|
|
|
|
What's Obvious Is Not Often Right
Most traders are inclined to place a stop right at the bottom of a double bottom or top of the double top. The conventional wisdom says that once the pattern is broken, the trader should get out. But conventional wisdom is often wrong.
Leaving the trade early may seem prudent and logical, but markets are rarely that straightforward. Many retail traders play double tops/bottoms, and, knowing this, dealers and institutional traders love to exploit the retail traders' behavior of exiting early, forcing the weak hands out of the trade before price changes direction. The net effect is a series of frustrating stops out of positions that often would have turned out to be successful trades.
|
|
What Are Stops For?
Most traders make the mistake of using stops for risk control. But risk control in trading should be achieved through proper position size, not stops. The general rule of thumb is never to risk more than 2% of capital per trade. For smaller traders, that can sometimes mean ridiculously small trades.
Fortunately in FX where many dealers allow flexible lot sizes, down to one unit per lot - the 2% rule of thumb is easily possible. Nevertheless, many traders insist on using tight stops on highly leveraged positions. In fact, it is quite common for a trader to generate 10 consecutive losing trades under such tight stop methods. So, we could say that in FX, instead of controlling risk, ineffective stops might even increase it. Their function, then, is to determine the highest probability for a point of failure. An effective stop poses little doubt to the trader over whether he or she is wrong.
Implementing the True Function of Stops
A technique using Bollinger Bands can help traders set those proper stops. Because Bollinger Bands incorporate volatility by using standard deviations in their calculations, they can accurately project price levels at which traders should abandon their trades.
The method for using Bollinger-Bands stops for double tops and double bottoms is quite simple:
- Isolate the point of the first top or bottom, and overlay Bollinger Bands with four standard-deviation parameters.
- Draw a line from the first top or bottom to the Bollinger Band. The point of intersection becomes your stop.
At first glance four standard deviations may seem like an extreme choice. After all, two standard deviations cover 95% of possible scenarios in a normal distribution of a dataset. However, all those who have traded financial markets know that price action is anything but normal - if it were, the type of crashes that happen in financial markets every five or 10 years would occur only once every 6,000 years. Classic statistical assumptions are not very useful for traders. Therefore setting a wider standard-deviation parameter is a must.
The four standard deviations cover more than 99% of all probabilities and therefore seem to offer a reasonable cut-off point. More importantly they work well in actual testing, providing stops that are not too tight, yet not so wide as to become prohibitively costly. Note how well they work on the following GBP/USD example.
|
|
More importantly, take a look at the next example. A true sign of a proper stop is a capacity to protect the trader from runaway losses. In the following chart, the trade is clearly wrong but is stopped out well before the one-way move causes major damage to the trader's account.
|
|
