By SANDER GERBER
Can artificial intelligence replace human intelligence in investing? Some people are betting yes, believing that quantitative analysis outperforms human analysis — that crunching numbers, machine learning, algorithms, backtesting, prediction models and other forms of computational wizardry will inevitably obsolete the need for human judgment when assessing risks and rewards in investing.
But step outside of investing for a moment and consider attempts to “robotize” other endeavors that call for the sifting of indicators and the weighing of human decisions. Perhaps you will agree with me that, while big data is useful for spotting patterns, people still play a vital role in interpreting information.
A classic example came in 2009, when Capt. Chesley “Sully” Sullenberger overrode landing options (calculated by air traffic controllers) at nearby airports and, based on his judgment and experience, decided that the best choice was to bring down his disabled US Airways plane in New York City’s Hudson River. Sullenberger’s skillful touchdown saved all of the 155 passengers and crew on board.
Sullenberger has said: “Not every situation can be foreseen or anticipated. There isn’t a checklist for everything.” His visual assessment of the predicament, his intuition, his vast experience as a fighter pilot and commercial airline captain, and his “feel” for what his best odds of success were — in short, his human judgment — is what produced a water landing dubbed the “Miracle on the Hudson.”
Take baseball, football, basketball or any other professional sport. Sure, some teams use statistical analyses like sabermetrics to identify and recruit players whose performance in certain skill categories indicate their true value. But those teams also use the human judgments of scouts and team executives to assay the less tangible assets, like a player’s leadership ability and level of desire.
I’m a sailor. In addition to using GPS and navigation instruments, I also manually make steering and other adjustments depending on how I read the wind and currents.
Statistical analysis is useful for charting the dots. But human judgment — at its best — connects the dots.
Legendary investors like Warren Buffett of Berkshire Hathaway and Peter Lynch of Fidelity combined close inspection of a business’s numbers with close inspection of human factors like management talents and marketing potential.
In investing, valuations are based on an uncertain future, and the decisions that are made in that future. Perhaps, in a stable, low-risk environment marked by little change, a computer can mine information to ascribe relative values better than a human, and quickly trade them for a profit.
Don’t go too long on artificial intelligence. Don’t go too short on human intelligence. The two are complementary.
However, there are limits to the intentionality of a computer — its ability to correctly discern information, to predict future management decisions and corporate actions, competitive market dynamics or consumer behavior changes, and other dynamic factors. Perhaps, in stable environments, the computer might appear to be making optimal decisions. However, in a high-risk environment, human judgment clearly makes an important contribution.
If you believe that investing in various markets and financial instruments will proceed in a stable environment in the foreseeable future, then lean on a heavily quantitative approach.
But if you agree with me that the only economic, social and political certainty going forward will be change and even chaos, then I suggest you hedge your bets.
Go with a combination of computational fact-gathering and interpretational human judgment to find the form of money management that produces for you the diversity, consistency and performance that you seek.
Don’t go too long on artificial intelligence. Don’t go too short on human intelligence. The two are complementary. –AN
