Assuming you're talking about intraday... I think the best is either an ATR based stop, or stop just outside the high/low you are trading off. In my experience if you're having lots of trades that run your stop a bit then turn around and head your way, and assuming the stops are reasonable to start (say not less than 1x ATR on avg), the proper response is to try and get better entries, or filter entries, versus trying to evade the problem by widening stops. It's relatively rare that I get barely clipped out on trades that turn around to work, and when I do, the stoprun usually sets up a second trade/entry. If you place stops based on high/low you do need to have some controls in place to keep your R:R and targets reasonable, like a max allowable stop. As to question 3, it really depends on what scale of move you're looking for. I think the idea of not doing any in-trade money management at all is sub-optimal, and at a certain point it becomes ridiculous to hold your original stop from a risk-reward perspective - e.g. you have a trade that's 19 points to a 20 point target, in which case holding the original 10 point stop would be risking 29 to make 1, which is nuts. But a simple mechanism (like an auto trail or auto breakeven after a given MFE) can help you keep R:R in line as the trade progresses without needing much or any attention from you.