Dreman references many different research papers that demonstrate that company earnings appear to follow a random walk. In this random walk scenario, past and future earnings show almost no correlation together. Additional research shows that stock analysts predominantly extrapolate existing trends into the future. They project a continuation of the existing trend, despite the number of the inputs available for consideration. If earnings follow a random walk and analysts are projecting existing trends forward, it is no wonder why the analysts have such a poor track record for estimating earnings. The implication is that since analysts base stock recommendations on their earnings forecasts, in aggregate they make poor stock recommendations.
According to Dreman, one important fundamental truth about competitive markets is the constant push towards average company returns. It is because of this truth that many high P/E stocks have their high P/E's reduced dramatically and many low P/E stocks have their P/E's rise substantially.
There are various forces that can act to disturb high P/E stocks, including increased competition (in a free market), government regulation, socioeconomic events, inadequate management and increased costs that cannot be passed on. Likewise, for companies experiencing current difficulties, often improved operational focus including reducing costs and slowing down expansion can help act as a catalyst for once again achieving above average growth.
Dreman concludes that since both high P/E and low P/E stocks represent extreme situations, there is a high probability that their returns will be pushed towards the average. He also opines that most investors won't see it this way! Due to human biases, such as the "recency effect", market participants tend to price adverse or sublime conditions on a continuing basis. Dreman observes that the consistent overreaction by most market participants towards current trends opens up a great opportunity for contrarian investors to profit.
Dreman introduces his "Investor-Overreaction Hypothesis" which essentially boils down to investors overreacting to primarily earnings disappointments (but also to other disappointments) by overselling stocks and bonds and thereby bringing prices of these securities down too far. These low prices offer an excellent opportunity for investors to buy in cheap.