Option Investor
Trader's Corner

Technical Indicators: MACD, a Last 'Oscillator'

Printer friendly version

Not that the Moving Average Convergence Divergence Indicator, generally known by it's initials 'MACD' and pronounced 'macdee', is last in any sense except that this Trader's Corner is the last of a series of three (LINKS to the earlier two articles at bottom) articles devoted to the 'oscillator' type indicators. The prior dates are the Option Investor Daily e-mails of 2/15 & 2/22.

Each of my three articles, examines the use of oscillators in general (used to determine price momentum as well as suggesting when an Index or stock is 'overbought' or 'oversold'), as well as looking at how each of the 3 most common oscillators are constructed; i.e., their formula.

The other two are the Stochastics Indicator, which means almost always the 'slow' variation or 'Slow Stochastics', and the Relative Strength Index or RSI. [When you see the Stochastics indicator on a chart, it almost always is 'Slow Stochastics'; the name is important only when you have to chose one to put on a chart and there are multiple 'Stochastic' choices.]

Sorry for the boring repetition readers, but in a nutshell Oscillators like MACD, measure how fast is the rate of (price) change, or up/down MOMENTUM. Readings of the RSI and Stochastics 'oscillate' only between zero and (a maximum high) of 100, although neither the Slow Stochastics or Relative Strength Index (RSI) Indicators will ever reach these absolute maximums.

Extreme readings, ranging from below 20 to above 80 in the Stochastics model, or below 30 and above 70 in the RSI, are typically used in default settings in PC charting applications or on web sites that use these studies, to indicate an 'overbought' or 'oversold' condition in stocks and stock indexes.

The Moving Average Convergence Divergence Indicator or MACD for short is the third of the oscillator type technical indicators. This indicator measures overbought/oversold also, but its formula is not set to produce a number between 0 and 100. With MACD we need to look at how low or high this indicator has been relative to past weeks, months and years. The MACD was originally thought (by it's originator) to be especially useful on Weekly charts.
More on the MACD further on ...

But first, a brief update on the current market. And we're still more or less whipping back forth, but the Nasdaq Composite (COMP) has come on stronger this week as it rebounds to play catch up with the S&P, while the narrow blue chip Dow 30 (INDU) has fallen back.

The Dow has been leading the market up, but the question in my mind has been whether this was a "solitary walk of the Dow" as its sometimes been called. It's been unusual for the Dow 30 Industrials (INDU) to lead the market higher for any prolonged periods, especially in the last decade or so.

Sure enough INDU this week has corrected after its run up to a new high and a lower (up) swing high was established on the first rally attempt. If INDU holds the 11050 area of its early-Jan high, as it did today by the close, the Average is maintaining a 'minimally' bullish chart; prior resistance 'becoming' support.

However, to re-establish its prior upside momentum, the Dow would have to pierce its down trendline at around 11100, as seen on the HOURLY chart below.

The S&P 100 (OEX) presents a mixed picture. On one hand, the rally from early-Feb is bullish on balance until this last price dip when the chart picture suggests to bulls to be on alert to OEX building another top.

The Index has to close above 588-589 to gain some upside traction again. A dip below 580 suggests that we're again at the top of OEX's trading range. Stay tuned on that! Looks like a top on the HOURLY chart below, but sellers aren't pushing it.

Calls were good bought in low-570 area of course but an exit was warranted after multiple tops were seen at 586-587. I haven't been convinced to buy puts. There is just not enough bearish news coming out to push the S&P down for long it seems and only mixed bearish signals technically.

The Nasdaq Composite (COMP) managed to clear today its earlier highs around 2315. The January peak just over 2330 is not too far away and an uptrend is intact. COMP is still bullish in its technical pattern as long as the uptrend seen on the HOURLY chart below remains intact and is not pierced.

COMP support around 2285 is key. The prior high in the 2332 area is pivotal on the upside. A close or two above 2330 is needed to suggest a bullish breakout. Volumes look too low to accomplish this currently. Buyers arent that aggressive, just doing some nibbling type buying after how much some of the key Nas stocks fell from their recent peaks. Many of them were OVERBOUGHT!

The Nas 100 (NDX) rallies have been lackluster as can be seen by on the HOURLY chart below, but the primary technical consideration is that hourly lows have established a bullish UP trendline; an uptrend is defined not only as a series of higher highs, but a series of lower pullback lows as well. An index can always take out the prior peak sometime down the road. But if the pullbacks start going to lower lows its time to be in puts on rallies.

NDX has to clear 1720 then establish some support on pullbacks to this same area to suggest that the Index was in a position to challenge its early January top around 1760. I'm not looking for this anytime real soon.

I always try to emphasize that the concept of 'overbought' and 'oversold' that you see associated with Stochastics, RSI and MACD indicates a POTENTIAL vulnerability (to a reversal) only. A stock or an Index can stay oversold or overbought for a relatively long period in a strongly trending market.

It is also true that RAPID and STEEP advances or declines are usually not sustainable for a long period. The market will correct at some point after a steep rise or fall, but WHEN is an open question.

This point is important because there is a tendency to think that the first time or two that an oscillator reaches an extreme reading, whether using an hourly, daily or a longer-term weekly chart, it might be time to EXIT that stock or Index.

However, we see many instances where the trend takes prices significantly higher or lower before there is a correction. Moreover, a 'correction' may turn out to be a SIDEWAYS move ONLY or a lateral consolidation. And, which has been going on for weeks, broadly speaking, in the current market.

Oscillators 'work' best in terms of timing and buying dips in a downswing and selling rallies in an upswing in a trading range market of two-sided price swings, rather than one trending strongly in one direction.

The MACD indicator is calculated by taking the difference between two 'exponentially smoothed' moving averages of closing prices of 12 and 26-day duration; usually these are the only values ever used, although I would note that Gerald Appel, the market technician who formulated the MACD indicator, suggested that a slightly different set of values to be used as a sell 'signal'.

A 'purist' who relies heavily on this indicator might be concerned about using a variation for this purpose. I haven't yet seen evidence that constructing this slight variation and applying it in declining trends adds significant value.

The MACD line is the difference between the longer average (26) and the shorter (12). A moving average of 9 periods then is calculated of this differential (the result of the subtraction), which is called the "signal line", resulting in a faster moving MACD line. The exponential smoothing technique weights the most recent prices changes more heavily and is therefore quicker to track the latest price changes.

The signal line will be slower because it is a simple moving average of the last 9 values of the differential and is not weighted. There is usually a third line plotted, which is a "histogram" (these are the vertical bars) used to show the difference above and below a midpoint line of the difference between the MACD line and the signal line; this line is included in the 'standard' MACD indicator to better see the points where there is an upside (the bars go above the midpoint) or downside (bars are below the center or zero line) CROSSOVERS.

Most often I dispense with the histogram because it clutters up my charts when the size of the chart window is small - the zero line can be seen anyway.

The chart below is from my book (Essential Technical Analysis). What the chart is doesn't matter, nor price levels, only the Indicator explanatory notes and it's PATTERN for now:

The reason to INCLUDE the histogram is that those bars may present a clearer picture of when the difference of the two moving averages is at the widest and narrowest points however, often it is difficult to see these details given how much is shown on the MACD indicator, which often occupies a small section of a price chart.

Its common as you gain experience, to use indicators with minor modifications such as I describe, but this is an individual thing and relates to how much practice you have or the way in which you use the indicator for example if you use the MACD in order to generate a crossover buy or sell 'signal' only, then the narrowing or widening of the differential is less important.

As the inventor of MACD, Appel thought that the MACD was especially useful on WEEKLY charts. From the 'good old days', here's another chart and explanatory notes of the MACD Indicator.

The use made of MACD is similar to the crossover technique of other moving averages a buy indication or 'signal' is generated when the MACD line crosses above the slower signal line and a sell indication is suggested when the MACD line crosses below the signal line.

This 'crossover' use is similar to the trend-following usefulness of moving averages that define and track the dominant trend, either up or down. I suggest MACD use on either daily or weekly charts but tend to favor its use on longer-term weekly charts as a good measure of the momentum of the primary trend.

Once there is a weekly chart buy or sell crossover signal and I am not in the market in question yet or want to add to my position, I may take the first MACD crossover signal on the daily chart, that is in the same direction as the weekly signal.

The foregoing is a use of the MACD as a trend-following indictor, but we are focused here on its use as an OSCILLATOR and, as such, want to employ it for bullish strategies when the index or stock we are following is oversold and for bearish strategies when the MACD suggests an overbought condition and confirmation is seen in price and volume patterns.

The way this is done is to define "overbought" and "oversold" zones as ABOVE or BELOW the Zero line respectively. There are times when the two MACD lines are above or better, well above, the zero line and have crossover 'signals'. Conversely, there are times that the MACD lines have crossover sell signals below or well below the zero line. These are the crossover buy and sell indications that are the most like the Stochastics indicator.

A couple of good buy and sell crossovers of the Weekly MACD is seen on the weekly S&P 500 chart below. Since mid to late 2004, the MACD hasn't shown SPX have been in a longer term oversold condition, absent a dip under the zero line.

There have been upside crossovers of the weekly MACD which may have reinforced major buy decisions. Some downside crossovers occurred also. 'Timing' options trades on weekly charts is not going to be of much practical use, so we need switch to a daily chart example.

There were a few excellent buy/sell crossovers seen in the daily chart below. There was one sell side (or buy put) signal that was worthwhile. As with other oscillators, like Stochastics and RSI, the MACD is best used not as a 'mechanical' signal but as a reinforcing indicator to what is showing up in the chart pattern; e.g., trendline breaks.

You may note instances above relating to when the MACD did not accompany prices to a new high, or a new low and thereby warning of an upside or downside REVERSAL to come. Bullish or bearish divergences can be seen in the MACD at times; e.g., the bearish price/MACH divergence of late-July/early-August.

Two Wednesdays' back (2/15)I went into the Stochastics Indicator use and its calculation; this is viewable by clicking here.

Last Wednesday, I wrote on the Relative Strength Index or RSI, which has an added usefulness in highlighting the occasional bullish or bearish price/RSI divergence which is a frequent 'divergence' well correlated with trend reversals that may lie ahead. Last Wednesday's (2/22/06) article can be found in your e-mailed Option Investor Newsletter (OIN) of last week; or, can be viewed online at the OIN web site by clicking here.

Please send any technical and Index-related questions for possible use in my next Trader's Corner article to support@optioninvestor.com with 'Leigh Stevens' in the Subject line.

** Good Trading Success! **

Trader's Corner Archives