Archive for the ‘. . . . 3 STEPS’ category

June 1, 2020



Bob Karrow discovered a different method of counting waves in 1976

There wasn’t much in the way of wave counting methods in those days other than the Elliott Wave Theory.  For want of a better name I have named my wave count method, Karrow’s Wave Count.

The following is Edson Gould’s verbatim wave counting technique as published in “Findings & Forecasts”.  “Findings & Forecasts” was Edson Gould’s advisory service.  His wave counting is a simple method and relatively good, but prone to mistakes and it  doesn’t explain the detailed breakdown of a wave counting event.  Every bull and bear market is different and the details of the wave count can’t be explained by Edson Gould’s technique.


THE THREE-STEP RULE as written by Edson Gould


“The theory, hypothesis and assumption that the stock market is essentially a reflection of investor psychology has led to the development of a number of principles and theories that over the years have yielded practical results.

The Three Step Rule is one example.  The rule is: “Expect three steps, but be prepared for a fourth. ” It is based on the principle that man, individually and in mass, can usually tolerate only three attempts.  The typical salesman, after ringing three doorbells and getting turned down, will go to the movies.  The typical fighter, after being knocked down three times, stays down.  If the salesman, after three failures, makes his fourth call or if the fighter gets up after the third knockdown, they are likely respectively to become sales manager or champion.

 So it is with the market.  Short-term, intermediate-term and long-term moves usually comprise three well-defined steps.  But sometimes there is a fourth.”



Edson Gould was a wave counter, but it’s debatable to what extent he counted waves???  His real-time wave counting techniques as revealed on his charts were cloudy and unexplained.  If he had a more detailed and precise technique, it was never revealed in his writings (“Findings and Forecasts”). 

I wish Gould had documented his techniques to be published after his death, but it never happened.  

Knowledge has a terrible habit of evaporating after a person’s death.  

I have a lot of Edson Gould’s published reports from “Findings and Forecasts”, 1973 to 1979.  There was a LOT of repetition in his biweekly reports and consequently I didn’t keep everything that he wrote, but I kept everything important.

During the period that I was subscribing to Gould, 1973 to 1979, I discovered a different method of wave counting that was simple and precise, plus the method doesn’t change between an up or down market (bull or bear).  Prior to this period I had tried Elliott Wave counting and found it was OK in hindsight, but terrible in real-time.  The Elliott exclusions and exceptions were so many that one could draw any conclusion that was deemed necessary.  This is the reason why a leading Elliott Wave technician has continually called for a catastrophic ending of the bull market. I have little doubt that he will be eventually correct, but in the meantime he has missed the beginning of a lot of market action.  His method has caused him to react rather than predict the market.

In 1982, he was perfect in calling the beginning of the bull market.  I too was a major bull at the August 1982 bottom and missed the exact bottom by being 2 days early. I had to sweat out the declines of Thursday and Friday before the Monday launch of the bull market.

Needless to say, I discarded Elliott Wave after having high hopes but unreliable results.



After my breakthrough discovery on a new wave counting technique (late 1976) , I found that KARROW’S WAVE COUNT was at odds with Edson Gould’s previous wave counts.  But it was the problems in Gould’s method of wave counting that forced me to invent a different method. For instance, he labeled step 3 of the 1973-1974 bear market as ending in December 1974.  My method determined that the October 1974 bottom was the end of step 3, which ended the 1973-1974 bear market.

The rally from October 1974 was step 1 up in the new bull market that began in Oct 1974.

How did I conceive of the new wave counting technique???

In 1976 using Gould’s method of wave counting, I was left hanging waiting for a 3rd step up that never appeared.  This was a major problem as I was still holding stocks long as the began a deep correction after the 1976 peak.  I was not too happy with this result.

If you look at the chart below and begin counting the steps up from December 1974, you will see there are only 2 steps up to its peak in 1976.  I have labeled the steps correctly in the chart below, which shows 3 distinct steps down after January 1973 and then 3 steps up to the 1976 peak.

1970 to 1978 DJ INDUSTRIALS

1970 to 1978 DJ INDUSTRIALS

As a result of my waiting for a step that never came, I researched the steps in the past and arrived at KARROW’S WAVE COUNT.

Gould states: “Expect three steps, but be prepared for a fourth.”

I agree on the 3 steps, but totally disagree on a 4th step.  There is either a 3 step, or a 5 step for a completed rally, or decline.  More on a 5 step later.

The first two large steps up after the 1974 bottom were the recovery precursor to the great bull market that kicked off in step 3, which began in August 1982 and continued until 2000.  Step 1 was October 1974 to 1976, step 2 was 1978 to 1981 and step 3 was 1982 to 2000.

Edson Gould died in 1987 prior to the market crash of October 1987.   It would have been nice if he could have lived to see the bull market to 2000 that he correctly called in 1974.  He was a GREAT stock market technician (the greatest of the 20th century, maybe of all time)



KARROW’S WAVE COUNT consists of 3 steps up and 3 steps down.  3 steps down is a conflicting idea with just about all wave counting methods that exist today.

But first a few rules 



  • Forget about Elliott when looking at my wave counts or you’ll NEVER understand the simplicity of the concept.  Fuhgeddaboudit
  • KARROW’S WAVE COUNT is simple and functions without a maze of exclusions (but it does have ONE exception, unlike Elliott Wave, which has a multitude of exclusions and exceptions and allows wave counts to proceed in many different directions).
  • I like things simple and I am a strict believer in KISS (keep it simple & stupid).


Here is how KARROW’S WAVE COUNT works:

  • Some times it’s as simple as counting 3 rallies (or 3 declines) on a chart . . . other times, not so easy, requiring more scrutiny.
  • In a downtrend the same rules apply except you are counting 3 declines instead of 3 rallies up.
  • To verify the wave counts, we have the channel concept.  Staying within the channel verifies a continuous wave.  This is an important concept and it rules the waves like Captain Bligh.
  • 3 steps must stay confined to a channel. On a long wave, a log chart may be required to see the channel.
  • Laying a straight edge on the chart will help you visualize the channel.
  • As the larger trend progresses, all of the steps (or waves) that make up the overall trend and will be confined to a larger channel.
  • When the market breaks its channel, regardless of the perceived wave count, the step is finished.   Small or large channel, this is true.
  • In a lengthy wave extension (5 steps), breaking the channel may be your best, or only indicator that the wave count has been completed.  This usually will be a product of a long bull market like 1982 to 2000, which was difficult to count in real time.
  • Here comes an exception (I hate exceptions) . . . 1 of the 3 waves can sub-divide into another 3 waves.  I call this an extension.  Extensions are frequently the rule rather than the exception during a bull market.
  • When an extension takes place: (1) the trend is still intact, (2) the overall channel will likely widened and (3) instead of a total of 3 steps, there will be 5 steps.
  • Why 5 steps and not 4 steps???
  • An illustration:  You have 3 steps in a completed wave.  ONE of these 3 steps subdivides into ANOTHER 3 steps.  You now have the two original steps plus the subdividing third step. This makes a total of 5 steps, which breaks down as: 2 original steps + the 3rd original step, which now consists of 3 more steps = 2 steps + 3 more steps = 5 steps).  Clear as mud???  I hope not, but try to visualize it, or use a paper and pencil and draw the illustration of what I’m saying and you’ll get it.

  • Extensions during a bull market can occasionally take a long time to complete the count.  The 1990s was an excellent example of never-ending subdivisions.  Correctly tracking the wave count can become almost impossible during continuous subdivisions (the 1990s).  The breaking of a wave channel will be the only conclusive clue that the wave has terminated.
  • There are always trends within trends from the short term of a few days to the super long term trend that may last centuries.  Bringing it all together gives you a remarkable market perspective.  I love perspective as it tells you the magnitude of an upcoming trend, but forecasted numbers will never be provided for steps to end their move.  It’s just over when it’s over.  In other words, it takes a visible reversal of trend to say a step is finished.
  • The correction following the second step is usually larger/longer than the correction that followed the first step.
  • Frequently I will use the terms “step” and “wave” interchangeably.  But I usually think of (1) a step as shorter in duration, (2) a wave consists of steps and is longer term than a step; such as there are 3 steps to a wave.
  • The steps/waves fit inside one another like a matryoshka (Russian nesting doll).
  • At the conclusion of a 3 step trend, there will be a corrective counter move.  This counter move MUST be of sufficient amplitude (or duration) to BREAK THE CHANNEL of the prior 3 steps.
  • Something to keep in mind.  If a channel break doesn’t occur after the corrective counter move is finished, you’ve probably miscounted the steps and the previous trend is probably still intact.
  • Breaking the channel is a very important concept.  If a channel wasn’t broken, it can be the only clue that your wave count was incorrect.
  • Here comes another exception (damn another one, I hate these things) . . . In other blog updates like “Death of a Gunslinger” , you can read how things really went wrong for me in 1987. The reason is in this exception. Up until 1987, every step resulted in a lower low or a higher high. In 1987, the countable steps didn’t always result in a lower low. This naturally threw my count off.  In this exception, every downward impulse should have been counted as a step.  So beware the possibility that a downward/upward impulse is sometimes good enough to warrant a step count even if it doesn’t result in a lower/higher low/high.  If I had used this principle in 1987, the results would have been extremely different, but I wouldn’t have learned the lesson about using too much leverage.  That unlearned lesson would likely have killed me at some later date and the money lost undoubtedly would have been far greater.  Why???  Because “Death of a Gunslinger” wouldn’t have taken place in 1987 and my string of unbroken profits would have likely carried on for some time and I would still eventually learn the lesson of leverage being a double edged blade.
  • Nobody ever talks about their losses, alway there profits.  If they don’t admit to being taken out to the woodshed and having the crap beat out of them . . . they are liars.  It’s a pillar of truth from Wall St that newbies always get beat up until they learn how to manage their risks. A newbie would be anyone that hasn’t been to the woodshed.  It took me 1955 to 1987 to stop being a newbie, 32 years (almost a Fibonacci number).


  • Below is a chart of a subdivided wave with a total count of 5 steps.
  • In this chart, the real uptrend ended at the conclusion of the 3rd step.  Steps 4 and 5 were part of a topping formation (sawtooth).  Steps 4 and 5 don’t have to be part of a topping formation, it just happened to be true in this example.  Occasionally 4 and 5 can be part of a blow-off phase.  Notice the use of the Pitchfork in denoting the channel in the chart below.  Sometimes pitchforks are useful in denoting the channel, but they don’t always work properly.




%d bloggers like this: