Hit the "More" button in the bottom
right corner of the graphing area. This causes a number of pulldown menus
and parameter fields to be displayed.
Hit the "Overlay" pulldown menu.
Select one of the overlays listed below the
horizontal line in the pulldown menu.
(The extra pulldown menus and parameter fields can
be hidden again by hitting the "Less" button.)
Each of the overlays can be parameterized through the input fields to the right
of the overlay button, as described below. After the parameters are set, it is
necessary to hit the "Apply" button for the overlay to appear in the graph.
Any or all of the overlays can be removed by clicking
the mouse when it is in the graphing area and selecting the overlay to be removed.
SMA: Simple Moving Average
Overview
The Simple Moving Average is the average of the n most recent prices. Moving
averages, in general, smooth out a data series and hide some of the "noise"
in the data. They make it easier to identify the direction of price movements
and the general trend trend. Because only previous data is used to compute a
moving average, a moving average always lags behind the actual prices. As a
result, moving averages will not predict a change in trend, but rather follow
behind the current trend. Therefore, they are best suited for trend identification
and trend following purposes, not for prediction.
Calculation
For each point in time, t, the Simple Moving Average, A(t), is defined to
be equal to
A(t) = ( P(t) + P(t-1) + ... + P(t-n-1)
) / n
where P(i) is equal to the price at time i, P(i-1)
is equal to the price at the time corresponding to the previous interval,
P(i-2) is equal to the price at the time corresponding to the interval before
that, and so on. In each case, the price taken is the mid-price between the
minimum bid and the maximum ask price obtained in the interval.
Parameters
The Simple Moving Average overlay has one parameter, namely n, the number
of most recent periods over which the average is calculated.
The time period over which the average is taken
also depends on the selected granularly of the graph. If parameter n
is 3, and the selected size of period is 5 seconds, then the Simple Moving
Average curve depicts, for each point, the average of 3 mid-prices for each
of the preceding 3 periods, and thus takes prices over the last 15 seconds
into account. If the , but if the selected size of period is 1 day, then with
n=3, prices over the last 3 days are taken into account.
Typical choice of parameter varies from trader
and depends in part on how many moving averages are overlaid at the same time:
5, 14, 21, 60, and 90 are typical choices
when using only 1 moving average
(4,20), (5,60), (7,90) are combinations often
used when overlaying two moving averages
(4,9,18), (5,20,60), (7,21,90) are combinations
often used when overlaying three moving averages.
Interpretation
Moving averages are primarily used to identify trends. However, they are often
used as part of more complex indicators, and some traders overlay multiple
moving average curves to predict changes in the trend. Some traders using
moving averages to generate buy/sell signals:
Price crossover signal: When
a single moving average crosses the exchange price, this is sometimes considered
as a signal. Thus, the moving average crossing below the price curve is
interpreted as a buy signal, and a moving average crossing above a price
curve is interpreted as a sell signal.
Double crossover signal: Some
traders identify trading signals at points where two moving averages intersect.
A buy signal is interpreted from when a shorter moving average (one calculated
with a fewer number of periods) moves upwards and crosses a longer moving
average, and a sell signal is interpreted from when a shorter moving average
moves downward and crosses a longer moving average.
Triple crossover signal: This
method uses three moving averages, each with its parameter set differently.
The intersection of the two faster moving averages is interpreted as a warning
signal, and the intersection of the two slower moving averages is interpreted
as a trading signal if it follows a warning signal.
EMA: Exponential Moving Average
Overview
Exponential Moving Averages are similar to Simple Moving Averages except that
more recent rates are weighted more strongly when computing the average. Hence,
Exponential Moving Averages tend to follow the rates more closely than Simple
Moving Averages.
Calculation
Given parameter n,
A(t) = (K * ( P(t) - A(t-1)) + A(t-1)
where
K is referred to as the smoothing function:
K = 2 / (1 + n)
P(t) is the mean between the highest bid price
and the lowest ask price during the interval t, and
A(t-1) is the value of the Exponential Moving
Average in the previous period
Parameter
EMA has one parameter, namely n that specifies how strongly the moving average
should follow the currency prices. If n is chosen small, then the EMA will
closely track the currency prices; if large, it will lag behind more.
Interpretation
The Exponential Moving Average can be interpreted similarly to the Simple
Moving Average. Sometimes EMA is used to identify breakouts: if the currency
prices cross above the EMA so that the lowest ask price of several periods
is above the EMA, then this can be interpreted as a buy signal. Conversely,
if the currency prices cross below the EMA so that the highest bid price of
several periods is below the EMA, then this would be interpreted as a sell
signal.
WMA: Weighted Moving Average
Overview
The Weighted Moving Average is a linearly weighted moving average where more
recent price points are weighted more heavily than with Simple moving
averages.
Calculation
The WMA overlay is formally defined as follows:
n is the number of periods over which the
average is to be calculated, and
P(t) is the currency price at period t, where
the currency price is taken as the mean between the highest bid-price and
the lowest ask price during the interval
Parameter
WMA has one parameter, n, specifying over how many periods the average should
be calculated.
BB: Bollinger Bands
Overview
Bollinger Bands are due to John Bollinger, and consist of three curves:
a simple moving average in middle
an upper band corresponding to a simple moving
average plus a constant, k, times the standard deviations
a lower band corresponding to a simple moving
average minus a constant, k, times the standard deviations.
The latter two curves are symmetrical and form an envelope around the first
curve.
Calculation
Creating Bollinger Bands involves four steps:
calculate the simple moving average, using
n as a parameter specifying the number of periods to include
calculate the standard deviation of the mid-prices
over the last n periods. See the section on Standard
Deviation on how to calculate it.
calculate the upper band as:
UB(t) = SMAn(t) + m * STDDEVn(t)
where SMAn(t) is the Simple Moving Average over n periods at time t,
STDDEVn(t) is the Standard Deviation of the mid-prices over the last n periods,
and
m is a second parameter.
calculate the lower band as:
LB(t) = SMAn(t) - m * STDDEVn(t)
where SMAn(t) is the Simple Moving Average over n periods at time t,
STDDEVn(t) is the Standard Deviation of the mid-prices over the last n periods,
and
m is a second parameter.
Parameters
Bollinger Bands entail two parameters:
n: the number of periods used in calculating
the Simple Moving Average and the Standard Deviation
m: a multiplicative factor specifying how
tight the bounds should be made around the Simple Moving Average.
Typical values for n are 14 or 20, and in that case
m is often set to 2. For higher values of n (such as 50 or 90), m is typically
set to a higher value, such as 2.5 or 3.
Interpretation
Because standard deviation measures volatility, the bands widen when the currency
prices are volatile and they are tight when there is not much volatility.
Some argue that changes in the trend change significantly when the bands are
tight; i.e. when volatility is low. Bollinger bands give a reference point
as to what is currently considered to be high or low. When currency
prices go over the band, they would be considered relatively high, and conversely,
when the prices go below the band, they are considered relatively low.
Bollinger Bands are not typically used to generate signals themselves; rather,
they are typically used together with, and as support for, other indicators.
