Efficient Capital Markets: II
Read the paperopens doi.org in a new tab
What they found
Twenty years after his original review, Fama updated the scorecard. He replaced the weak/semi-strong/strong labels with tests of return predictability, event studies, and tests for private information, and openly conceded that returns are somewhat predictable from dividend yields, term spreads, and past returns, and that the size and January effects were real puzzles. He also named the joint-hypothesis problem as the central difficulty: you can never test efficiency alone, only efficiency plus a model of expected returns.
What you can use
- Even the leading defender of efficiency accepted that some return predictability exists; the question is whether it is exploitable after costs.
- Any claim that a strategy beats the market depends on the risk model you compare it against.
- Long-horizon predictability (years) is much better supported than short-horizon (days), which is the horizon most active traders care about.
Caveats
Survey, not new evidence. Data largely from U.S. equities before 1990. Many of the predictability results discussed have weakened out of sample.
Tags: efficiency, theory, survey
Summaries are our own reading of the paper, not the authors' words. Educational only, not advice. Discuss it in Book Club.