Moving Average - how to successfully smooth out shorter fluctuations

Moving averages are one of the most fruitful approaches to determining and profiting from trends.

A moving average is computed for each consecutive period interval and represents a fixed period average of prices.

The resulting line is smooth and reflects the consecutive, average prices.

One of the main reasons for using moving averages is to smooth out erratic data, making it easier to see the actual underlying trend and focus on the time horizon that fits with the trading approach.

The nature of a moving average is such that it just one value representing a set of specific past numbers.

While the majority of moving averages of prices are based on closes, they can be calculated on highs, lows, or any other value as long as the price used is constant throughout the calculations.

Moving averages are plotted on a price chart and the short term oscillations are smoothed so that one can see the main trend without being distracted by the small, erratic movements.

What’s the most important thing we look for?

A rising moving average signals a bullish trend, while a falling moving average signals a bearish trend.

Even though the moving average helps us determine a trend, it does so after the trend has started.

Since one wants to be trading with the trend, we have to remember that the moving average is a lagging indicator that will always give us some delay in signaling a change in trend.

The secret to choosing the most useful length

Moving averages can be calculated for different lengths of time, - but which length is the most useful?

A longer time period provides more information because it includes more data observations, with each bar’s data becoming less significant in the calculation.

Consequently, a single large change in the value does not have a large influence on the longer moving average.

If this large move is an irregular outlier - it can represent an advantage.

However, if this large change represents the start of an important trend change, it would take longer to discern the underlying trend change.

The bottom line is this.

The longer the interval the less responsive moving average is at picking up trend changes but also less likely to signal false trend change because of a short-term data blip.

What variations exist and why?

The simple moving average (SMA), otherwise known as arithmetic moving average, is the most commonly used type of moving average.

It is calculated by adding a group of data and then dividing by the number of events in the period being examined.

A 10-day moving average, for example, is just one value representing all the prices for the last 10 days.

It filters out each one of the prices during the last 10 days and shows us how the group of 10 days is behaving, rather than its separate parts.

Each one of the prices for the last 10 days is filtered out and the moving average tells us how the set of 10 days is behaving, instead of its separate parts.

Using a 10-day SMA as an example, the data contained in the price for each of the twenty bars is given equal importance.

In some contexts, however, the more recent price could have more bearing on the future market direction than the twenty -day old market price does.

If more recent price data are more relevant than earlier observations, one would want to weight data in favor of the more recent price.

In case of a weighted moving average, the most recent price is weighted more heavily, giving the most recent observation more importance in the calculation.

LWMA - linearly weighted moving average

A twelve-bar linearly weighted moving average, for example, multiplies the twelfth bar data by 12, the eleventh bar by 11, and the tenth bar by 10, and so forth. The total of these numbers are added up and divided by the sum of all the multipliers.

When using this twelve-bar moving average weighting approach, the most recent pricing data is given two times the importance of the price six days earlier and twelve times the importance of the price twelve bars earlier.

As we calculate the twelve-bar linearly weighted moving average for bar 13, the prices for trading bars 2–12 again will be weighted and the earliest bar data is dropped from the data set being used in the calculation.

Solving the “drop-off” problem of with the exponential smoothing

For some analysts, dropping off the earliest trading day’s data that occurs with an SMA or linearly weighted moving average is problematic.

In some cases, discarding the earliest bar’s data that happens with moving average represents an issue.

If the more recent price shows very little change, but the earliest price, now being discarded, reflects a considerable change, the moving average can be unevenly influenced by the omitting of the older data.

So what’s the point?

A significant change in the moving average that comes from the omission of early data potentially triggers a false signal.

Known as the “drop-off effect”, it is the most unflattering aspect of SMA.

While it can be argued that very early data is not necessarily as relevant in determining future price movement as the most recent observations, it is still data that may be valuable.

Both the SMA and the LWMA ignore this older data located outside of the length of the moving average.

The exponential moving average is used to address this problem and to keep this older data in the moving average calculation.

Let us take a look at the calculation of the exponential moving average based on a simple ten-bar moving average.

To calculate the exponential moving average, we will use both ten-bar SMA and the closing price for bar 11, - the total of 11 bars of price information, with each bar’s data having a weight of 9.09% or 1/11 in the calculation.

A larger weight, however, needs to be placed on information that is more recent.

If we want, for instance, the price from bar 11 to weigh twice as much as it would have in SMA, it would have a weight of 18.18%, or 2/11.

The total of all the weights must add up to 100%, leaving 100% minus 18.18%, or 81.82% weight to be placed on the ten-bar moving average.

Wilder method of finding trends faster with fewer whipsaws

Welles Wilder proposed a very simple method to calculate a moving average that weights the most recent data more heavily.

The formula for calculating Wilder’s moving average is as follows:

MA[bar i] = ((n – 1) × MA[bar i–1] + Price[bar i] ) ÷ n

Geometric moving average (GMA)

Used primarily in indexes, the geometric moving average is an SMA of the percent difference between the previous bar and the current bar over some fixed period.

Its range or dimensions remains unaffected by the choice of percentages rather than points.

However, the problems of equal weight and lag remain unsolved.

Double smoothing the data with triangular moving average

A doubly smoothed moving average is a result of applying a moving average calculation process to an already calculated moving average.

Basically, we take a series of prices and apply the moving average formula, then, to make it for an even closer fit, we apply the averaging formula one more time.

The triangular moving average starts with an SMA of a fixed number of bars and then calculates a moving average of a length of half the initial number of bars, such as a 20-day SMA smoothed in a ten-day SMA.

What makes this so special?

The benefit of this method is that it better represents the trend by doubly smoothing the data.

Responding to volatility challenge with variable EMA

This moving average is similar to EMA, the difference being the weighting scheme that is adjusted based on the volatility to reduce the number of adverse signals during a trading range.

During narrow volatility trading ranges variable EMA becomes shorter but expands when the price starts to trend.

4 strategies built with the help of moving averages

Being a basic tool with a wide range of applications, moving averages are used extensively.

They are used in four basic ways:

as a trend determination,
to see the support and resistance,
to spot price extremes, and
for explicit trading signals.

Determining trend

Comparing the moving average with the current price represents the most common usage of moving averages.

If the market is above its n-bar moving average, the trend is defined as bullish.

Conversely, if the market is below the n-bar moving average, the trend is thought of as bearish.

Determining support and resistance.

Prices seem to halt at the vicinity of moving averages.

What it means is this.

The moving average regularly duplicates the trend line by acting as support or resistance.

Therefore, it can be used as practical approach for trailing stops to determine when a position should be reduced or liquidated.

Determining price extremes.

The moving average also delineates the price extremes.

Any reversion to the mean will have a tendency to approach the moving average, because the moving average itself is a mean.

Here’s the thing.

When the current price has significantly moved away from the moving average, the reversion that follows can be traded profitably.

Price habitually reverts to the mean.

Since prices are likely to revert to that mean, the divergence becomes an opportunity to open position against the trend.

The reversion also offers a chance to trade in the direction of the dominant trend when it happens.

When prices move away from the trend substantially, it is often an indication that the trend direction may be changing.

If these occurrences are in sync with the main trend, they offer reduced risk and increased profit potential.

Giving specific signals.

Moving averages can also offer specific signals.

These happen:

when prices cross a moving average,
when a faster moving average crosses a slower moving average, and in some cases,
when a third, even faster, moving average crosses two slower ones.

We do not have to limit ourselves to the information offered by a single moving average.

Looking at various moving averages of different lengths simultaneously can increase our data set.

For example, a support or resistance often coincides with the crossover of two moving averages.

The point is this.

When the faster moving average crosses above the slower is often taken as a sign that the price trend is bullish.

Similar, it is considered a bearish signal when the faster declines below the slower.

Using moving averages as the key determinant of trend and then applying faster moving averages as signals represent a foundation upon which a number of successful trading strategies are built.

In other cases, chart patterns are used to signal entry and exit after moving averages are used to determine the trend.

But it all comes down to this.

A trend has to be in existence for a moving average a crossover to be profitable.

Otherwise, a number of signals will be false and produce small losses while expecting the one signal that will produce the large profit.

This approach can be a highly profitable if one has the discipline to handle with the small losses, and it often is the foundation for many long-term trend systems, demonstrating how one can profit while still losing on a majority of small trades.

Conclusion

Moving average signals are most effective primarily during a trending market, which is where the majority of the traders make their money.

Therefore, trend determination is the most practical use of moving averages.

As soon as a trend has been identified, the next step is to use breakouts and price patterns in the direction of the trend for the timing of entries.

This approach will minimize losses and accrue profits, but may lag behind the major tops and bottoms.

Try It Yourself

After all the sides of the indicator were revealed, it is right the time for you to try either it will become your tool #1 for trading.

In order to try the indicator performance alone or in the combination with other ones, you can use Forex Tester with the historical data that comes along with the program.

Simply download Forex Tester for free. In addition, you will receive 23 years of free historical data (easily downloadable straight from the software).

Please let us know about your experience using moving averages. We would love to get your feedback. Any comments or observations would be greatly appreciated.