hi bxurbx,
there are a lot of questions here. let me see what i can do:
1. measure of trend... yeah that's not easy. i dont like up weeks... forget ADX or DMI or anything like that. you might be able to create a proxy trend indicator from a fast MA - slow MA type indicator, a slope of MA indicator, or a linear regression line. the problem with all of these is that they will often boil down to a binary (yes or no) answer with regard to trend... and i don't think it's always a yes or no. when you play with it and tweak it it doesnt capture the uncertainty at the turns that a trader would see.
it's hard to beat higher highs and higher lows as a measure of uptrend... but it will take some work to code. my results using systemized versions of this have been less than satisfactory.
2. yes. maybe risking to a previous pivot on the individual stock... maybe risking a % of average daily range... make sure you're comfortable with whatever you choose and that you really understand the implications.
3. targets. i think for a system like this maybe no target is appropriate... not sure... you could scale out at multiples of your risk. i'm a big fan of that.. you never have your full size for the big moves but you will have many trades that would be losers that you end up making money on.
i dont think this is "basic stuff" and anyway i love "basic stuff". the most successful traders i have trained are the ones who have most immersed themselves in the basics... hand charting, hand backtesting (a big fan of that btw, but you have to do it right), keeping a journal, reviewing their trades, etc etc.
success isn't about the sexiest trading system or having the highest yielding limit order script on some website somewhere. it's about relatively simple concepts coupled with behaviors that go against our nature for the most part... and about practicing until those behaviors are second nature. basics and fundamentals over and over. very few people will submit themselves to the discipline, shed their preconceptions and do what it takes to be successful. put me in front of a screen with nothing but the last price of a stock flashing every trade, give me a pencil and paper, and i can outtrade most of the systems, algorithms and other traders out there because of years of focus on the basics.
never apologize for basics!!!
Quote from xburbx:
talon,
Hopefully these questions arent too off base in the discussion. I know your time is limited so I will try to keep it short.
I have a system that I believe makes sense and is logical which is your number 1 in the 3 steps. When looking to test it and backtest it without using too many parameters, what are the keys to focus on for a simple TA system like the following. I am looking for what parts of the system I should be testing to make it effective.
Some description of the current study.
1. A trend following system that play pullbacks and correlates to the trend of a major index. (What measurements should I use to gauge the trend being in effect? Up weeks? Pivot breaks?)
2.Draw down - A blowout stop and a soft stop. (A measurement of stop sizes based on range or s/r or something else?)
3. Targets - How to determine the most effective targets? Predetermine amount, average range, standard deviations?
I know this is basic stuff compared to what was discussed earlier, but as of the last 6 months this strategy I have been using has been working and is slightly discretionary.
This is based on a system that scans for stocks with no concern for sector.
