The Value of Using Targets
|September 10, 2012||Posted by Oddmund Grotte under trading topics|
Previously I wrote an article about the detrimental effect of using stop-loss. What happens if we reverse it? Does using target create any value? When trading it’s mentally very hard to see a nice gain turn around into a loss. It’s easy to kick yourself for not having used a target. But in the long term, does it have any real value?
First, let’s define what we mean with a target. Target is a predefined level in which you sell (or cover) at a profit. Let’s say you bought SPY at 100 and have established a target of 2% in your strategy. Then you sell if SPY reaches 102. Quite simple. It’s the opposite of a stop-loss.
Now, if you want to use a target you have to consider the following: First, you can’t trust the data you’re testing on. Open and close have a bigger chance of being correct than high and low. High and low quotes has a lot more bad data into it. And high and low needs to be used when using targets (at least intraday, but we can also use targets on close on swing trading). Second, we have the aspect of curve fitting. Some strategies works a lot better with targets, others don’t. Hence, be careful so not to overfit the data. Three, and this is the main reason, the strategies perform better with other exits. A simple time exit is a lot better. You’re simply cutting winners short by using target. In some strategies just the best 5% of the trades are the main source of good returns. Looking at the distribution of the returns you can have an idea if target is good or not. But be very careful!
I always test a strategy with and without target, but it’s usually 40/60 if it improves performance or not (more likely the strategy is better without a target).
Another point to bear in mind is that targets are not very efficient if trading for example oversold strategies. In these occasions there are often some terrific bounces which are cut too short if using a target.
The best advice is to test strategies both with and without target. Then you can get look at your own numbers and draw your own conclusions.