Basic Interpretation of the MACD.

INDICATORS: MACD

One of the recurring themes of this book is that traders must have a complete understanding of any tool they use. It is not constructive to use a tool that is simply a mysterious wiggly line without really understanding what it is saying about price action.

With this in mind, Appendix B takes a deep look at how the modified MACD indicator is constructed, and how it reacts to specific price patterns. If you have background and experience with this or a similar tool, this material may offer a new perspective on the use of this indicator. Though I have found this tool useful, many traders will have other favorite indicators that they will prefer to use. 

This is fine, and many of the applications of the modified MACD I discuss can be applied to other momentum indicators. However, be sure you understand what your indicator is really measuring. In general, most indicators fall into two broad categories: momentum indicators such as MACD, rate of change (ROC), and Momentum, or overbought/oversold indicators such as the Stochastic oscillator and the Relative Strength Index (RSI). 

The MACD is a momentum indicator and these techniques can be adapted for other momentum indicators, but do not try to apply them to overbought/oversold indicators. Many people do not think beyond the fact that their indicator has two moving lines. As always, think deeply and understand the nuances of the tools you are using.

Basic Interpretation of the MACD

Let’s turn our attention to some practical applications of the MACD, and how it might be used in actual trading. Formally and precisely, the MACD measures the changes in momentum of prices, but there tends to be some persistence in momentum, so it is acceptable to treat the MACD as a proxy measure of momentum. (That is to say, an increase in the rate of change of momentum will likely lead to higher momentum and vice versa.) 

Most of the benefit from using this tool comes from its ability to pick out inflections that might not be clearly visible on price charts, to identify swings that are more or less likely to have continuation after consolidation, and to mark overextended points where the momentum is potentially exhausted. There are a few specific applications and patterns that can be applied to this tool or most other momentum indicators. 

Fast Line Pop One of the core ideas of technical analysis, and indeed of fundamental analysis, is that markets move to new prices in response to new information. This movement usually takes the form of a series of alternating waves and retracements. Simply put, a strong momentum move will usually result in another move in the same direction following a period of consolidation. This pattern is the foundation of market structure, but it is also possible to use a momentum indicator to define these conditions precisely. 

The most important concept here is that momentum precedes price, meaning that a sharp momentum move in a market pressing to new highs will usually be followed by higher prices after a consolidation, and the reverse is true to the downside. Be clear on this point: a momentum indicator will not somehow lead prices—this is not possible because the indicators are calculated from past prices. There are no true leading indicators in technical analysis; what does lead price is a strong move in the market itself, not the reading on an indicator. 

One of the simplest ways to use a momentum indicator is to look for it to make a significant new high or low, and then to enter the pullback following that new momentum extreme. One complication is that the MACD is an unbounded indicator, so it is not possible to set fixed reference levels. In practice, it is enough to use rough intuitive guidelines created by comparing the indicator to its own recent history. A more systematic approach could, for instance, express the indicator’s value as a percentile of a look-back window, and perhaps enter retracements following an excursion into either the top or the bottom decile. 

If you are just starting to look at this indicator, use the 40-bar history as a starting point, but you should be able to read the indicator without precise reference to this history very quickly. Again, the momentum pop on the MACD is just a visual representation of information that is already in the price bars, but it can be a useful confirmation in some situations. Figure 7.11 shows the MACD applied to weekly bars of the Financial Sector SPDR (NYSE: XLF). 

At point A the market had made a strong down move, which also pushed the MACD line to a significant new low relative to its recent history. (Do not focus too much attention on the fact that it is a new low on this particular chart. The chart boundaries are arbitrary.) After an extended consolidation, the market made new lows into the point marked B, at which time the MACD again made a lower low. Following more consolidation, the market traded lower until the sharp reversal pushed the MACD to significant new highs (C). 

This upside momentum should have been a warning to shorts to not look to add short exposure into the next pullback. In this case, the market staged a substantial multiyear rally from this point. This is an important concept, so let’s review a more complex example, this time using a 5-minute chart of the E-mini S&P 500 futures. At the point marked A in Figure 7.12, the MACD confirmed the price weakness by making a significant new low, and prices slid lower with scarcely any consolidation. 

The action into 11:00 AM was perhaps slightly suggestive of a selling climax, so very short-term shorts would have been well advised to take profits into the small flush. Point B raises an important issue: is this a significant new high on the MACD? Perhaps so, as this is the highest reading that has been seen so far in this trading day, but there were other considerations (on the higher time frame) that suggested continued weakness. 

What if a trader has misread this MACD line and had entered a long trade? A losing trade would have been the result, but losing trades are a normal part of any trading plan. If the risk on this trade had been managed appropriately, the loss would have been small. Points C and D both show new price lows on the day, 

accompanied by at least marginal new lows on the MACD, but it is clear that the character of the market has changed somewhat. Though there may have been money shorting into retracements following these two points, they are not examples of the best possible trades.

Fast Line Behavior in Climax

Consolidations or pullbacks following climaxes are not high-probability trades for trend continuation. Reversals or long, extended consolidations are more likely following these points. Though it is not always possible to distinguish a climax from good strength or weakness, it is important to keep this consideration in mind. If an extremely large momentum move emerges that is out of all proportion to the recent history of the market, be very careful of entering retracements. This is a case where the details of the mathematical construction of the MACD become significant. It is an unbounded indicator, so extreme market conditions can push it to theoretically infinite levels. In these extreme situations, the indicator will rescale so that recent history is highly compressed in a tight wiggly line. Rather than trying to read any significance into that line, accept that it shows that the market has made such an extreme move that the indicator should be disregarded. Figure 7.13 shows this principle in action on a daily chart of Silver futures. After pressing to new highs, the market collapsed in a very sharp sell-off at point A. Note

that the MACD fast line definitely made a significant new low, but this was clearly a climactic condition. Naive application of the MACD might have had a trader entering short in the first bounce off those lows, and a disciplined profit-taking plan would actually have locked in some profits in this case. The point is clear: this is not a reliable pattern to short. Consider also that a bounded indicator (e.g., stochastics) would simply have gone to the boundary and stayed there, rather than making this dramatic pop. Traders who use the MACD on intraday charts will see these types of distortions frequently if the markets they trade have large overnight gaps. It is usually best to disregard the indicator until the market has settled down and more normal momentum conditions emerge. 

Fast Line Divergence

If the best pullbacks are preceded by the MACD fast line making a significant new high (or low in the case of a downtrend), then it follows logically that pullbacks that are not set up by a new high or low on the MACD are less likely to have good continuation. This is the concept of momentum divergence: a market makes a new high in prices that is not accompanied by a new high on the indicator, and vice versa to the downside. Figure 7.14 shows clean momentum divergences in Wheat futures. At A, price pushes to a new high, but the MACD fast line is clearly lower. At point B, prices

drive to a dramatic new low (selling climax?), and the MACD barely moves below the previous low. C sets up yet another example of a divergence, as price makes a new high unaccompanied by a new high on the MACD. Last, D shows a new price low against a higher low on the MACD. In all of these cases, the momentum divergence would suggest that the dominant group in the market may be losing their hold and that a reversal is more likely than continuation. 

Any book on technical analysis will show you that this is the way divergence is supposed to work. There is only one small problem: divergences fail about as often as they work. Figure 7.15 shows a darker side of momentum divergence. At least four momentum divergences set up against this trend, all of which failed. Any extended trend will set up multiple momentum divergences, most of which will fail. This is, in fact, one working definition of a trend: just as a trend breaks support or resistance, trends will also roll over momentum divergences. Traders using these divergences to set up counter-trend trades will incur steady losses, but paying attention to these divergences can be equally damaging to traders who are only trading with the trend if they use them as warning signs to exit their positions prematurely. This does not invalidate the concept of momentum divergence, but it does highlight the importance of having a trading plan that respects

the reality of market action. Clearly define how and when you will use momentum divergences, and, if you will actually set up trades using them, have precise plans for managing those trades. A skilled counter-trend trader could likely have executed shorts on each of the momentum divergences in Figure 7.15 and emerged at the right edge of the chart with only small losses, but this is a difficult way to make a living. This tool, like any other momentum indicator, measures potential divergences between subsequent swings in the market. 

If there are no swings, there is nothing to measure so we cannot have divergence. Many times, markets will set up patterns that resemble valid divergences, but there is no retracement in between the two points being measured. Consider the divergences marked in Figure 7.16. Though the indicator shows the classic patterns of divergence, at the corresponding points on the price bars the market was simply sliding higher. 

A few further refinements will wrap up our consideration of divergences. It is possible to see momentum divergences without referring to the indicator, especially after the trader internalizes many patterns and variations of charts. There are subtle relationships between the length of swings, angle of swings (a visual indication of rate of change, assuming that vertical scaling is consistent), and the specific formations at the turn of

swings that can reveal momentum conditions to the trained eye. The MACD fast line divergence is a crutch, but it is a useful crutch, especially for new and developing traders. Also, divergence is often a temporary and fleeting condition. 

A divergence obviously does not mean the market will never make another move in the same direction, so give some consideration to the timing and duration of divergences, which will vary depending on the specific indicator used to define them. 

With this MACD, a divergence can usually be assumed to have reset or to have fulfilled its expectation when one of three conditions occurs: price returns to an intermediate-term (e.g., 20-period) moving average, the MACD fast line crosses the zero line from the divergent state, or an extended period of time (10+ bars) elapses. If you have executed a countertrend trade based on a divergence, be protective of any open profits once any of these conditions emerges. Last, there are a few logical inconsistencies to consider when trading divergences. 

One of the most important is that since pivot highs are matched to indicator values in an uptrend, we are comparing the highs of the bars to the indicator value; in a downtrend, the lows of the price bars are compared to the indicator value. This problem is that highs and lows of price bars are being compared to an indicator that is calculated from the closes of the price bars; the indicator is completely blind to any information encoded in the extreme points of the bars. There are a number of potential solutions to address this problem. 

A simple one would be to calculate the indicator off the midpoint of the bar (average of high and low of the bar), or off the average of the high, low, and close (this is called the typical price in some analytical systems). As a much more extreme solution, you could calculate separate indicators for bullish and bearish divergence, basing the tool for bearish divergence off the highs of the price bars and the other off the lows. A word of warning is in order, too: if you do something like this, make sure the extra information gained from the tool compensates for the additional complications—simplicity is greatly to be desired.