OK, torturing the data sounds better but everyone is a spin doctor.
In this morning's Quick Takes Pro (why have you still not taken a free trial?) I wrote about the S&P 500 reaching, or at least getting close enough to an important Fibonacci retracement level. Let's not debate the efficacy of Fibonacci here and stipulate that there are some attractions in the major retracement levels.
Anyway, we've all seen charts with a spaghetti pattern of Fibo retracements, Fibo time projections, Gann Fans, speed lines and possibly all of them together. It is like trying to see the price chart behind several pair of fishnet stockings (you pick your own metaphor). The more lines you dump on, the more likely you will find a line or combination of lines that hits a major turning point.
I remember doing a presentation to Knight-Ridder corporate folks when I was a newbie analyst working for Knight-Ridder Financial. Basically, I cherry picked the right length of data that made the trend of Knight-Ridder stock point up. After the dog and pony show ended, the boss of KR Financial leaned over to me and said appreciatively, "I don't know how you made the stock look good." After all, for the past year or maybe two the stock went absolutely nowhere.
"Lies, damn lies and statistics" - (possibly Benjamin Disraeli but there is debate)
This is why we agonize over log vs. linear. Over where to start drawing a trendline. And over what exactly it means to be "overbought." Indicators can let you have any market opinion you want.
I cope by trying to figure out what circles, lines and arrows actually describe the action as it is right now. When some disagree I believe neither. But I do toss them onto their own sides of the clip board along with everything else I use to see if one side - bull or bear - has a significantly longer list. If it does, then there is a decision to be made. And if there is not, well, welcome to today's market. Pick the side with the small advantage, use good risk control and off you go.
No comments:
Post a Comment