Efficient Market Hypothesis

The three forms are cumulative rather than alternative. The weak form assumes price reflects all past market data: historical prices, returns and volume. If it holds, technical analysis cannot generate risk-adjusted excess returns, because the patterns it exploits are already priced. The semi-strong form widens the set to all publicly available information: financial statements, disclosures, macroeconomic releases and analyst commentary. If it holds, fundamental analysis based on public data also fails to deliver abnormal returns consistently, though it remains the mechanism by which prices become informative. The strong form extends to private and insider information as well, implying no group of investors has privileged access to profit.

Because each set contains the one below it, strong-form efficiency implies semi-strong, which implies weak. Rejecting the weak form would therefore reject all three; rejecting the strong form says nothing about the other two.

Weak-form tests examine whether returns are predictable from their own history, using serial correlation of successive returns and runs tests on the sequence of price changes. Trading-rule studies ask whether filter or moving-average strategies survive transaction costs; most do not.

Semi-strong tests rely on event studies. The analyst defines an event date, estimates each security's normal return over an estimation window using a market model, then computes the abnormal return as the actual return minus the return predicted by that model. Abnormal returns are averaged across firms and cumulated over the event window to produce a cumulative average abnormal return. Efficiency implies the price adjustment is immediate and complete at announcement, with no systematic drift afterwards that a trader could capture.

Strong-form tests compare returns earned by corporate insiders, exchange specialists and professional managers against benchmarks.

If prices already reflected everything, no one would pay to research securities; but if no one researched, prices would reflect nothing. This is the Grossman-Stiglitz paradox: markets can only be efficient because a population of analysts competes to be first to information, and their gross returns must at least cover their costs. Efficiency is therefore a limiting condition approached most closely where analyst coverage is dense, trading is liquid, short selling is permitted and disclosure is timely. Small-capitalisation stocks, emerging markets and thinly traded instruments show measurably slower adjustment.

Documented anomalies complicate the picture. Value and small-firm effects, momentum, post-earnings-announcement drift and calendar effects have all survived some scrutiny, but many weaken out of sample, disappear after transaction costs, or are reinterpreted as compensation for risk factors rather than free returns.

Candidates should be able to place a described piece of information into the correct information set and state which form of efficiency it tests. A trading rule based on last month's prices tests the weak form; a study of price reaction to an earnings release tests the semi-strong form; profits earned by directors before an announcement test the strong form.

The practical implication is the joint-hypothesis problem: any test of efficiency is simultaneously a test of the asset-pricing model used to define normal return, so abnormal performance may reflect a mis-specified benchmark rather than inefficiency. Investment consequences follow directly. Under semi-strong efficiency, indexing, low turnover and cost control dominate, and the role of the active manager becomes risk-matching the client rather than forecasting. Behavioural finance supplies the standing challenge, explaining persistent mispricing through overconfidence, herding, anchoring and loss aversion.