Market efficiency, its three forms and anomalies

Nine thirty-one

At 9:31 a.m. a newswire reports that Tamarack Foods is recalling a product line. By 9:32 the share is down 6 %. Nobody on the floor of anything decided that; thousands of orders arrived within seconds, each from someone who read the headline and knew roughly what it was worth. That is the efficient market hypothesis (EMH): security prices reflect available information, and reflect new information almost as soon as it arrives. "Efficient" does not mean the price is correct — Tamarack may be worth more or less than the market's 9:32 guess. It means the price is unpredictable from what is already known: everything known is already in it, so only new information moves it, and new information is by definition a surprise. Prices therefore follow a random walk — the changes are unpredictable, with a positive expected return as the reward for bearing risk. The levels trend up with that return; it is the changes that are random.

Three forms, nested

How much information is "available"? The hypothesis comes in three strengths, each containing the last:

Weak form — prices reflect all past prices and volumes. If true, studying charts for patterns (technical analysis) cannot earn an excess return: any pattern in the past is already priced. The test: do past returns predict future ones? Mostly they do not.

Semi-strong form — prices reflect all publicly available information: financial statements, news, analyst reports, the recall at 9:31. If true, fundamental analysis of public data cannot earn an excess return either, because by the time you have read the annual report the price has read it too. The test: does the price adjust fully and quickly to announcements? Largely, yes.

Strong form — prices reflect all information, including private (inside) information. If true, even an insider could not profit from what only they know. The test is direct: do insiders earn excess returns? They do — and insider trading on undisclosed material facts is illegal precisely because it is profitable (Lesson 5.5). The strong form fails; the evidence for the weak and semi-strong forms is considerable.

Two other words for efficiency

Operational efficiency is a different thing: how cheaply and quickly a trade executes — spreads, commissions, settlement in T+1. A market can be operationally efficient (fast, cheap) without being informationally efficient, and vice versa. Allocational efficiency is a third: capital flows to its best use. The EMH is about informational efficiency only.

Anomalies

Researchers have found patterns that a semi-strong-efficient market should not have: a January effect (small firms outperforming in early January), a small-firm effect, momentum (recent winners keep winning for some months), overreaction (long-run losers rebound). Two things about them: they are small after transaction costs, and they tend to fade once published — which is itself the EMH at work. An anomaly is an exception that tests the rule; it is not a form of efficiency, and it is not a reliable strategy.

What follows for an investor

If prices already reflect what you know, the expected reward for picking shares is the cost of trying. The practical answer for a non-expert is a diversified, low-cost portfolio that accepts the market's return — the index fund of Lesson 13.7. Module 14's experiment tests that claim with real prices.

Assign the strategies

Charting a 50-day moving average → defeated by the weak form. Buying after a strong earnings release → defeated by the semi-strong form. Buying on a tip from a director before the release → would be defeated only by the strong form, which the evidence rejects (and the trade is illegal). Cold: the strong-form test is whether insiders earn excess returns; they do, so it fails.