There are several issues floating around here. Let's look at them briefly:
1. Initial stop. From backtesting, you determine that a trade that goes n amount against you seldom comes back with in the relevant time frame. That is where you place your initial stop. Wealthlab has a nice feature called MAE , max adverse excursion, that gives you a nice visual representation of that concept.
If that stop is too wide, then you have to come up with another entry method or use a smaller timeframe and retest.
I Like to use this initial stop as a disaster stop that is seldom hit.
2. Risking 1-2% per trade. Many experts advise that you risk no more than 1 or 2% of your account per trade. $10,000 account, you can risk $100-200 per trade. This is used for position sizing however, not for determining the initial stop. That depends on the market, time frame and entry method.
3. What triggers an exit?
a. For a trade that has not cleared the noise zone around the entry price, I am looking for anything that invalidates the entry criteria. For example, if my entry is based on crossing a pivot point, if it reverses back below the pivot, then the trade is on thin ice and probably should be exited pretty quickly. waiting for the stop to be hit can be costly, but for systems traders, that may be a requirement. Otherwise the system testing is invalidated.
b. For a trade that clears the entry zone, I pull the stop to break even. This puts into place the important adage "never let a profit turn into a loss." The exercise then becomes one of managing the profit.
c. Taking profits. Kachingo, baby, this is what we live for. There are a lot of adages here too, such as no one ever went broke taking a profit, or if it's not booked, it's not jack. But don't lose sight of "Let your winners run." Hopefully youhave some sense from backtesting how far to let your winners run. I'm not a big fan of profit targets, except in crappy range-bound markets. The big moeny comes from catching a big move and sticking with it. Some ideas for getting out are a nasty reversal bar or pattern, clearing obvious resistance (if long) then pulling back, an oscillator divergence, a trendline break or the close coming up. A good guide is never let more than 50% of a big profit get away from you.
Hope this helps.