Last Updated on Tuesday, 07 July 2015 04:04 GMT
4.4 Exponential Moving Averages
We noted earlier that the Simple Moving Average suffered from ‘lag’ and ‘spike’ problems, and then show a solution by WMA. However, in WMA the weight or importance given to each value decreases linearly as we move back in time. The linear decrease in the importance of the values as we move into the past still keeps the lag factor considerably high.
Here comes the Exponential Moving Average! (We’ll call this EMA, OK?) How does the EMA solve kid brothers SMA and the next one i.e. WMA’s problems?
It does this by refining the method of calculation so as to give much greater weightage (read: ‘importance’) to more recent prices. It adds a percentage of yesterday’s moving average to a percentage of today’s closing value. This is done with exponentially decreasing importance to the closing prices as we move backwards. The most recent closing price will have very high importance and this importance keeps decreasing exponentially. Now, please….don‘t…..panic! I don’t calculate it, nor will you, we’ll just leave it to the computer software. However, there is nothing wrong in knowing, right?
Calculation of Exponential moving Average
The first step is to calculate the Exponent or the factor which would changes the weight. Well, this exponent depends on the number of period e.g. days for daily EMA.
Exponent = 2 / (number of periods +1)
Let’s say you wish to calculate EMA for 22 periods then the exponent will be 2 / (22+1) = 0.08695
Current periods’ exponential moving average = (Previous period’s closing price * Exponent) + (Previous period's EMA * (1 - Exponent))
Let’s assume the following to make an example:
- You wish to calculate daily current EMA while using EMA of 22 periods
- Let’s say today is 20th April
- The closing price of April 19th was 135.80
- EMA of April 19th was 132.60
So the exponent = 2 / (22+1) = = 0.08695
Hence using the above formula, the EMA of April 20th = (135.80 * 0.08695) + (132.80) * (1 - 0.08695) = 135.41
You may now ask as to what happens if we are only on day 2 and wish to calculate the EMA as in that case the previous day’s EMA is not there? Simple, we take the closing price of the first day as the previous day’s EMA.
If we are on third day, we can take first 2 days’ simple moving average as the EMA of second day, to calculate the EMA of the third day.
Here’s a chart showing the price, SMA and EMA all plotted together:
See how the EMA picks up the trend quicker than the SMA. The circles mark the points where a strong trend takes hold. Within each circle you can see how the EMA turned in the direction of the price sooner than the SMA. This is because by giving more weightage to recent prices the EMA is more sensitive to a change in price trend than the SMA.
Another feature of the EMA is that it does not jump in response to the discarding of older data, and therefore spikes get eliminated.
However, don’t write off the SMA just yet! It is better suited to pinpoint zones of support and resistance because it relies on an average of the whole time period.
You’re thoroughly confused, and you’re praying, ‘by chance, just by chance, will they give me a checklist’? You bet we will, and I’ll do better – a whole chapter, the next one, will tell you how to choose among these two averages.
- Introduction of Exponential Moving Average (EMA)
- Which Moving Averages of What Time-frames are Best?
- Recap - Moving Averages - Q&A