This article originally appeared in the Ask Jack column on Bidnessetc.com. Each article answered a question submitted by readers.Do you think one day computers will be able to routinely beat even the best human traders in the market? I see this as a real possibility, since research has shown that the human mind is irrational (e.g., making bad choices in risk/reward decisions). As we know, a computer beat world chess champion, Gary Kasparov, in 1997, and computers have become much more powerful since then. Given current technology and the growth curve in computing power, is it possible that in the future, there will be no humans involved in this competitive auction market?

Will human traders become obsolete? That is a very interesting question. Before attempting an answer, let me make one minor modification in your question changing “one day” to the “in the next 50 years.” The growth rate in computing power is so extraordinary that it would be folly to rule out computers making humans obsolete in any decision-based endeavor within an open-ended “one day” timeframe. I think if I were to answer the original question within the “one day” context, I would be tempted to respond, “yes,” and leave it at that. Not much of an article. Answering the question for the next 50 years instead still leaves the reply totally relevant, while making the answer more interesting.

First, we need to begin by acknowledging that computerized trading approaches already compete with the very best traders. The Market Wizards I have interviewed have included both discretionary traders and traders who have used computerized approaches. It would seem that with computerized approaches already highly competitive with great discretionary traders, this observation coupled with the astounding growth rate in computing power suggests that computerized trading will increasingly dominate human traders over time. And this may well be true. But the key question is: Will there still be a place for human traders to effectively compete in the markets? Or, will trading be like chess in that, at some juncture, computers surpass humans and continually widen their advantage over time so that there is no longer any contest between the two?

It is generally acknowledged that, at present, the most powerful computer chess programs (now referred to as “chess engines”) can decisively beat the best grandmasters in chess. If computers were allowed to compete in chess tournaments, one could easily envision computers driving out human competitors entirely. But, of course, computers are not permitted to compete in chess tournaments, which is fortunate for human chess players. However, human traders have no such protection. Computers can and do participate in trading and increasingly so over time. Why then not conclude that the days of human traders are growing increasingly limited? Ironically, it is human nature itself—the very irrationality that you allude to in your question—that may prevent any imminent domination of computers over humans in the world of trading.

There is a big difference between using computers to play chess and using computers to trade the markets. Before explaining this distinction, it is useful to compare predictions made by astronomers versus those made by economists. If a large meteor passes the earth with a close miss, astronomers can predict very precisely when and how close the meteor will be to earth on its next pass. Even though slight variations in the meteor’s path can mean the difference between another miss on the next orbit and a catastrophic impact, astronomers can make their predictions with virtual certainty. In contrast, economists are often unable to beat a flip of the coin in their forecasts. The difference between the predictive skills of astronomers versus economists has nothing to do with relative intelligence. If all the world’s astronomers had majored in economics and all the world’s economists had majored in astronomy, the situation would be exactly the same.

There is a clear explanation for the great dichotomy between the accuracy of astronomical predictions versus economic predictions: The laws of physics are precisely defined and unchanging, while the relationship between economic variables, at best, can be only roughly approximated and, even worse, continually change over time. Part of the problem in creating accurate economic forecasting models is the great multitude of interrelated factors (interest rate levels, currency values, economic growth variables, employment statistics, government policies, etc.) made dramatically more complex by the interrelationship of these same factors in multiple countries in a globalized world economy. The much larger problem, however, is the effect of human behavior. Fluctuations in pessimism versus optimism can drastically alter decisions by both consumers and companies. Factors such as the propensity to save versus spend and the willingness of companies to hire new employees can vary drastically based on shifting psychology. At the extremes, shifts in human behavior can result in bubble economies and depressions. The fact that the vagaries of human nature make it impossible to derive precise equations to describe economic behavior makes the task of achieving accurate economic forecasts exceedingly more complex than making predictions based on the laws of physics.

The difference between forecasts in astronomy and economics is entirely analogous to the distinction between using computers to play chess versus using computers to trade the markets. The common denominator is that, in each case, one of the two items being compared is based on human behavior and thus far less predictable. The rules of chess are very precisely defined and don’t change. In contrast, there are no static rules governing market behavior because of the role that human behavior plays in determining market prices. Even if all the relevant factors could be precisely defined, the exact same set of factors could result in wildly different outcomes depending on prevailing human emotions. As just one example, during the Internet stock bubble and its aftermath, the valuation the market placed on these stocks increased sevenfold (based on the index) in a mere 17 months and then decreased by an equivalent amount in the next 18 months, all of this price action occurring in a virtual absence of any major fundamental changes in these companies in either the bull or bear phases. Instead, these enormous price fluctuations were almost exclusively the consequence of human behavior, which swung from euphoria to extreme pessimism.

It is the complexity and unpredictability of human behavior that will limit the degree to which computerized trading can improve and that limitation will leave a role for the human trader. Determining the correct trade is not a problem that can be solved in the way that determining the correct chess move can be, regardless of how much computing power is available – at least not now, and probably not in our lifetimes. Computerized approaches can achieve strong market edges, but so can talented discretionary traders.

Submit your own trading or investment question:

[email protected]

Information on FundSeeder

http://fundseeder.com