The second enhancement was to change the length of the channels. It turns out that longer channels can work much better for timing entries.
In the late 1980s I started using what I called the 40/20. When there was a break of the 40-day high or low, we entered long and short positions. The 40/20 was the same system as the 4WR, except that the time was twice as long. You could say that entries were made using an eight-week rule.
However, the exits were on a break of the 20-day high or low depending on whether you were long or short. This meant, for example, that you would enter long when the market made a new 40-day high and you would exit that trade when it made a new 20-day low. Note that you did not go short when the market made a new 20-day low; you only exited your existing long position. You would be flat the market if you got stopped out. This is different from the classic channel breakout. With the classic channel breakout, you are always in the market while the new method has you on the sidelines.
This was a major breakthrough in channel trading. Often a market would be in a nice strong rally and then quickly dip to the bottom and get short on what was really just a retracement in a major bull market. The
rally on the British Pound chart in September is an excellent example. In hindsight, we wouldn’t want to get long on that break to new highs. We would have preferred to stand aside. I’ll show you how to avoid those dings later.
I used the 40/20 for many years. Studies and my own experience showed it to be a nicely profitable system. I later changed to a 55/20 after the release of the Turtle System to the public. The Turtles were a group of traders trained by legendary traders Richard Dennis and William Eckhardt. They were basically trend followers and their core techniques were channel breakouts. They had settled on 55/20 parameters. The Turtles were some of the most profitable traders, so I was intrigued to find out that they were users of channel breakouts. In addition to using 55/20, they were also using the classic 4WR!
I love to learn, particularly from great traders. Dennis and Eckhardt certainly fit that description, having made hundreds of millions of dollars by the mid-1980s. It was amazing that they achieved that much success using the old-fashioned 4WR.
But they had also found out that using a longer-term entry parameter combined with a shorter-term exit parameter was an improvement over the classic 4WR. They had also created other rules for trading but this was the core upon which everything else was measured.
I had not checked which parameters were best so I tested whether the 55-day parameter was better than my 40-day parameter. It was! Actually, just about everything between 50 to 60 days had about the same results. So I enhanced my trading by switching to the parameters that the Turtles were using. Basically, I was shifting from a two-month breakout to almost a three-month breakout. I was going to be trading less often but presumably entering only when there was a monster move in place.
To repeat: Entries are made using 55 days but exits are made with 20 days. Let’s look at an example. Figure 3.2 is the same chart as shown in Figure 3.1, but now I’ve added a 55-day channel. If the 20 and the 55 channel high and/or low are the same, then you will only see one line. An example of this is the month of July. The top channel for both the 20-day and 55-day high are exactly the same so you only see one line. But note that there are two lines below the prices. The higher line is the 20-day low line and the bottom line is the 55.
Let’s follow the same process we did earlier but this time add in the 55-day parameter.
First, note that we do not get short in early April but wait until early May. This shows a couple of things. First, 55-day breakouts will occur far less often than 20-day breakouts. In this case, the market had to drop for another month to create a strong enough bear market to trigger the 55-day low. Second, this actually hurt us in this trade because we sold at a price lower than we would have using the 20-day low. On the other hand,
I actually don’t mind the lower price for the entry because it means that I didn’t have my money tied up for about a month with a market going nowhere. I don’t want to invest in things that are going sideways. It’s a
waste of time and attention.
We get short then in early May and get stopped out when the market moves above the 20-day high near the end of June for a loss of about 450 pips. Ugh. Now we are on the sidelines, not long. The market then moves sideways and then breaks the 55-day high and gets us long in the middle of July. That trade quickly fails and we are stopped out right at the beginning of August for about a 650-pip loss. A few days later, we get short when we break the 55-day low on August 8. The market then collapses and we show a profit until the sharp rally stops us out near the end of September. We still make about a 750-pip profit. We get short again in early October and would be showing a nice profit as the chart ends.
Note that this shift to 55 days eliminated the short position in early June. We lost a little more on the losing trades than when we used the 4WR, but we were able to eliminate one of the losing trades. We are probably a little bit more profitable by adding in this new, longer entry channel. Over the long run, we will be much more profitable.
A lot of people ask me what will happen if these two parameters become very popular. Won’t that degrade the performance? There is no question that a method can become too popular and the performance begins to degrade. But this technique is robust. First, few traders are looking at 55-day highs. Most traders are using short-term support and resistance to time their trades. They are looking at 10-day moving averages and the most recent high or low. Few are looking back over the last 55 days. Still, it is possible that the 55-day high or low could have occurred in the last 10 days and that means that a lot of traders will be looking at that point to buy or sell, which will degrade performance slightly. I’m always concerned about dealers and other short-term traders moving a price to just above the 55-day high, triggering stops, and then having the market collapse. The breakout will have occurred due to short-term price manipulation rather that a significant change in the fundamentals. I’ll tell you how we can counteract that next.
The 55/20 technique is so robust that you can use just about any parameter between 50 and 60. If you think that everyone is entering on 55-day highs and lows then shift to 54 or 56. Heck, you can go all the way to 50- or 60-day highs or lows and it won’t make any difference. The basic technique is catching some deep human action; shifting the parameters slightly won’t change the performance.
In addition, 55/20 are not necessarily the optimal parameters for each pair at all times. The best overall parameter is 55/20 because it works the best for all pairs over the longest period of time. I’ve found that the use of 20 days for the exit works well for most pairs most of the time so I don’t fiddle with it. However, I have seen the optimal length of the entry channel range from about 30 to about 70. I’ve seen the optimal entry parameter change from year to year by half or double.
The problem with constantly changing the entry parameter is that you will always be using the parameter that worked over the recent past, not the parameter that will work in the near future. Sure, a parameter that
worked the best last month or year will likely be the parameter that will work the best in the coming month or year. But I don’t like to fiddle with the parameters because it makes the method less accountable and effectively turns the trading method into a discretionary method. I’ll talk more about this concept in Chapter 9 on the psychology of trading. Chasing varying parameters is generally a waste of time. Finding a strong and robust set of parameters that works is the key and I’ve presented those parameters.
Source: How to Make a Living Trading Foreign Exchange: A Guaranteed Income for Life (Wiley Trading)
