Application #6 - Chapter 14 - Momentum II: Individual Indicators - MACD

A trend-deviation indicator can be used for overbought and oversold zones but its defining strength comes into play when used together with trend lines and moving-average (MA) crossovers. Trend-deviation is also known as a "price oscillator" in some charting products. But what does trend-deviation mean?

Trend-deviation is a fancy way of saying price is moving away from the trend it's being viewed in relation to. For example, when price dips underneath an upward trend line that's connected by troughs (shown as point A in the hastily drawn Image 1, below).

Image 1: A visual representation of trend-deviation

Therefore, it's safe to conclude that a trend-deviation indicator helps identify true breaks in a trend. If it wasn't possible for false breakouts to happen, then trend-indicators probably wouldn't exist as drawing a trend line is more straightforward.

If you're going to use an acronym, it's got to be for the mouthful that is the Moving-Average Convergence Divergence (MACD) trend-deviation indicator. This isn't the only trend-deviation indicator available but it is one that can be seen as a standard in many user interfaces.

MACD works by using two exponential MAs (EMAs), the shorter being subtracted from the longer. Exponential means that the most recent periods are more heavily weighted, unlike in a simple MA where there is no weighting. An additional line is then added which represents the two EMAs, smoothed by a third, known as the "signal line". The name is straightforward to understand, since when the two lines crossover is when a buy or sell signal is generated. See Image 2 to see how these are inputted in TradingView.

Image 2: MACD Construction. Lines 1 & 2 represent the MACD line. Line 3 represents the signal line.

12, 26, 9 seem to be a common standard in trading products. Gerald Appel suggests settings for daily charts buy signals at 8, 17, 9 and sell signals at 12, 25, 9. It seems that signal smoothing at 9 is the common ground between all three. MACD can be used on any time frame but these are suggested on the back of considerable research. As a caveat, this research was published in 1974 so there are more than likely to be updates to this given advancements in computing capability.

Image 3 - Example buy and sell signals on hourly BTCUSD. Green = buy/long. Red = sell/short. Yellow arrow indicates fakeout.

Using the default 12, 26, 9 setting on the hourly Bitcoin chart in Image 3, buy and sell signals have been plotted as examples. A one bar confirmation is used before marking it as a signal i.e. compulsory minimum of two opposing consecutive colors.

Notice that, while all the signals indicated result in positive or near positive outcomes, there is one "fakeout" semi-signal pointed out by the yellow arrow. "Semi" in this instance is used to describe the fact that while there is a signal line crossover, it occurs in the upper half of the MACD chart so is prone to being unreliable. It is therefore discounted as a signal.

There are a few examples in Image 3 where using other pieces of evidence may have resulted in the correct action being taken earlier. It's clear, therefore, that MACD should be used alongside other evidence (notably trend lines) to provide the greatest accuracy and returns.