Quote from AAAintheBeltway:
Deciding on the optimal stop size is dependent on a number of factors. What is right for one trader will not work for another. Personally, I like to use a mental stop as disaster insurance rather than as a routine exit. Ideally, I would be out of a dubious trade before the stop was triggered.
This is where backtesting software can be very useful. You can vary the stop size and see how it affects profitability. For most any type of intraday entry signal, there is a point that winning trades almost never go. In other words, a winning trade tends to go against you by only so much. This is called maximum adverse excursion or MAE. Some s/w, like Wealthlab, will display MAE for your system and you can see the tradeoff between win percentage, biggest loss and profitability.
Obviously, different entry signals will expose one to varying levels of MAE. Time of day, market volatility, data releases, etc all come into play as well. If you have to use a very tight stop, you will probably need to restrict your trading to less volatile times. A one point stop would be pointless immediately after a FED release for example.
The idea that small stop levels reduce risk is illusory. They might reduce the exposure on any particular trade, but over time they tend to reduce profitability, both by increased numbers of losing trades and by knocking you out of potential big winners.
Thank you for taking the time to write that post.
Nexen
