pattern recognition

During my brief peeks at the blogosphere while on assignment the last few days, I saw that a theological battle had broken out.  Or, should I say, broken out again.

The thread jumped from Phil Pearlman to Josh Brown to J.C. Parets and beyond.  To get the feel of it, read the blogs and all the comments too, but it amounted to a squabble among the major securities analysis religions, to borrow a phrase that Brown had used just a few weeks ago.

The technical analysts point fingers at the fundamentalists, the fundies point back, while the academics and indexers look askance at everyone.

Here’s the deal:  Tell me what you do.  Tell me why you do it and why you expect it to be successful.  Show me how you do it.  Let me see the results and pick away at them.  And give me a chance to poke holes in what you do.

I don’t care what you believe.  But you have to convince me that it’s something more than faith and something greater than luck.  No matter what tribe you come from.

If I ask a portfolio manager what the role of technical analysis is in his decision making and the answer is “nothing,” then I might wonder why he has a stock chart on his screen when I walk into his office.  The mere presence of that image triggers pattern recognition and we are easily anchored by the smallest of things.  Similarly, the technical analyst who looks out his window and sees the trucks streaming by on their way to North Dakota may be influenced in unexpected ways when interpreting the stock of an E&P company that’s active there.

And so, say quants and efficient-marketers, we just do the numbers.  But there are patterns there too, often more powerful because of the allure of “proof” that comes with numbers all tied up with statistical bows.

We are in the pattern recognition business, one way or another.  The more we understand the benefits and detriments of that inescapable tendency, the better off we are.