I'm a total novice at backtesting, say I want to backtest something simple like the breakout of daily chart-patterns, what would be the best thing to go about this. Which programs would I have to use or buy, where can I get my data, do I have to link them together or could this be (as much as possible) automated, all input is wellcome as I practically don't know anything about this.
I recommend that you look into E-Signal. It has an automated code generator which asks you questions about what you want to test. That works for simple ideas like BOs. And you can code yourself for more sophisticated ideas. Before you do so, however, let me recommend that you read a book or two first. My favorite is the late Bruce Babcock, Jr.'s "The Dow Jones-Irwin Guide to Trading Systems". Another is Robert Pardo's "Design, Testing, and Optimization of Trading Systems". You'll save yourself a lot of trading losses by reading those first and avoiding common trade system optimization errors.
What advice would members give on the sorts of criteria one should apply to backtesting prices, say YM or ES? Thanks.
You need to put so much time into learning and developing your own system, be it single strategy system or combination of 10s strategies system. Tradestation is best fitted toward beginners with no programing background, all documentations, educational material and data are extensive and cleanly presented. Tradestation forum is the only forum that Index all infos (Threads are relatively categorized based on trading fields, and updated) where no searching is required anymore https://www.tradestation.com/discussions/Topic.aspx?Topic_ID=24092
I don't often use the word 'backtesting' and when I have, I have probably had a different meaning in mind. Certainly I assemble and check historical prices and related data up to the present time and can find answers to questions that matter. Of course the efficacy of backtesting raw data has been mentioned or debated as a side issue in various threads. If you are seeking a day-on-day predictive formula or system, by computer backtesting of raw data for a given market, I suggest on that approach, you will never produce a tradeable or sufficient or worthwhile or sustained basis for prediction. However if there are any of you who wish to follow your excitement at the prospect of such a pursuit do not let me deter you.
I am not sure what do you mean by '' Day on Day predictive formula " but it is possible to backtest indicators values and get profitable entry and exit criteria.
A note from another novice for a book with good resell value : Professional Stock Trading: System Design and Automation by Mark R. Conway, Aaron N. Behle .