Capital Ideas — Chapter 1: Are Stock Prices Predictable?

1. 1. Louis Bachelier's 1900 dissertation argued that stock prices follow a random walk, making them unpredictable.
2. 2. Bachelier's random walk model predicted that price fluctuations grow proportionally to the square root of time, which held true for U.S. stocks over 60 years.
3. 3. The Dow Theory, developed by Charles Dow, assumes stock price trends persist until a reversal signal appears, but it often generates late or conflicting signals.
4. 4. Alfred Cowles 3rd's 1933 study found that 16 financial services, 20 insurance companies, and William Peter Hamilton's Dow Theory calls all failed to outperform a buy-and-hold strategy.
5. 5. Cowles concluded that the average forecaster performed worse than pure chance, and even the best forecasters could not demonstrate skill.
6. 6. The Cowles Commission's stock index, covering 97% of NYSE market value, was statistically superior to the Dow Jones Averages but less timely due to computational limits.
7. 7. Despite evidence against predictability, investors continue to seek forecasting services because they want to believe someone knows the future.