movingaverages

Guide to Moving Averages

Moving
average is a term that is often being used when there is talk about trading,
but what exactly is it and how can you use it in your own analysis? You
probably know how to calculate the average of a pool of numbers but let us
start with the basics.
Average- and moving average- calculation
When you
have a series of numbers and you for some reason want the average of those
numbers you simply add all those numbers and divide the sum by how many numbers
you have. E.g. you have these five numbers: 8, 9, 10, 8 and 10. The sum of
these numbers is 45. Divide 45 by 5 (as you have five numbers here) and you get
an average of 9.
If the five
numbers listed above were price data for company ZZZ we could say that the last
five days average price for ZZZ is 9. Next day we read that ZZZ closed at 13.
If we were to calculate the new five day’s average it would become:
13 + 10 + 8
+ 10 + 9 = 50 and five day average will now become 50 / 5 = 10 (Fig. 1). What
we have done here is to “move” the five day period one step forward by
including the latest close and excluding the now sixth day close thereby
creating what we call a moving average. This procedure is repeated for every
new day so that we will get a series of moving average numbers that we can plot
as a line in our charts.
Fig. 1. Five day average calculation.
Plotting moving average
When doing
technical analysis we analyze charts where price and other data are plotted. To
be able to get something useful from moving average we should plot the data in
our charts together with the price curve (Fig. 2). In figure 2 you see price data
plotted along price and volume (light blue bars). On top of the price bars we
have plotted a five day moving average as a dark blue line. Taking a closer
look at this line we see that it is smoother than the actual price movement. If
we were to plot a longer moving average using e.g. 50 days moving average we
would see an even much more smoothed curve as the value plotted for a specific
day is the average of the number of days we have used in the calculation for
the moving average.
moving averages
Fig. 2. Plotting of a five day moving average. The
blue line seen in this chart is the five day moving average of the closing price
for this stock.
Types of moving average
The moving
average we have talked about in this article so far is what we call a simple
moving average. There are many other types of moving averages as well but the
most common one beside simple moving average is exponential moving average.
When applying exponential moving average, last days values are being weighted more
than earlier days thereby making last days movements “more important”. The
result of this is that the graph will react quicker to the price changes that
occur. In Figure 3 we have plotted two moving averages. The blue line is a 20
day simple moving average and the green line is a 20 day exponential moving average.
It is easy to see that there is a difference between the two lines. Even though
it is the same number of days that is being used in the calculation of the
moving averages we see that the green line (exponential moving average) is
following the price development a bit more closely than the more smoothed blue
line.
We earlier
said that we plot moving average along with price data, but it is also
important to remember that moving averages can be applied on all kinds of data
and not just on price data. Other examples could be moving average of volume
and relative strength index (RSI) to mention a few.
Fig. 3. Plotting of 20 days simple moving average
(blue line) and 20 days exponential moving average (green line). We see that
the green line react more to the price changes that occur than the blue line.
A basic moving average trading system
To see how
a moving average trading system could be used we will now design a very basic
moving average trading system consisting of two moving averages. The coding and
plotting of data shown here is done in a software called Amibroker
which is a technical analysis
software where you can plot data, program trading systems, optimize- and test-
your own trading ideas.
In this
example we will use a short 5 day simple moving average and a medium long 25
day simple moving average to form the framework for our trading system. What we
want to do next is to design our trading system to buy the stock when the 5 day
moving average crosses above the 25 day moving average and sell when the 5 day
moving average crosses below the 25 day moving average. The coding done to make
this system is seen in figure 4.
Fig. 4. Amibroker code for the moving average trading
system
When we plot
these two moving averages we see why we want to buy the stock when 5 day moving
average crosses above the 25 day moving average and sell when the opposite
occur. Basically what we want is to be on the “right” side as the stock price
moves upwards or downwards (Fig. 5).
Fig. 5 We want to own stocks in this company when the
blue line is higher than the green line, and sell (or short) the stock when the
blue line falls below the green line. The light blue bars seen in the lower
part of the chart is the trading volume for the different days.
From the
plot it looks like this system is quite promising with regard to being
profitable but to make sure that it is actually profitable we need to do some
more testing. What we want is a trading system that will perform well over time
and in all kinds of market conditions. Although the trading system seems
promising at a first glance the testing of this system shows something else. As
seen from the test report produced by Amibroker the system performance was
rather poor over the test period from mid June 2004 to mid June 2008. With a
win percent of 33.3 this is probably a trading system that we should leave and
not trade as is today. Although the trading system does not perform as wanted
the overall idea for the system might still be something to build further upon
(Fig. 6).
Fig. 6. Profit distribution for our trading
system. We see that we have a few nice returns of investment here but when we
just have a few winners we choose not to go further with this system. What we
want is a much more even profit distribution that could be e.g. 20 winners in
the 5% to 10% range rather than one winner of 100%. Only then can we say that
we have a more robust trading system.
It has to
be said that the testing done here is very simple and
stocktradersbulletin will look deeper into testing of
trading systems in an article at a later stage. Also it should be noted that
the testing here was done on one stock which is not a good enough foundation
for validating trading systems. 
Conclusion
Moving
average is a tool that could be used in a meaningful way if it is applied correctly.
From our
analysis in the previous section we can say that using moving average alone for
your trading will probably not give you the profit you would like to see in the
long run. You will simply see too many false signals for when to buy and sell.
To minimize the false signals you should use moving average in conjunction with
other technical analysis tools. What tools and how you want to do that is
entirely up to you but what you should remember to do if you have the chance is
to test, test and test your system and make sure that you are not just seeing
wishful things before you go ahead and do actual trading with real money.
Final note
What you
have read here is just one method of how you could use a moving average. As we
have seen the method presented in this article is not a very profitable one but
the intention here is to give you some information about moving average and
possibly also give you some inspiration from where you can continue with your
own ideas and development of trading systems. 

Posted by Judy Romero