Thanks for your comment learner2007, I use thinkorswim Ondemand to manually back test. How much sample data do you use when manually backtesting?
As I said, I usually pick 100+ cases, depending on the difficulty of the setup, to determine the probability. But after doing this for so many years, I can often reach a decision with fewer examples. Bedtime.
Given sufficient time those can be done as well. First time to define 'objectively' the type of relationship between one price move and another price move and variations that meet your criteria for 'too much' or 'too little', the various cases - again objectively defined. Will take time to all objectivize - the programming comes easy after that.
I really appreciate everyone response. In conclusion what I have learned from other who may read this thread: 1. Anyway you prefer to trade can be programmed, automatically backtested and live automated. This will require programming skills or hiring a programmer. 2. As learner2007 recommended, you can manually back test your ideas with +100 sample points during bear and bull markets time frame. I prefer to back test for +100 trades then program the strategy to test over the past +10 years.
Sounds like what you're looking for is Quantopian. It let's you a Look into Quantopian. It sounds like what you're looking for. It only automate trades on IB and robinhood but there's a low learning curve for beginning programmers when it comes to building algorithim.
Thanks Steve, I believe it can be trival if i put a EMAs on my chart to identify trend and then trade how I been doing. I been discretionary trading breakout with no indicators, just using my eyes to identify trend. But if I can program my strategy, I can try countless EMAs combinations to see which one gives better results. Man, the power of learning to program.
Thank you Maka, Is it really that simple? https://www.quantopian.com/algorithms/584e16351fc8d500120007fb I can say is WOW. I have been discretionary forward testing one strategy for 1.5 years on and off and I click that link above and ran the sample back test and in 1 minutes the results of that sample strategy shows the performance of that strategy. lol, I am speechless right now. So much time spent, but atleast I learned during the process.
Hello, I have a breakout strategy that have been manual backtesting with no indicators. I just identify the trend direction with my eyes. Well, I can not program and automated backtest what my eyes see. I would like to use EMA(s), but not sure which ema(s) to use. Can different ema(s) combinations be programmed and automated backtested to measure the performance of the strategy? This could take forever with manual backtesting. Thanks,