Developed by Gerald Appel, Moving Average Convergence/Divergence (MACD) is one of the simplest and most reliable indicators available. MACD uses moving averages, which are lagging indicators, to include some trend-following characteristics. These lagging indicators are turned into a momentum oscillator by subtracting the longer moving average from the shorter moving average. The resulting plot forms a line that oscillates above and below zero, without any upper or lower limits. MACD is a centered oscillator and the guidelines for using centered oscillators apply.
The most popular formula for the "standard" MACD is the difference between a security's 26-day and 12-day Exponential Moving Averages (EMAs). This is the formula that is used in many popular technical analysis programs, including Sharp Charts, and quoted in most technical analysis books on the subject. Appel and others have since tinkered with these original settings to come up with a MACD that is better suited for faster or slower securities. Using shorter moving averages will produce a quicker, more responsive indicator, while using longer moving averages will produce a slower indicator, less prone to whipsaws. For our purposes in this article, the traditional 12/26 MACD will be used for explanations. Later in the indicator series, we will address the use of different moving averages in calculating MACD.
Of the two moving averages that make up MACD, the 12-day EMA is the faster and the 26-day EMA is the slower. Closing prices are used to form the moving averages. Usually, a 9-day EMA of MACD is plotted along side to act as a trigger line. A bullish crossover occurs when MACD moves above its 9-day EMA, and a bearish crossover occurs when MACD moves below its 9-day EMA. The Merrill Lynch (MER) chart below shows the 12-day EMA (thin blue line) with the 26-day EMA (thin red line) overlaid the price plot. MACD appears in the box below as the thick black line and its 9-day EMA is the thin blue line. The histogram represents the difference between MACD and its 9-day EMA. The histogram is positive when MACD is above its 9-day EMA and negative when MACD is below its 9-day EMA.
MACD measures the difference between two Exponential Moving Averages (EMAs). A positive MACD indicates that the 12-day EMA is trading above the 26-day EMA. A negative MACD indicates that the 12-day EMA is trading below the 26-day EMA. If MACD is positive and rising, then the gap between the 12-day EMA and the 26-day EMA is widening. This indicates that the rate-of-change of the faster moving average is higher than the rate-of-change for the slower moving average. Positive momentum is increasing, indicating a bullish period for the price plot. If MACD is negative and declining further, then the negative gap between the faster moving average (blue) and the slower moving average (red) is expanding. Downward momentum is accelerating, indicating a bearish period of trading. MACD centerline crossovers occur when the faster moving average crosses the slower moving average.
This Merrill Lynch (MER) chart shows MACD as a solid black line, and its 9-day EMA as the thin blue line. Even though moving averages are lagging indicators, notice that MACD moves faster than the moving averages. In this example, MACD provided a few good trading signals as well:
- In March and April, MACD turned down ahead of both moving averages, and formed a negative divergence ahead of the price peak.
- In May and June, MACD began to strengthen and make higher Lows while both moving averages continued to make lower Lows.
- Finally, MACD formed a positive divergence in October while both moving averages recorded new Lows.
MACD generates bullish signals from three main sources:
- Positive Divergence
- Bullish Moving Average Crossover
- Bullish Centerline Crossover
A Positive Divergence occurs when MACD begins to advance and the security is still in a downtrend and makes a lower reaction low. MACD can either form as a series of higher Lows or a second Low that is higher than the previous Low. Positive Divergences are probably the least common of the three signals, but are usually the most reliable, and lead to the biggest moves.
A Bullish Moving Average Crossover occurs when MACD moves above its 9-day EMA, or trigger line. Bullish Moving Average Crossovers are probably the most common signals and as such are the least reliable. If not used in conjunction with other technical analysis tools, these crossovers can lead to whipsaws and many false signals. Bullish Moving Average Crossovers are used occasionally to confirm a positive divergence. A positive divergence can be considered valid when a Bullish Moving Average Crossover occurs after the MACD Line makes its second "higher Low".
Sometimes it is prudent to apply a price filter to the Bullish Moving Average Crossover to ensure that it will hold. An example of a price filter would be to buy if MACD breaks above the 9-day EMA and remains above for three days. The buy signal would then commence at the end of the third day.
A Bullish Centerline Crossover occurs when MACD moves above the zero line and into positive territory. This is a clear indication that momentum has changed from negative to positive, or from bearish to bullish. After a Positive Divergence and Bullish Centerline Crossover, the Bullish Centerline Crossover can act as a confirmation signal. Of the three signals, moving average crossover are probably the second most common signals.
Even though some traders may use only one of the above signals to form a buy or a sell signal, using a combination can generate more robust signals. In the Halliburton (HAL) example, all three bullish signals were present and the stock still advanced another 20%. The stock formed a lower Low at the end of February, but MACD formed a higher Low, thus creating a potential Positive Divergence. MACD then formed a Bullish Moving Average Crossover by moving above its 9-day EMA. And finally, MACD traded above zero to form a Bullish Centerline Crossover. At the time of the Bullish Centerline Crossover, the stock was trading at 32 1/4 and went above 40 immediately after that. In August, the stock traded above 50.
MACD generates bearish signals from three main sources. These signals are mirror reflections of the bullish signals:
- Negative Divergence
- Bearish Moving Average Crossover
- Bearish Centerline Crossover
A Negative Divergence forms when the security advances or moves sideways, and the MACD declines. The Negative Divergence in MACD can take the form of either a lower High or a straight decline. Negative Divergences are probably the least common of the three signals, but are usually the most reliable, and can warn of an impending peak.
The FedEx (FDX) chart shows a Negative Divergence when MACD formed a lower High in May, and the stock formed a higher High at the same time. This was a rather blatant Negative Divergence, and signaled that momentum was slowing. A few days later, the stock broke the uptrend line, and the MACD formed a lower Low.
There are two possible means of confirming a Negative Divergence. First, the indicator can form a lower Low. This is traditional peak-and-trough analysis applied to an indicator. With the lower High and subsequent lower Low, the uptrend for MACD has changed from bullish to bearish. Second, a Bearish Moving Average Crossover (which is explained below) can act to confirm a negative divergence. As long as MACD is trading above its 9-day EMA, or trigger line, it has not turned down and the lower High is difficult to confirm. When MACD breaks below its 9-day EMA, it signals that the short-term trend for the indicator is weakening, and a possible interim peak has formed.
The most common signal for MACD is the moving average crossover. A Bearish Moving Average Crossover occurs when MACD declines below its 9-day EMA. Not only are these signals the most common, but they also produce the most false signals. As such, moving average crossovers should be confirmed with other signals to avoid whipsaws and false readings.
Sometimes a stock can be in a strong uptrend, and MACD will remain above its trigger line for a sustained period of time. In this case, it is unlikely that a Negative Divergence will develop. A different signal is needed to identify a potential change in momentum. This was the case with Merck (MRK) in February and March. The stock advanced in a strong uptrend, and MACD remained above its 9-day EMA for 7 weeks. When a Bearish Moving Average Crossover occurred, it signaled that upside momentum was slowing. This slowing momentum should have served as an alert to monitor the technical situation for further clues of weakness. Weakness was soon confirmed when the stock broke its uptrend line and MACD continued its decline and moved below zero.
A Bearish Centerline Crossover occurs when MACD moves below zero and into negative territory. This is a clear indication that momentum has changed from positive to negative, or from bullish to bearish. The centerline crossover can act as an independent signal, or confirm a prior signal such as a moving average crossover or negative divergence. Once MACD crosses into negative territory, momentum, at least for the short term, has turned bearish.
The significance of the centerline crossover will depend on the previous movements of MACD as well. If MACD is positive for many weeks, begins to trend down, and then crosses into negative territory, it would be bearish. However, if MACD has been negative for a few months, breaks above zero, and then back below, it might be a correction. In order to judge the significance of a centerline crossover, traditional technical analysis can be applied to see if there has been a change in trend, higher High or lower Low.
The Unisys (UIS) chart depicts a Bearish Centerline Crossover that preceded a 25% drop in the stock that occurs just off the right edge of the chart. Although there was little time to act once this signal appeared, there were other warnings signs prior to the dramatic drop:
- After the drop to trend line support, a Bearish Moving Average Crossover formed.
- When the stock rebounded from the drop, MACD did not even break above the trigger line, indicating weak upside momentum.
- The peak of the reaction rally was marked by a shooting star candlestick (blue arrow) and a gap down on increased volume (red arrows).
- After the gap down, the blue trend line extending up from Apr, 1999, was broken.
In addition to the signals mentioned above, a Bearish Centerline Crossover occurred after MACD had been above zero for almost two months. From 20 Sept on, MACD had been weakening and momentum was slowing. The break below zero acted as the final straw of a long weakening process.
As with bullish MACD signals, bearish signals can be combined to create more robust signals. In most cases, stocks fall faster than they rise. This was definitely the case with Unisys (UIS), and only two bearish MACD signals were present. Using momentum indicators like MACD, technical analysis can sometimes provide clues to impending weakness. While it may be impossible to predict the length and duration of the decline, being able to spot weakness can enable traders to take a more defensive position.
In 2002, Intel (INTC) dropped from above 36 to below 28 in a few months. Yet it would seem that smart money began distributing the stock before the actual decline. Looking at the technical picture, we can spot evidence of this distribution and a serious loss of momentum:
- In December, a negative divergence formed in MACD.
- Chaikin Money Flow turned negative on December 21.
- Also in December, a Bearish Moving Average Crossover occurred in MACD (black arrow).
- The trend line extending up from October was broken on 20 December.
- A Bearish Centerline Crossover occurred in MACD on 10 Feb (green arrow).
- On 15 February, support at 31 1/2 was violated (red arrow).
For those waiting for a recovery in the stock, the continued decline of momentum suggested that selling pressure was increasing, and not about to decrease. Hindsight is 20/20, but with careful study of past situations, we can learn how to better read the present and prepare for the future.
One of the primary benefits of MACD is that it incorporates aspects of both momentum and trend in one indicator. As a trend-following indicator, it will not be wrong for very long. The use of moving averages ensures that the indicator will eventually follow the movements of the underlying security. By using Exponential Moving Averages (EMAs), as opposed to Simple Moving Averages (SMAs), some of the lag has been taken out.
As a momentum indicator, MACD has the ability to foreshadow moves in the underlying security. MACD divergences can be key factors in predicting a trend change. A Negative Divergence signals that bullish momentum is waning, and there could be a potential change in trend from bullish to bearish. This can serve as an alert for traders to take some profits in long positions, or for aggressive traders to consider initiating a short position.
MACD can be applied to daily, weekly or monthly charts. MACD represents the convergence and divergence of two moving averages. The standard setting for MACD is the difference between the 12 and 26-period EMA. However, any combination of moving averages can be used. The set of moving averages used in MACD can be tailored for each individual security. For weekly charts, a faster set of moving averages may be appropriate. For volatile stocks, slower moving averages may be needed to help smooth the data. Given that level of flexibility, each individual should adjust the MACD to suit his or her own trading style, objectives and risk tolerance.
One of the beneficial aspects of the MACD is also one of its drawbacks. Moving averages, be they simple, exponential or weighted, are lagging indicators. Even though MACD represents the difference between two moving averages, there can still be some lag in the indicator itself. This is more likely to be the case with weekly charts than daily charts. One solution to this problem is the use of the MACD-Histogram.
MACD is not particularly good for identifying overbought and oversold levels. Even though it is possible to identify levels that historically represent overbought and oversold levels, MACD does not have any upper or lower limits to bind its movement. MACD can continue to overextend beyond historical extremes.
MACD calculates the absolute difference between two moving averages and not the percentage difference. MACD is calculated by subtracting one moving average from the other. As a security increases in price, the difference (both positive and negative) between the two moving averages is destined to grow. This makes its difficult to compare MACD levels over a long period of time, especially for stocks that have grown exponentially.
The Amazon (AMZN) chart demonstrates the difficult in comparing MACD levels over a long period of time. Before 1999, Amazon's MACD is barely recognizable, and appears to trade close to the zero line. MACD was indeed quite volatile at the time, but this volatility has been dwarfed since the stock rose from below 20 to almost 100.
An alternative is to use the Price Oscillator, which shows the percentage difference between two moving averages:
(12 day EMA - 26 day EMA) / (26 day EMA)
(20 - 18) / 18 = .11 or +11%
The resulting percentage difference can be compared over a longer period of time. On the Amazon chart, we can see that the Price Oscillator provides a better means for a long-term comparison. For the short term, MACD and the Price Oscillator are basically the same. The shape of the lines, the divergences, moving average crossovers and centerline crossovers for MACD and the Price Oscillator are virtually identical.Since Gerald Appel developed the MACD, there have been hundreds of new indicators introduced to technical analysis. While many indicators have come and gone, the MACD has stood the test of time. The concept behind its use is straightforward, and its construction is simple, yet it remains one of the most reliable indicators around. The effectiveness of the MACD will vary for different securities and markets. The lengths of the moving averages can be adapted for a better fit to a particular security or market. As with all indicators , MACD is not infallible and should be used in conjunction with other technical analysis tools.