The Not-So-Wisdom of Analyst Crowds

There is a fun paper out that compares the performance of teams of equity analysts with that of individual analysts. Contrary to “wisdom of crowds” tenets, the individuals beat the teams. Interestingly, however, the market (wrongly) thinks teams are better:

We compare the performance of teams of sell-side analysts with that of individual sell-side analysts. We find that teams forecast earnings less accurately than individuals forecast earnings. Based on team theory, we  hypothesize that the relatively poor performance of teams is due to less accountability for team versus individual forecast performance. Consistent with our accountability hypothesis, we show that larger teams forecast less accurately than smaller teams, and that analysts are more likely to get terminated or demoted for their poor individual forecasting performance relative to their poor team forecasting performance. We investigate the market perception of team forecasting performance, and show that stock prices react more strongly to earnings revisions by teams compared to revisions by individuals.


  1. The Wisdom of Crowds theory by Surowiecki says that groups will, on average, make better decisions than individuals as long as the group:
    -is diverse – if they’re all from similar backgrounds, it would fail the test
    -is independent – if they’re influenced by the opinion of others (as would happen w/ close collaboration), it would fail this test
    -has a good way of aggregating opinions
    I’m guessing the equity analyst teams are prone to groupthink and hence make bad calls.

  2. That’s because the same disease that afflicts investors — herd mentality — afflicts the average dummy Wall Street analyst.
    One example: BIDU, the weak internet content co out of China.
    Goldman just took them public and raised 4 billion smackers.
    Everyone and their grandma was all over this stock.
    At $140 a share, I called it “short of the year.”
    Look at the stock today.
    I’m sorry, the way I make money in the market is by avoiding co’s that do 2 million net income and somehow attain nosebleed market cap.
    I have a price target of $30 on Baidu and thats being generous…