And this is why computer programs that claim they can pick stocks (or, more appropriately, drive trading behavior via software) simply don’t work, and companies hawking said programs are doing it on flimsy pretense and some snake-oil salesmanship. It’s not the code nor logic that’s really the issue — it’s the data. Without lots of historical data to train the algorithmic model, the model is guessing as much as your average investor.
The closest thing to the intelligent and pragmatic grafting of predictive mathematics into investing is quantitative investment, but that’s by no means a magic widget to pick stocks and nullify the stock-trading elements of relationships, trust and gamesmanship. Quant is a supplemental science for talented investment managers, not a replacement.
This area is a particular one of interest for me, as I have several friends who play quite heavily in the stock market. A few have purchased a $5,000 program that gives them green, yellow, and red lights to tell them when to buy and sell stocks, and they tried to pitch it to me. I told them that the company that sells the program is in the business of selling fraudulent software to naive investors, and not at all in the business of actually understanding how trading works. They didn’t listen, and they bought the software after attending a full-day seminar (read: sales pitch to stay-at-home moms and retirees).
As near as I can tell, they haven’t even recouped the cost of the software 18 months later.
6 responses so far ↓
michaelsanford // November 25, 2006 at 11:38 pm
This is the only non-human (and even that’s in question, given the skill of the user) stock chooser anyone should use: it likely has the same accuracy as high-priced software and costs almost nothing.
Jeff Ventura // November 26, 2006 at 12:23 am
michaelsanford — haha. That reminds me of the contest the Wall Street Journal ran a few years ago.
Check it out.
michaelsanford // November 26, 2006 at 1:07 am
Exactly what I had in mind! Some students at Berkley (Princeton?) actually ran a “study” along these lines–comparing a dart board to investor picks: differences were stastically insignificant.
thywordistruth // November 30, 2006 at 9:04 am
Thanks for posting this info. I wish more people with your obvious intellect would take the time to post about such matters more often.
By thw way, the best method for picking stocks is to employ the old fasionshed way: With patient observance!
Lord bless,
Jim
https://thywordistruth.wordpress.com
Snoop // November 30, 2006 at 10:26 pm
Of course you are correct but the irony of these stock picking programs is that if they did in fact work, market forces would embrace them and their advantage would end.
Your blog doesn’t suck, peace.
Jeff Ventura // November 30, 2006 at 10:58 pm
Snoop — exactly. Efficient market theory.
Thanks for your comments. I appreciate them.