Hi Guys,
In a 1997 6-game chess challenge, IBM’s Deep Blue computer program beat Grand Master Garry Kasparov in a tipping point match between man and machine. The machine won. Was this the harbinger of a machine takeover, especially of the decision making process? Maybe, maybe not.
After all, man might just use these programs to supplement his decision making process. Kasparov, armed with access to the Deep Blue program, should logically whip Kasparov without that tool. I suppose that would be true if Kasparov slavishly accepted Deep Blue’s analysis. However, being human, Kasparov would likely succumb to compromising behavioral biases and reject the computer’s advice under some circumstances.
Today, mathematical algorithms are outperforming human decisions in several disciplines. That’s a little surprising given that these same algorithms are formulated by curve fitting decision factors used by the same humans that they subsequently defeat. Decision consistency is not a strong human attribute.
Phil Tetlock, of Hedgehogs and Foxes fame, has been conducting a
5-year Intelligence Advanced Research Projects Activity (IARPA) forecasting study for over two years now. In a second stage, he formed
5 teams of 12 men each from his best expert forecasters from a first stage test series. A fundamental commitment here is a belief in the wisdom of informed crowds.
His research unit also developed algorithms using factors that his experts identified as influencing their decision making and forecasts. At present, these algorithms are slightly outperforming the expert teams. The study continues. Here is a Link to a Tetlock 2012 video interview on his IARPA work titled “How to Win at Forecasting”:
http://edge.org/videos/year/2012The text from this Edge interview is also available.
In Daniel Kahneman’s “Thinking Fast and Slow” book, he references a huge body of studies that continue to demonstrate that algorithms outdistance expert human judgments in a diverse group of disciplines like medicine and investing.
In medicine, machine diagnosis, on average, are more accurate than on-site doctors. In investing, hedge fund manager Ray Dalio admits that when he and a computer analysis that he trusts make a disparate forecast in an investment decision, the computer program is right two-thirds of the time. Why is the machine superior?
Well, humans get tired and focus drifts. We are subject to household and health problems. We are influenced by current events and the medias selective hype on these events. We are also guilty of overconfidence in our own skills. We trade too often. Several studies have shown that equity accounts that are more or less dormant for years outdistance others that are frequently traded.
Investors also abandon our well crafted rules and policies when the going gets tough. That’s true of both individual investors and professional money managers. Our market timing decisions are often a catastrophe. We are inconsistent in choosing and navigating our selected stars.
John Hussman is a prime recent example of a mostly successful wizard who failed to continue his planned march without introducing a disastrous route change. He added historical data sets to make his model forcibly agree with his predetermined decision; the facts be damned in this instance. That’s a classic, first-order sin when doing forecasting work.
What to do? John Bogle has it right: just stay the course. Don’t junk your rules and process just because an outcome has been a disappointment. Nobody is ever 100% correct with the possible exception of MFO’s Ted (am I having fun at Ted’s expense?).
When change is justified as the investment world constantly evolves, change slowly and incrementally. That’s applying Kahneman’s System Two deliberate slow thinking advantage. When investing, our System One fast response heritage gets us into deep water far too often.
It requires near perfect timing to be either all In or all Out of the marketplace for periods, and still recover long term market returns. Not many of us are successful at this process. We vacillate. This is not the coward’s approach, but it is the prudent approach.
Your comments are solicited. Thank you.
Best Wishes for your continuing investment success.
