Momentum Precedes Price #2: Momentum Divergences

Mar 11, 2007: 4:03 PM CST

Momentum can also be a leading indicator when divergences between price and momentum arise and can lead to some profitable, high probability trades.

Please understand that the set-ups I am about to discuss are countertrend tactics, and as such, you must employ a hard stop in the event that the trend reasserts itself and you are on the wrong side. Contrast this tactic with the principle: “Trends have a higher probability of continuation than reversal.”

When you play for a momentum divergence trade, you are always playing for a small target and playing for a possible shift in buying/selling pressure. You can find various other sites that describe the concept of divergences with various indicators, and before attempting any such trade, I suggest researching further on this potentially profitable topic.

Some of the most popular indicators for uncovering price divergences include the MACD, stochastic, RSI, Ultimate Oscillator, momentum, rate of change, price oscillator, etc. You have to discover which indicator works best for you. Indicators are used as ‘training wheels’ until you can develop an intuitive sense of determining where the buying and selling pressure (momentum of the move) are diverging with the price action. This process takes time, yet indicators can help highlight these conditions. There is no perfect indicator to do this. I am using a fast MACD oscillator in my chart examples. I also play divergences in the RSI oscillator.

In this sense, momentum precedes price in that a slowing of momentum indicates that a possible change in price is yet to come. Do not get caught in the trap of searching for momentum divergences all over the chart. Examine them at the (possible) end of mature trends for greater probability. Again, we are not seeking the end of a trend move (reversal), but just a retest and a small target. In fact, we are playing for a simple retracement swing against the direction of the prevailing trend. This illustration may help:

The top pane always shows price. In this case, we are in a mature uptrend and price is continuing higher. A situation develops where the buyers are becoming less aggressive in their momentum (force of buying pressure) and momentum is declining while price is not.

Of importance to note (and the reason behind the divergence in the oscillator) is also price based. Note the steep rise of the previous swing up (creating heightened oscillator/indicator readings) and then the more gradual rise of the second swing up (creating a lower peak in the mathematical oscillator). This sets up the divergence while the reason for it is declining momentum.

If momentum precedes price, then in this case, a decline in momentum forecasts a decline in price as the most probable swing play. If buyers are less aggressive to raise their offers, then it won’t take much effort for price to fall and those who own the stock will begin to sell.

Here are some charts which highlight momentum divergences:

Ebay 2006 Daily Chart

GOOG 15m Divergence

Divergences are difficult to quantify for a mechanical system, so this is one area discretionary traders may have an edge over programmers.

Chart #1: EBAY Daily (end of 2006). With momentum (buying pressure) decreasing, a countertrend divergence trade sets up to test the most recent swing low. Target achieved.

Chart #2: EBAY 60 min chart (October 2006). Buying pressure (momentum) is declining and we can play for a small target with a tight stop.

Chart #3: GOOG 15 min chart (March 2-6, 2007). After declining for a few days, a four-point touch divergence develops in the oscillator and a flatline base forms (this is where experience over oscillators triumphs – the decline in selling momentum is best picked up by the oscillator, yet the basing area is easy to spot on the chart). Even though price continued above our target, the divergence play is only good for a small target. Note the new momentum high and reaction against it on this 15 minute chart.

I did want to highlight another point through the use of various time-frames. Divergences and momentum concepts are valid across all timeframes.

There are a few caveats to be aware when identifying momentum divergence plays:

  • Momentum divergences are invalidated (and nonexistent) in rangebound, consolidating markets
  • Only look for momentum divergences in the context of a mature trend (however short the time-frame)
  • Momentum divergences work best after a “three-impulse” pattern in a trend
  • Momentum divergences can be used in conjunction with Bollinger Bands or Keltner Channels (for increased probability)
  • Momentum divergences are to be played for a SMALL target (price correction) and NOT for a reversal in trend direction
  • The best divergences resemble “double-top” or “double-bottom” chart patterns
  • Keep a tight stop in the market close to entry in the event that the strong trend reasserts itself and causes great losses.
  • Exit divergence trades which do not resolve within 15 bars (create a time stop parameter)

Trading momentum divergences is a complex strategy and should only be attempted after repeated exposure and internalization of the price behavior that sets up the pattern.

12 Comments

12 Responses to “Momentum Precedes Price #2: Momentum Divergences”

  1. jake Says:

    hi corey im a newbie and appreciate all the helpfull advice. im a bit confused regarding the oscilator, would you be able to explain what kind of oscilator your using and how it works? thanks.

  2. Corey Says:

    Sure, Jake. Thank you for the comment.

    While divergences can be measured using standard settings for Rate of Change (ROC), RSI (one of my favorites), momentum, MACD (Moving Average Convergence-Divergence), and even Stochastics, I am using a modified MACD indicator. The default setting for the MACD is a 26 period exponential moving average subtracted from a 12 period moving average, with the signal line being a 9 period smoothing (average) of the MACD line (difference).

    I am using an indicator created by LBR Group which is simply a modified MACD that subtracts a 10 period EMA from a 3 period EMA and plots the signal line (‘lookback’ or smoothed line) at 16. This is called the “3/10 Oscillator” and helps alert to divergences and momentum readings a little quicker than some of the traditional indicators. It’s almost like a hyper-MACD. It attempts to capture “swings” in price better than other indicators. It’s still a MACD, though, and can be plotted with standard software by modifying your inputs.

    Corey

  3. Afraid to Trade.com Blog - Overcoming Stock Market Fears » Quick Daily Commentary: March 14 Says:

    […] the climax and the divergence, the market created a New Momentum High which is the precursor for the “Impulse Buy” […]

  4. jeff Says:

    Thank you for your fascinating blog! It looks like you are using the MACD histogram (rather than the MACD line itself) to detect divergences. Is that so?

  5. Corey Rosenbloom Says:

    Hey Jeff,

    Thank you for the comment and question.

    Actually, no, the indicator is rather unique. You can achieve a similar indicator for free at StockCharts.com by typing in the standard MACD indicator and changing the values to 3, 10, 16 (in the three boxes from left to right).

    I have customized the 3/10 Oscillator so that it does not show the histogram on my panel. What you see is actually the identical reading of the 3/10 line (outer white line) and I have programmed in a duplicate line that I have set as a histogram, but it is only for ‘show’ and to be pleasing to the eye, and is absolutely no different that the ‘swinging’ white line. You will see that this is not the case on StockCharts.com, as it uses the actual histogram, which I do not use (though others do).

    I consider a divergence to be when the outer white line on the oscillator makes a lower low when price makes a higher high. I am comparing swing highs to swing highs to look for confirmation/non-confirmation in the oscillator. The theory is that if momentum is fading, as evidenced by a shorter price swing up than previous ‘up-swings,’ then we can expect at least a nominal correction or a possible reversal in direction.

    With these, it is best to play for a small target only, and not for an actual reversal. Divergences only highlight loss of momentum, and momentum may pick up following a natural correction or retracement, and so it is best to use absolute stops in case a large trend movement is building.

  6. jeff Says:

    THank you very much!

  7. jeff Says:

    Could I ask you another little technical question? what are the white lines that border the highs and lows of the price bars? is this some indicator?

  8. Corey Rosenbloom Says:

    Hey Jeff – No problem at all!

    Ahh. What I have done there is simply plot a one-period simple moving average of the HIGHS of the daily price and the LOWS of the daily price (I plotted two averages).

    That – for me – turns it into a ‘swing’ chart where I’m looking at the price highs and lows and relative ‘swings’ in price instead of the open, close, candlesticks, etc. It’s just a neat way to organize the data.

    I’m looking at swing divergences, and this seems to make them jump out a bit more at me. I use TradeStation.

  9. Anonymous Says:

    1. how far away are the divergence consider a true divergence?

    2. if divergence appears in daily(or weekly) chart but not on the weekly(or daily) chart, is that still a true divergence?

  10. Corey Rosenbloom Says:

    Anon,

    1. Testing shows that a ‘good’ divergence (or a valid divergence) is between swing highs and swing lows with minimal to no price interruption. In other words, the best divergences are clean, observable swings, rather than containing erratic price movement in between. I don’t have a strict definition for ‘how long is too long’ between peaks, but experience will help you in this area as well as seeing multiple examples. I will say that if you enter a possible divergence trade and it does not hit its target within 13 or so bars (weekly, daily, etc), then it’s best to take a time stop and not try to hold the trade any longer. Most divergences work relatively quickly.

    2. A divergence will be valid if there is non-confirmation by momentum (or other oscillator) between swing highs or lows (if price makes a high but momentum makes a lower high).

    Divergences are relative to time frames, and a weekly observed divergence will play a larger role on the price structure than a daily or hourly divergence, but they are all still divergences with the same implications (loss of momentum on the prevailing force).

  11. Spa Dog Shampoo Says:

    If momentum precedes price, then in this case, a decline in momentum forecasts a decline in price as the most probable swing play. If buyers are less aggressive to raise their offers, then it won’t take much effort for price to fall and those who own the stock will begin to sell.
    Thanks for sharing.

  12. Manhindara Says:

    Momentum is good in Forex, it really make us run away with the situation but also a wrong direction of momentum is pretty bad as sometimes it gets too tough to stop the trouble. I mostly want to gain good profits with keeping momentum but at the same time I like to have safety and that’s why I use rebate service, it is provided by one of the finest broker OctaFX, so we can earn up to 15 USD per lot size trade.