|
An
exponential moving average is another type of moving average.
In a simple moving average, the price data has an equal weight
in the computation of the average. Also, the oldest price
data is removed from the moving average as a new price is
added to the computation. The exponential moving average assigns
a weight to the price data as the average is calculated. Thus,
the oldest price data in the exponential moving average is
never removed, but it has only a minimal impact on the moving
average.
This
study displays the exponential moving averages as a crossover
system. You may select up to three different averages. Generally,
the lengths are short, intermediate, and long term. A commonly
used system is 4, 9, and 18 intervals. An interval may be
in ticks, minutes, days, weeks or months; it is a function
of the chart type.
A
buy signal occurs when the short and intermediate term averages
cross from below to above the longer term average. Conversely,
a sell signal is issued when the short and intermediate term
averages cross from above to below the longer term average.
You can use the same signals with two moving averages, but
most market technicians suggest using longer term averages
when trading only two exponential moving averages in a crossover
system.
Another
trading approach is to use the current price concept. If the
current price is above the exponential moving averages, you
buy. Liquidate that position when the current price crosses
below either moving average. For a short position, sell when
the current price is below the exponential moving average.
Liquidate that position when current price rises above the
exponential moving averages.
As
you use exponential moving averages, do not confuse them with
simple moving averages. An exponential moving average behaves
quite differently than a simple moving average. It is a function
of the weighting factor or length of the average.
Parameters
(4,9,18)
- Period1
- the number of bars, or period, used to calculate the first
moving average. FutureSource uses a default setting
of 4.
- Period2
- the number of bars, or period, used to calculate the second
moving average. FutureSource uses a default setting
of 9.
- Period3
- the number of bars, or period, used to calculate the third
moving average. FutureSource uses a default setting
of 18.
Computing
the Exponential Moving Average
The
formula to calculate an exponential moving average is as follows:
EMAt
= EMAt-1 + (k * (Pt - EMAt-1))
EMAt
is the exponential moving average for the current period.
EMAt-1
is the exponential moving average for the previous period.
Pt
is the price for the current period.
k
is the exponential smoothing constant.
FutureSource
does not ask you to specify the smoothing constants. It asks
you to specify the length of the moving average. You can then
determine the smoothing constant from the formula listed below.
If
you specify an exponential moving average length of 10, the
smoothing constant is 0.18. The formula to determine the smoothing
constant is:
k
= 2 / (n + 1)
k
is the smoothing constant
n
is the length of the moving average.
Now,
substitute the above values in the formula.
k
= 2 / (10 + 1) = 2 / 11 =.1818
Conversely,
if you know the smoothing constant, you must derive the length
of the moving average. In this example, use a smoothing constant
of .125 You can solve the above equation for the value of
n, which produces the following formula:
n
= (2 / k) - 1
Now,
substitute the above values for the equation.
n
= (2 /.125) - 1 = 16 - 1 = 15
Please
remember that FutureSource always asks for the length of the
moving average, not the smoothing constant. If you know the
smoothing constant, you use the above formula to determine
the length of the moving average. If you set the length of
the moving average and want to know the smoothing constant,
use the formula to solve for the smoothing constant or k.
Note:
There are several variations to the above formula.
|
