Determining the direction of the trend is one of the most important parts of most good technical trading systems. Without the existence of a trend the market would essentially be random and impossible to trade profitably. Trades that are taken in the direction of the main trend therefore have a statistical advantage over trades taken against the trend. When the trend is moving in different directions on different time frames the longer term-trend usually wins with the shorter-term trends usually reversing to move in the same direction as the larger trend. Determining the trend is usually very easy as it is simply a matter of asking if the market is moving up, down or has no clear direction; some traders will simple answer this question for themselves with a quick glance of a chart (which is usually enough), others however seek a set of clearly defined hard and fast rules to answer the question. Some common methods for doing this are -
1. Moving Averages,
There are 3 ways that moving averages are typically used to determine the direction of the trend. The first and most common is the moving average crossover. A moving average crossover involves having two moving averages, one shorter than the other, and comparing the position of the shorter moving average to the longer one. If the shorter moving average is above the longer moving average the trend is said to be up – if the shorter moving average is below the longer moving average the trend is said to be down; the trend is therefore said to change direction when the moving averages 'cross'. One variation of the duel moving average crossover is the single moving average crossover where the direction of the trend is determined by which side of the moving average the price is on, that is, the trend is said to be up if the price (day/candle) closes above it's moving average and down if the price closes below it's moving average. Another variation of the moving average crossover method is the triple moving average crossover that involves having 3 moving averages of different sizes; the trend is only said to be up if the shorter moving average is above both the medium term and longer term moving averages and the medium moving average is also above the longer moving average – the trend said to be down when the shortest moving average is below the long and medium term moving averages and the medium moving term average is also below the long term moving average. Unlike the duel moving average and single moving average trend detection methods, the triple moving average has periods where no trend is given. Systems that use triple moving averages as entry signals are therefore not always 'in the market', where systems that use duel or single moving averages as entry signals are always giving a trend direction saying the trend is always up or down. Systems that use triple moving averages tend to be correct more often than the duel or single moving average systems as they are more likely to avoid periods where the market is 'choppy', but they are also usually slower to detect the existence of trends and provide less trading opportunities. As we would expect, the longer the moving averages time frame the more reliable they are.
5 Day Single Simple Moving Average
50 Day Single Simple Moving Average
5 Day/10 Day Duel Simple Moving Average Crossover
25 Day/50 Day Duel Simple Moving Average Crossover
5 Day/10 Day/15 Day Triple Moving Average Crossover
15 Day/25 Day/50 Day Triple Moving Average Crossover
2. Donchian channels,
A Donchian channel is an technical indicator developed by Richard Donchian, it consists of two lines that are formed by taking the maximum daily high and the minimum daily low of the last x number of days. A long signal is given when the upper Donchian channel is broken by a new high exceeding the highest high in the previous x number of days, and the lower Donchian channel is broken by a new low that is less than the lowest low in the previous x number of days. A variation of this channel that I have used in a number of systems listed in this website's Trading Systems Library is a channel consisting of the highest and lowest closing prices in the previous x number of days, where the channel is said to be broken when price closes outside the previous x number of days highest or lowest closing price. This slight modification helps to avoid false breakouts on shorter time frames – on longer time frames however, the modified system tends to be less effective than the standard Donchian system. The shorter the time frame the less significant a breakout is, using the closing price as we do in the modified Donchian channel system helps to avoid these false breakouts. However, on larger time frames where the breaking of the Donchian channel is more significant, the gains from avoiding false breakouts are not out-weighted by the pips that are given back or lost by waiting for the closing price instead of entering and exiting immediately. The system used by the Original Turtles involved using stand Donchian channel breakouts of 20 or 55 days. As with moving averages, the larger the period of time used for the channel, the more reliable the signal.
Here are the results of using standard and modified Donchian channel's on the EUR/USD currency pair from the beginning of 2000 to the end of 2009. The system is always in the market and goes long when the price breaks the upper channel line and stays long until the price breaks the lower channel line, and vice verse for shorts.
Standard Donchian Channel – Entering immediately, as soon as the high or low is broken (EUR/USD Results 2000 – 2009) -
Simply comparing today's price to the price a number of days ago is probably the simplest method of determining the direction of a trend. For example, we might say that the short-term trend was up if today's price closed above the closing price 8 days ago, and down if the price closed below the price that it closed at 8 days ago. For a medium-term trend we might use a time frame of 30 days and perhaps about 120 days to determine the direction of the long-term trend.
Here are the results of following the trend on the EUR/USD currency pair from the start of 2000 to the end of 2009 –
