Muscular Investing

When to Use a Computer and When to Use Your Brain


It’s well-established that computers outperform the best human players in chess, poker, Go, and other skills. The same is true in the game of investing. • We can use this fact to boost our long-term performance. Simple computer formulas outdo financial experts. Their minds and ours are clouded with ‘behavioral biases’ that economists have proved are baked into our human decision-making processes.

• Part 3 of a series. Parts 1 and 2 appeared on May 7 and 9, 2019. Photo by VG Stock Studio/Shutterstock. •

In Parts 1 and 2 of this series, we’ve seen that:

  1. People who are brain-damaged and cannot experience fear make much bigger gains in stock-market investing games than people with normal brains.
  2. Paper-trading (making theoretical buys and sells with no money involved) may not help us overcome our behavioral flaws, since people who are paper-trading take on much more risk than people who are trading serious money, thus changing the outcomes.

We can learn from these two facts to improve our own investing results. The field of “behavoral finance” — which has recently won the Nobel Prize in Economics for some of its pioneers — reveals that our human brains make decisions using a lot of influences other than logic. To overcome our baked-in instincts, which are very often wrong about financial data, we can use computerized models, also known as mechanical investing.

It’s widely known that the fastest computers routinely beat the greatest human champions of chess, poker, Go, and many other competitions where there are clear winners and losers. Individual investing is yet another competitive field where computers make better decisions than our fallible human brains do.

Learning to trust the formula rather than our own opinions

A new study by Nizan Packin of the City University of New York’s Baruch College shows that individual investors have heard about most mutual funds and hedge funds not beating the S&P 500. In her April 2019 white paper, Packin tested 800 volunteers in the following way:

  • Half of the subjects were told, “You decide to invest 15% of your savings in the stock market” — a relatively small amount.
  • The other half were told, “You decide to invest 60% of your savings in the stock market” — a much more serious chunk of one’s assets.
  • Half of each group was then told one of the following statements: “You find a reputable stockbroker who makes investment recommendations,” or “You find a reputable online automated investment adviser who makes investment recommendations.”
  • Each participant was then asked, “How confident are you that you got the best recommendation possible for your investment?”

Regardless of the dollar amount invested, the participants were more confident in the recommendations of the computer than those of the human adviser. Apparently, the repeated finding that index funds return 99% of the gain of any given market, while human advisers seriously underperform due to behavioral biases, has worked its way into the minds of most people.

To test this level of trust under stress, the subjects were then told:

  • “The recommendation regarding the investment did not turn out as successful as you had hoped, going down 30% in value. How likely are you to use the same service again?”

Even after a very disappointing loss, the respondents still had more confidence in the computerized formula than in the human adviser. The participants apparently know that, just as in chess, a computer can compute a winning investment strategy that’s superior to the opinions of even very experienced professionals. Grandmaster Garry Kasparov learned the advantages of computers the hard way in 1997, when he became the first reigning world chess champion to lose a public exhibition series against IBM’s Deep Blue.

Using a model versus risking your money on human experts

It’s not just board games and investment strategies where computer models outperform the best-paid human professionals. A white paper by Wesley Gray — the CEO of Alpha Architect with almost $1 billion under management, and a finance professor at Drexel University’s School of Business — sums up the intellectual pursuits in which computer predictions outdo expert opinions. To name just a few:

  • Which convicts should be paroled?
  • How much brain impairment has a patient suffered?
  • What will be the academic performance of college enrollees?

Gray cites a meta-analysis of 136 published studies that judged the accuracy of computer models against the opinions of human experts. The results may shock you:

  • The experts beat the models in only 6% to 16% of the cases.
  • The models matched or exceeded the experts in at least 84% of the cases.

We can use computer models to make our investment performance better than human advisers can pull off — and far better than the typical amateur investor can manage. When you’re playing against a chess grandmaster, you want to use a computer, not your own novice guesses. And when you’re playing against the best traders in the world, you want a computer formula that has proven itself through bull markets and bear markets alike.

To be sure, computers can’t do everything. There will always be a need for a human to monitor any computerized activity. In her paper, Packin stresses the importance of a second opinion from time to time. She cites the downing of Iran Air Flight 655 by the USS Vincennes in 1988. The ship’s computer had marked the passenger plane as an F-14 fighter, despite the fact that a human observer could have recognized the much larger shape of the civilian aircraft. All 290 passengers on board the flight died.

But if your investment decisions don’t involve split-second, life-and-death decisions — the vast majority of individual investors’ portfolios do not — simple computer formulas have been shown time and again to outperform the opinions of human minds, which are subject to behavioral blind spots.

In the fourth and final part of this series, we’ll see exactly how to apply that fact to our 401(k)s, IRAs, and taxable accounts for greater long-term gains.

Packin’s paper, “Algorithmic Decision-Making,” is available as Social Sciences Research Network document 3361639.

Gray’s study, “Are You Trying Too Hard?”, can be downloaded from SSRN 2481675.

The meta-analysis of 136 studies is by Grove, Zald, Lebow & Nelson, 2000.

• Part 4 appears on May 16, 2019.

With great knowledge comes great responsibility.

—Brian Livingston


Send story ideas to MaxGaines “at”


Brian Livingston
About the author: is a successful dot-com entrepreneur, an award-winning business and financial journalist, and the author of Muscular Portfolios: The Investing Revolution for Superior Returns with Lower Risk. He has more than two decades of experience and is now turning his attention directly on the investment industry. Based in Seattle, Livingston is now the CEO of, the first website to reveal Wall Street's secret buy-and-sell signals, absolutely free. He first learned computer programming on an IBM 360 in 1968 at age 15. Learn More
Subscribe to Muscular Investing to be notified whenever a new post is added to this blog!
comments powered by Disqus