Blog - Investing Notes
October 10, 2008 - Converging Guru Accuracies
Do stock market gurus tend to anchor on bullish or bearish outlooks, regardless of market trends? If so, the distribution of their stock market forecasting accuracies should diverge when the market persists in one mode over a long period and converge when the market changes modes. The results at Guru Grades offer a limited way to test these suppositions. Based on trends in the 50 Guru Grades accuracy rates over the past year, we tentatively find that:
The following chart compares the distribution of 50 guru stock market forecasting accuracies, measured as described at Guru Grades, now and approximately one year ago. The distribution is now more compact than it was a year ago (after a mostly rising market during 2003-2007). Over the past year (as the market has substantially declined), the grades of persistent bulls (bears) have fallen (risen), compressing the distribution. Note that:
- Measuring the accuracy of squishy market forecasts is messy.
- Some gurus were publicly inactive (via selected sources) over the past year, so their measured accuracy rates do not change.
- The number of forecasts graded for some gurus (for example, for the lowest value in the distribution) is very small, so confidence in the associated accuracy rates is very low.
- Details show that not all gurus anchor on bullish or bearish outlooks. Some switch outlooks abruptly.
- The overall average forecast accuracy of the gurus is roughly constant at 48% over the past year, whether equally weighted or weighted by number of forecasts.
In summary, limited evidence suggests that, in aggregate, stock market gurus may tend to anchor on bullish or bearish positions, regardless of actual market trends.
This result further suggests that, in aggregate, stock market experts form beliefs based on information irrelevant to future stock market behavior.
For related research, see Blog Synthesis: The Wisdom of Analysts, Experts and Gurus.
