Definitions...
dbphoenix said:
So you don't know what you're looking for, you haven't defined your setup, you don't know what to expect from whatever kind of backtesting you're doing, and you're not clear about what a hypothesis is?
Ok, here goes the definition part. I'll be using some parts of #5 in the Trading Journals thread, copyrights reserved
To put it straight: "One is more likely to get what he wants if he knows exactly what it is that he's looking for."
Perhaps the problem is the difference in opinion about what a setup consists of.
"Only after the setup is defined and tested (...) can one even begin to think about where to enter, what the target ought to be, where the stop should be placed, and so on." A setup to my knowledge was everything that made up for a trade, including the entry, stop, target. So I got rid of that first. I posted something that looks like a setup in #307. There's no mention of target, stop, etc... only a description of what one would look for in a chart. So in this particular case my hypothesis could have been: "If price breaks through S/R on a WRB (by x number of points or ticks) and after that retraces to the S/R zone, price is more likely to continue it's up move." Is this the kind of hypothesis you want to be seeing? Because if it's not, no point in continuing, than I'm already stuck at the first step. I did think that something in the likes of "if a breakthrough S or R happens with price moving away further then 3 ticks, price will continue to move in that direction" sounds pretty much like a hypothesis to me... So the independent variables are support and resistance, the dependent variable is price. There's no volume on the chart, but if there was I'd probably argue being volume the independent and price the dependent, but let's not get into that.
Ok, next step was to limit myself to one strategy which is what I did, abandoning points 2 and 3 written in #78. So each day I put up the chart with S/R of day/week/month before and looked for entries on that day. Most of the time I didn't see any, some of the times I did. At that point we somehow seemed to be getting out of sync. The next step would have been to determine the ratio of P/L and W/L, but I'd have to have more data for that purpose. Hence me saying that I would be "backtesting", in that context refering to "looking at old charts to see where the setup appears".
Ok, I won't make this any longer then necessary and answer your questions:
-> I'm looking to see what price does after my setup appears, where does it go, how far does it move, does it retrace, reverse,...
-> I expect from backtesting to 1) in my context give me more data to check out where the setup appears and perhaps refine that setup in order to determine a better stop or target; 2) in your definition a way to define the setup
-> I've stated two hypotheses, so if this isn't what you had in mind, I'm not sure what is...
Feel free to reply whenever you have the time...