For me , stop levels are closely related to :
-risk/reward ratio
-which is related to the ATR
-the position size
-your account size and the overall risks you want to take.
Without going into too much detail (because I think I could write a book about it), here's in short how I use stops.
1. Ill never put more than 2% of my account at risk. That is, on all open positions. So if you have a 100k account I'll never would have at any moment in time more than 2k at risk. 2k is the risk and not the money you put on stake. (Talking now purely about day trading without overnight positions).
2. I normally limit the number of open trades to 2, in very rare circumstances to 3. So, if I'm not in the market, and I see an opportunity, I'll put between 1 and 2% of trading capital at risk. This, to stay compliant with the first rule. If I open a position with 1% risk, I still can open a 2nd one with 1% risk.
3. Following the first 2 rules, would mean, a stop loss of 1k (1%) as soon as I enter the trade.
4. Now that I have determined the maximum amount of money I want to put at stake, I have to determine the position size. This is where the ATR, the volatility and support and resistance lines come into play. As an example I could compare two stocks. Both trading at 20$. The first one is a high volatility stock with a large spread, let's say an ATR of 1. Something like TASR , RIMM, BIIB, OSIP etc. A so called "thin stock". The second one is a thick stock with a small spread and a small ATR, for example 0,1. Something like INTC, MSFT,CSCO etc.
On the thin stock, I would have to give in 0,3-0,5 in case I have to run for the exit . Also , as mentioned before by someone else, when that time comes , even when trading intraday, you can get a very bad fill when your stop is hit. So, for that stock I would use something like 0,5/share risk. Given the total that I want to put at stake, that makes 1000/0,5 = 2000 shares.
On the thick stock, in the same situarion, where I have to run for the exit when the shit is going to hit the fan, I would calculate my risk at around 0,1 (Even if you have to dump 10000 shares of MSFT, nothing will happen to the stock. Noone will even notice. Now try that with VXGN or ABGX and you're into BIG trouble)
So in the second case I could lift 1000/0,1 = 10.000 shares.
So we now could have 2 situations:
2000 at 20, stop at 19,50 or
10000 at 20 stop at 19,90.
Same hypothetical risk of 1k.
But......
5. Before I enter the trade, I see that the thin stock has a very hard support level at 19.70. For example, when it has hit that level already 4 or 5 times that day and rebounced back up. In that case, I would/could raise my stop loss to 19.60 (depending also on highs/lows and other support/resistance levels that may have occured over at least the last 2-3 trading days). I also give a quick glimpse at the daily chart to see if there are no other , more long term support levels in that area.
I now have 2 options and a choice to make:
5.1 I could say: well, if this really goes wrong and it gets to that level (19,60), and it breaks, then really everyone will be running for the exit, so I'ld better stick with my risk as calculated above. So, no change: 2000 shares. or...
5.2 I could consider that there is a very low probability that it will break that level of 19,60 and therefore lower my risk to 0,4. In which case I could lift 1000/0,4 =2500 shares.
The opposite can also be true: suppose that there is a strong support level at 19,30. This would mean that my stop at 19,50 is of much less use then the one at 19,60 in the previous situation.
Now we have 3 2 choices.
5.3 I stick with my risk as determined before, and take 2000
5.4 I'ld set my stop at 19,20. (so 1000/0,8= 1250 shares).
5.5 I'ld set my stop at 19,20, lift 2000 and thereby increasing my risk from 1k to 1600$. Possible, as long as I stay within my 2% total risk (see rule 1).
But wait. if the ATR is 1 and I take a risk of 0,8 , it means that my "predictable" reward would be 0,2. Also if , there might be a strong resistance level at 20,20 . In such a case the risk reward ratio wouldn't be acceptable . So in these case, I wouldn't take the trade at all, and just continue "sitting on my hands" (which is what I'm doing 90% of my trading day anyway).
With this explanation I think you can work out for yourself how to handle a "thick" stock (although I don't see how you can make a real living out of daytrading thick stocks, but that's another story).
Then comes the part of "how to get in" (bold, pyramid, etc), how to trail your stop, and how to get out. But that's maybe for my next book.
Have a nice trade..
TFD