Objective research and reviews to aid investing decisions
 Looking for information on debt consolidation loans? Want to know more about reverse mortgages? |
Do individuals whose stock market forecasting records are good (bad) gain (lose) attention? The "pro" argument is that investors/traders, seeking a market timing edge, eventually flock to good forecasters and ignore bad ones. The "con" arguments are that loud noise (for example, marketing-related or entertainment-driven) swamps information, and/or investors/traders do not or cannot measure accuracy, and/or investors/traders are more interested in ideas than forecasts. As a simple, partial test these arguments, we use two data series: (1) the stock market forecasting accuracies of the 43 named gurus in Guru Grades; and, (2) the attention paid to these same individuals as defined by the number of matches found by a Google query on ("[guru name]" "stock market"), with the "stock market" term intended to filter out potential namesakes and relate each name to the forecasted variable. We find that:
The following scatter plot relates attention (Google links) and stock market forecasting accuracy for all 43 named gurus. The Pearson correlation for this distribution is 0.03 and the R-squared statistic is 0.00, indicating no relationship between forecasting accuracy and attention. However, the sample is small, and the extreme attention garnered by some gurus (Jim Cramer and Marc Faber, and perhaps Jim Jubak and Bill Cara) might obscure any pattern among the not-so-famous gurus.
What happens if we omit these four "stock stars"?
The next plot relates attention (Google links) and stock market forecasting accuracy for the 39 not-so-famous gurus. The Pearson correlation for this distribution is 0.01 and the R-squared statistic is 0.00, again indicating no relationship between forecasting accuracy and attention. However, if we exclude only Jim Cramer and Marc Faber, the correlation is -0.28 (less accuracy means more fame) and the R-squared statistic is 0.08. This instability of results derives from the small size and large dispersion of the sample.
Are there ranges of attention which relate reliably to forecasting accuracy?
The final chart offers a different perspective on the relationship between attention (Google links) and stock market forecasting accuracy. It orders the former series from lowest to highest values over the entire sample. This format facilitates understanding whether the relationship might be more significant or noisier for some levels of attention than for others. The chart generally confirms the non-relationship shown in the above scatter plots. The lack of trend in forecasting accuracy indicates that it is unconnected to Google name recognition.
The graph of Google links indicates that investors/traders collectively concentrate attention on just a few "star" gurus.
In summary, there is no clear relationship between the forecasting accuracy of and the attention paid to stock market gurus. It seems that other factors drive attention, which tends to concentrate on a relatively few gurus.
Note that the significance of this analysis depends on the reliability of our judgments regarding guru forecasting accuracies, the meaningfulness of a Google-based definition of attention and sample size. Google name recognition varies considerably from prior tests for some gurus, most notably for Ken Fisher. The search ("Ken Fisher" "stock market") generated only 31,700 results for this analysis, compared to 178,000 in January 2007, 138,000 in October 2006 and 599,000 in February 2006. This dramatic decline is puzzling.
For related research, see Blog Synthesis: The Wisdom of Analysts, Experts and Gurus.
