no indicators, only mathematics I can calculate or have already calculated in excel as my decision making tool. Side question… Taking math out of the picture, what do you think happens day to day that gives insight on what is going to happen tomorrow?
First question, In order to predict one full day open to close, How much information should I use (Open, High, Low, Close, and Volume data given by Oanda) ? 1 day, 2 days, 3 days… 20 days? in the past I only used 1 day, the previous before the day I am predicting.
Second question, More information, less information, or a combination of the two? What I mean by this is 1m, 5m, 15m, 30m, 1hr, or 4h data, etc. Each set of data, If I am advised to use more than one would be calculated separately in its own sheet. I could then also create a comparison sheet to compare the time frames (have done this in the past)? What do you guys advise.
I have spent thousands of hours in the past testing each new idea I had. The problem with that is it takes an incredible amount of time just to test a single idea. In order for me to take my trading to a professional level regardless of capital, I think its most important to come up with lots of ideas, conceptualize them, then test the ones that appear to have the most potential before I go OCD and test everything that comes to mind big or small. Like most everyone else, I don’t have unlimited time to test just any idea, Its about picking the right ones is what I have learned. So before I just kill myself building my best, smartest system yet, I would like to really plan it out and make sure it makes sense before I kill a bunch of valuable time.
That’s enough for now, I have some more advanced questions after I get some feedback. Thanks for your time! I know yours is as valuable as my own. I look forward to reading the responses I receive.