Capital market theory has passed through two distinctly different paradigms in the past 80 years and is experiencing the rise of a third. Each paradigm has attempted to better explain the movement of market prices. The currently ascendant paradigm, based on new research in the field of behavioral finance, promises to offer superior guidance to investors and advisors who hope to harness the pricing distortions created by widespread cognitive errors.
The first paradigm in this progression was launched by Graham and Dodd (GD) in 1934 with publication of their now seminal book, Securities Analysis, which provided the first systematic
approach to analyzing and investing in stocks. GD argued that it was possible to build superior stock portfolios using careful fundamental analysis and a set of simple decision rules. These rules were based on the emotional mistakes made by the market that could be identified via fundamental analysis. The success of GD is all the more impressive because their book appeared in the depths of the Great Depression, when stocks were crashing and market volatility was reaching levels not seen before nor since.
GD’s dominance lasted 40 years, until the ascendency of modern portfolio theory (MPT) in the mid-1970s. MPT agreed that there were many emotional investors, but there were enough
rational investors to arbitrage away pricing mistakes. Therefore market prices were “informationally efficient.” A consequence of this theory was that it was not worth conducting a GD-type of analysis, or any analysis for that matter. Instead, an investor should simply buy and hold an index portfolio.
MPT immediately ran into problems with the publication of two studies, with Basu (1977) demonstrating that stocks with low price-to-earnings ratios outperformed high PE stocks and
Banz (1981) showing that small stocks outperformed large stocks. MPT had no answer for these anomalies. In order to save the model, the two were sucked into MPT as “return factors.” It has been downhill for MPT ever since, with study after study uncovering one anomaly after another.
Full PDF see here: SSRN