“When clouds appear, wise men put on their cloaks” writes Shakespeare in King Richard III (http://goo.gl/PAFO9).
Such a rational response to information is one of the tenets on which economic analysis is built. However, as we know, modern technology has enabled us to predict with a high degree of confidence that clouds will appear, letting us foretell when cloaks might be needed.
It is just this kind of foresight that has been the subject of an increasing number of studies in recent years. The additional information that may enable us to do so might be provided by what people are trying to learn. These days, the most common way to accomplish our thirst for information is using search engines.
For a number of years now, Google has provided a service, the “Google Trends tool” (http://goo.gl/rd1il), allowing anybody to find out how many times certain search terms have been used (if, that is, the search volume is large enough).
Hal R. Varian, professor emeritus at Berkeley and currently the chief economist at Google, has been presenting studies of the power of this knowledge around the world. One of these engaging “Predicting the Present” talks can be watched here: http://goo.gl/2m0Gc. A research paper published last year by Choi and Varian (available at: http://goo.gl/yikrQ) demonstrates how the data can be utilized “to forecast near-term values of economic indicators,” including unemployment statistics, car sales, vacation trends and the confidence of consumers.
As those who have attended presentations of the underlying idea can tell, the results are usually astonishing enough to elicit “oohs and aahs” from even initially skeptical crowds.
More mind-boggling evidence of the power of crowd-search behavior was published this week.
Three professors in the U.S. and England presented their results on “Quantifying Trading Behavior in Financial Markets Using Google Trends” (the very readable paper is publicly available at: http://goo.gl/J8f2K).
In essence, the team identified certain search terms, such as debt, inflation or bonds (they list many more) and used hypothetical trading strategies based on the frequency of those searches in the U.S. They were able to discern “early warning signs” based on how those search patterns changed.
In case you are skeptical, here is a tidbit of information. An investment strategy (taking hypothetical short- and long-positions) “based on the volume of the search debt … would have yielded a profit of 326 percent.” At the same time a traditional (simulated) “buy and hold” strategy would have yielded 16 percent.
Of course, there are all kinds of caveats (the authors didn’t have to pay trading fees, etc.), but this approach demonstrates how the aggregated thirst for knowledge by crowds and the resulting big-data sets may inform us rather than the oft-quoted “wisdom of the crowd.”
As the authors wrote: the “data did not only reflect aspects of the current state of the economy, but may have also provided some insight into future trends in the behavior of economic actors.”
However, please don’t go changing your investment strategies based on search trends just yet. Theories about efficient markets tell us that once the “genie is out of the bottle” and sophisticated traders have taken such new knowledge into account, arbitrage or profit opportunities will quickly disappear.
Or as Shakespeare put it in Macbeth (http://goo.gl/us7W7): “By the pricking of my thumbs, something wicked this way comes.”
Dr. Michael Reksulak teaches economics and public finance in Georgia Southern University’s College of Business Administration. He may be reached by email at mreksula@georgiasouthern.edu.