How to Pick the Best Trends
Everyone’s heard that trading with the trend is the easiest way to put the odds in your favor. Yet there’s uncertainty over how to define trends and how to pick the best trends to trade. Here’s a few tips concerning the best definitions and filters to use when trading with the trend.
Defining a Trend
It is surprising how much dispute there is over the question of how to define a trend. You can find a wide range of different opinions. One popular choice as a filter to determine trend is the 200-period exponential moving average. Another is the 50-period simple moving average having crossed above or below the 200-period simple moving average. If you search, you can find a lot of other technical, indicator-based definitions. In fact, the dictionary definition of a trend can be summarized as a consecutive series of higher lows and highs (an uptrend), or lower lows and lows (a downtrend). Unfortunately, that is rather difficult to define mathematically, although most traders who have put in a reasonable amount of time reading price charts can tell you whether an attractive trend exists just by using their own eyes. The question remains, is there a way to define the existence of any trend, which we can use to at least identify that some type of trend exists, before we try to apply filters to pick out the most useful, profitable trends? I believe there is a simple answer: a basic upwards trend exists if the price is above where it was 3 months ago, and downwards if below. In markets, the rise of fall of price over a relatively extended period such as this has been shown to provide an edge. As the Forex market is more mean-reverting than most liquid speculative markets, the optimal period to use is a little shorter than it is in other markets. Results can be improved by stipulating that the trend must also be above or below its level measured 6 months ago. A true upwards trend, for example, has the price above where it was both 3 and 6 months ago.
Using these measurement as a baseline, let’s look at some statistics for the two most liquid currency pairs, EUR/USD and USD/JPY. Assume that over the past 16 years, you had bought each pair at the start of any week it was above its prices from both 3 months and 6 months ago, and exited the trade at the weekly close, or sold vice versa if below these historical prices. Ignoring spreads, commissions, and any possible positive or negative overnight swap, the results would have been as follows:
This trading methodology is not presented as a complete strategy, just as an indication of how profitable a trend can be. There is clearly an edge here: most weeks saw moves from open to close in the direction of the prevailing trend. The question is, can this simple strategy be improved by filtering the trends somehow, and only taking the trades when the trend is somehow seen as stronger or more reliable?
Using the ADX Indicator as a Trend Filter
The ADX indicator purports to show the “strength” of a trend. It does this by measuring the total amount of directional movement in a single direction over a given recent period. The value of the ADX indicator can range from 0 to 100. Typically, a trend is said to be strong when the value of ADX is 25 or higher. It seems like an appropriate filter to apply to our trend definition. I applied it to the most recent 3-month equivalent by using a period of 13 weeks (the “short term” component of our defined trends), and examined the results that would have been achieved by using ADX levels of 25 and 30 as filters, which were as follows:
Both ADX levels improve the win percentage and average profit per trade, the latter increases considerably. Note that although ADX 30 produced a lower average profit, its win percentage was slightly higher than what would have been achieved using ADX 25.
Using “Blue Sky” as a Trend Filter
“Blue sky” is an area of price that has not been visited for a long time. An ancient belief of traders says that the price moves more quickly and directionally through price areas which have recently been empty. This is the theory behind breakout trading. After all, if the price makes a new 6-month high by breaking out about that level, by definition, it has not traded there for at least 6 months. Perhaps we can apply the following filter to our advantage: examine the weeks where the price made a new 3-month high or low price during the previous week. In other words, there was a breakout last week of the 3-month price channel in the direction of the prevailing trend. Here are what the results would have looked like:
Interestingly, this “blue sky” filter would have given even better results than using strong ADX values as a filter. When both filters are combined, the results are even better.
The most reliable non-discretionary definition of whether a trend exists is the simple measurement of whether the price is both higher and lower than it was using a historical lookback. Three-month and six-month time periods have worked very well in Forex markets in recent years. Stronger trends produce more reliable short-term trading results than weaker trends, and the strength of a trend can be easily measured using the ADX (Average Directional Index) indicator.