In 1948 Warren Weaver, of the Rockefeller Foundation, wrote an article in which he described science “as progressing through successive eras, defined by the three types of problems—simple, uncertain, and complex—that they solved. Simple problems address a few variables that can be reduced to a deterministic formula.” Newton’s laws are examples. “By the late nineteenth century, scientists shifted their attention to problems of uncertainty, such as the motion of gas particles in a jar.” They used probability theory and statistical analysis to predict how large numbers behave in aggregate, “paving the way for advances in thermodynamics, genetics, and information theory.” That left us with the most difficult set of problems, those dealing with complexity. (pp. 9-11)
The financial markets are complex adaptive systems. They cannot be described by deterministic formulas. They don’t lend themselves to statistical prediction. What, then, is an investor or a trader to do?
MIT and Stanford professors Donald Sull and Kathleen M. Eisenhardt offer some suggestions in their forthcoming Simple Rules: How to Thrive in a Complex World (Houghton Mifflin Harcourt, 2015). Their book will inevitably be compared to Daniel Kahneman’s Thinking, Fast and Slow and Malcolm Gladwell’s Blink. But it is more practical than its predecessors, written for people (especially business people) who have to make tough decisions.
Simple rules often work best. “In contrast to complicated models, simple rules focus on only the most critical variables. By ignoring peripheral factors and tenuous correlations, rules of thumb eliminate a great deal of noise. The absence of noise results in decisions that work reasonably well across a wide range of scenarios, rather than being optimized for a single situation. … In very complex systems, like the stock market or the economy as a whole, where causal relations are poorly understood and shift over time, the risks of overfitting past data are particularly acute. Statisticians have found that complicated models consistently fail to outperform simple ones in forecasting economic trends, and the accuracy of their predictions has not improved over time. When it comes to modeling complex systems, sophisticated does not equal effective.” (pp. 34-35)
Simple rules “are particularly effective when the situation is in flux, flexibility trumps consistency, and the benefits of seizing opportunities exceed the cost of making mistakes. “ (p. 44)
Admittedly, there are situations in which complicated decision-models work better than simple rules. For instance, “decisions that can be made by computers, such as via automated trading programs, are better candidates for complicated models than those that rely on human willpower to implement.” (p. 37) In general, however, simplicity wins the day.
Effective simple rules can be sorted into six broad categories: boundary, prioritizing, stopping, how-to, coordination, and timing. The authors give examples of each type of rule, drawing on a range of behaviors (from which house to rob to when to sell a stock, from how to deal with out-of-control forest fires to how starlings flock).
The authors describe ways to develop simple rules in business, non-profit, and personal settings. These rules, of course, cannot be created in a vacuum. “Investing the time upfront to clarify what will move the needles dramatically increases the odds that simple rules will be applied where they can have the greatest impact.” (p. 144)
Investors and traders who want to simplify their overly complex systems or who want to create an efficient rule-based system or plan will be well served by this book. The task will remain difficult (or not, if they opt to follow the 1/N rule). In any event, the recommendations in Simple Rules should keep the investor or trader from straying too far off course.