When we are talking about riding winners we assume there is some type of medium to longer-term trend in place and the problem is one of determining the right time to exit. There is another approach for dealing with this issue that attempts to ride trends by breaking them into a series of short-term trades. This approach assumes one has available a swing/position trading system that generates enough trades along the trend with fixed - although not necessarily constant for each trade - exit levels. This is an article that describes this approach: Trend Trading with Short-Term Patterns
I've not read every page of this thread, so apologies if this has already been mentioned, but one possible technique is the so-called "chandelier" stop. It's a fairly simple idea. If you're long, your trailing s/l is n x ATR below the highest high for the last d days, and n x ATR above the lowest low for the last d days. Typically n is set to 3, and d could be 20 days on a daily chart. This method keeps you in trends and allows for increased volatility. The drawback is of course the bend at the end of the trend, where you can potentially hand back a decent amount of open profit. It's worth programming it and seeing what results you get from backtests. You can adjust n and d as you see fit. If you find parameters that work, this method will let you ride the occasional trend a lot further than you otherwise might.
On the issue of moving stops to breakeven, I've tested a large number of different strategies and have NEVER found any benefit to moving stop to breakeven at any stage, in fact, the opposite (you're far more likely to get kicked off a winning trade than save yourself money the other way). I think it's one of those things that sound sensible, but in trading not everything is intuitive, or designed to make you feel comfortable.
Very often it boils down to being able to identify what is noise and what is a reversal. Obviously, you want your stop loss outside the noise and ideally you want to get flat/reverse when you spot a reversal. The more time you spend trading/watching your market, the better you can tell what is noise and what is not.
To be fair I should qualify my last response to this. If I'm jumping into a momentum trade, such as shorting a strong breakdown on volume, the trade is immediately profitable and I always place my stop at b/e right away. This works every time, but that's because of the nature of the price action in this type of trade, and I want to be protected against a sudden S/R bounce out of the move.
Me too find useful break-even stop or protecting X% of open profits, and me too tend to trade momentum, a technique not often backtested I think (at least for very short time frames). Yes GhostofCutten the stop should be placed where you are probably wrong, but I have two questions: - If you place your stop at key levels,couldn't it be guessed by stop hunters? - Can we really identify when or where we are probably wrong? That is, do key levels we see give us a real edge? For these reasons I like stops placed a bit randomly,as it happens when saving X% of open profits.
interesting - this is exactly my philosophy/practice for years. the reason is that i never figured out anything better than that. the problem with this mental approach is that you are going to (partially) miss on big moves up while being (partially) exposed to big moves down - therefore the strategy is similar to selling puts in a trending market. the premium for writing these puts exactly offset the costs of straddles which you implicitly own in a rangebound market via the long gamma strategy. these markets are also the best for the performance of the advocated scale out/in approach.
To clarify my last post: I look at signals (and to some extent, risk) to enter (better to say, add to) a position; to exit (=reduce) the position, instead, I mostly look at risk, nearly ignoring where the market is going. This because signals I look for appear when they want, possibly too late for the risk I can bear. An example: suppose I am able to determine where the market is going only through big volatility and volumes. So I enter a short long-term position in 2003 in equity markets. Then volatility and volumes drop, and market starts a steady rise. Before I can see any signal, I'm broken.