Apr 2006
This is the second article in a series. Click the following for part I, What is Technical Analysis.
Now that we have a little more clarity, let's move forward. It's time to ask the question, "What is wrong with Technical Analysis?" And by that I mean, what is wrong with TA as it is commonly conceived or generally understood.
I believe the flaws boil down to four things: Popular TA is too confusing; too distracting; too prone to false confirmation; and, last but not least, too easily mistaken for reality. (We'll address the last one in Part III.
TA Problem #1: Too Confusing
Show of hands. How many of you think technical analysis is overly confusing, or at least has the potential to be such? (I know you're reading this, but ask yourself whether you would raise your hand. Chances are you would.
There are so many bells and whistles, it's easy to be overwhelmed. There are literally dozens and dozens of indicators, oscillators, chart patterns and candlestick formations to choose from... and that doesn't even touch the exotic stuff, like Gann Wheels (ahem) or Fibonacci Retracements (cough cough) or Elliott Waves or Camarilla Equations (gack!). When you combine all the possibilities, it's enough to make your head explode... and can quickly lead to ‘analysis paralysis', the dreaded deer in the headlights effect.
I came across a great example of TA recently.This mock phone call was posted on a trader's message board (transcripted from a monk-e-mail) and it does a good job of getting the point across.
Hey yeah, this is Joe Again... Hey you gotta check out this chart on, uh, DDP... uh, it's a triple-top head-and-shoulders-bottom outside-reversal.
Now if the MACD crosses when the stock's on the 50 day, you gotta buy with both hands. But if it breaks the 50 day, you probably don't have support until the 200 day, which we won't know unless we get some kind of a selloff confirmation here.
See the Bollinger bands? Yeah yeah, they're confirming it's a tightening flag pattern, which only further confirms the potential for support at this level.
So, you know, I'm not really sure what you should do with this, I just think it kinda looks cool... oh wait a second, uh, I got the wrong symbol.So, uh, call me on those Yank tickets.
I love the ending. "I'm not really sure what you should do with this, I just think it kinda looks cool..."
To further illustrate the point, Victor Sperandeo tells this sad but amusing tale in Methods of a Wall Street Master:
...in 1974, I was approached by a technical analyst named Leo who told me that he thought I would profit from his knowledge. I hired him on a trial basis for $125 a week and planned to pay him a percentage on his recommendations that paid out well. Anyway, Leo was in the office at 6:30 A.M. every day, doing over 120 charts in living color, using methods I still don't understand. He worked 16-hour days, and he obviously knew a lot about the markets, maybe too much for his own good.
When I asked him for recommendations, he would show me the charts and say things like, "This stock might be forming a bottom," or, "This stock looks like it may fill in the gap." But whatever he said, it was always indefinite and always offset by a confusing array of other possibilities. I would ask, "Okay, but what should I do, buy or sell?" Leo just couldn't give me a simple, straight answer. And what I remember most about him is that the ends of his shirt sleeves were frayed, and that he ate homemade tuna fish sandwiches for lunch.
TA Problem #2: Too Distracting
Technical Analysis can not only be confusing, it can be highly distracting.
One of the funniest movies of all time (in my opinion) is Monty Python & The Holy Grail. Like King Arthur, many traders have their own quest for a Holy Grail -- that magic formula that unlocks the door to vast trading riches.
In line with this, there is a phenomenon I call "grailitis" -- an unhealthy obsession with one particular factor or element of trading, lifted irrationally above all others in terms of importance, and perceived as the ultimate key to financial success. (For those inflicted with grailitis, the market tends to play the role of taunting Frenchman. Now go away or I shall taunt you a second time!)
Grailitis can take many different forms. All that's needed is an irrational overemphasis on a single area, to the exclusion of other areas, with the fervent belief that this ‘one thing' will somehow lead to the promised land. Technicians aren't the only ones prone to it either; I've seen guys make holy grail quests out of fundamental analysis, macro analysis, personal psychology, simplistic market platitudes, and even money management principles.
It just so happens that, because the possibilities of TA are so broad and wide ranging (and the bells and whistles so mesmerizing), indicator grailitis has become one of the most common trader afflictions.It's all too easy to think there is some magical shortcut, some secret combination of oscillators and patterns, that will make the dollars roll in... if you can just tweak it to perfection.
Like poor Leo with his frayed shirt sleeves and tuna fish sandwiches, devoted technicians can spend their entire market lives digging for that elusive grail in the lines and squiggles... all the while ignoring the true building blocks of success.
To make matters worse, the distractions of needless TA complexity are often self-perpetuating. There are so many stocks to choose from, for example, the pure technicians' starting point is often too broad to be whittled down by chart and volume patterns alone.The ‘simple' formations turn up dozens, or sometimes even hundreds, of candidates for whatever screen is run.Far too many!
So the technician resorts to more complicated oscillator and indicator combinations -- invasion of the bells and whistles -- to whittle down the selection pool.
Then, when final candidates from, say four different areas pop up -- biotech, semiconductors, homebuilders and money center banks perhaps --the technician often winds up subconsciously picking the industry he or she ‘likes', or else chooses one at random without considering the deeper merits of the situation.
Two problems are thus created: needlessly distracting complexity is layered onto the decision making process, and hidden selection bias has crept in through the back door.
TA Problem #3: Too Prone to False Confirmation
In the corporate world, most everyone is familiar with the "Yes Man," who always tells the boss what he wants to hear. The Yes Man is also referred to as Apple Polisher, Boot Licker, Suckup, and a few other less polite terms.
A corporate boss who surrounds himself with yes-men has created a bad situation.When the boss thinks he is getting useful feedback, in reality he is only getting echoes.
Boss: Jenkins, I'm thinking of replacing our entire customer service division with an automated phone system. I've found a company who says they can install it for $1500. What's your opinion?
Jenkins: Oh yes sir, excellent idea sir.
See how this might lead to a problem? In terms of TA, many popular indicators do the same thing as Jenkins. They are the "Yes Men" of the charting world, providing an echo rather than a validation.
If you see a breakout on a chart, it is tempting to look to an oscillator for confirmation -- and then get excited when you see that the RSI or the MACD or the Percent R or whatever is breaking out too.
But is it truly a confirmation? Or just an echo?
If a bar chart breakout is based on graphical representation of price behavior, and an oscillator breakout is based on graphical representation of that same behavior, are they not two different flavors of the same thing? There is a difference between new information and repackaged information.
Remember that a typical chart tracks two variables -- price and volume. That's two variables. Three if you count time. So if you decide to load up your chart with ten different indicators and oscillators and overlays, how many variables are you working with now? When you think about it, still just two or three! Most (if not all) of that stuff is based on price and volume over time... just repackaged in various colorful ways.
In some ways, the pitfalls of TA are similar to the pitfalls of statistics. With a little practice, it's easy to work the numbers -- or work the indicators -- until you see what you want to see.
Next in the series: the Chart is not Reality.
