CompanyGraph
A price can be approximately right and still be difficult to improve upon. Efficiency depends on who can observe, trade, finance, and wait for information to matter.
What the hypothesis says
The efficient-market hypothesis is not one binary doctrine. In weak form, past prices and trading data should not reliably predict abnormal returns. Semi-strong tests ask whether publicly available information is incorporated quickly. Strong-form claims extend the idea to private information, a much harder proposition that is contradicted by the existence of informed insiders and trading restrictions.
Fama's review of efficient capital markets presents the theory and the joint-hypothesis problem: a finding that prices appear wrong may instead reflect a faulty model of risk or expected return. A price that later rises is not, by itself, evidence that it was previously mispriced.
Efficiency is therefore a claim about a defined market, period, information set, benchmark, and cost of trading. A liquid large-cap stock in a developed market can be efficient for one public earnings release and less efficient for a long-term change in customer behaviour that is expensive to analyze.
Trading is the correction mechanism
Prices incorporate information through participants who research, trade, and bear risk. For a discrepancy to close, someone must have capital, a position that expresses the view, a way to execute without excessive price impact, and enough time for the thesis to work. The trade itself can be costly or move the price before the position is complete.
Shleifer and Vishny's limits-to-arbitrage analysis explains why even an informed arbitrageur may be unable to correct a mispricing. Funding withdrawals, margin calls, short-sale constraints, uncertain timing, and investor redemptions can force a position out before the price converges.
Market design creates temporary demand
Index changes illustrate a mechanical demand shock. Index-tracking funds must buy additions and sell deletions according to the index rule, regardless of their own valuation. A New York Federal Reserve study of the S&P 500 index effect reviews evidence of price and demand effects around inclusion, while noting disagreement about their persistence and information content.
The event does not prove that the company became more valuable on the announcement date. It shows that market structure can move price when a class of investors has a required trade. Whether that effect is temporary, permanent, or offset by information requires an event study and a defined benchmark.
Why anomalies are hard to own
An apparent anomaly may disappear after publication, survive only in a small sample, require leverage, or be too crowded to trade at the original return. A small company can be cheap because information is scarce, but the investor may face poor liquidity, wide spreads, weak governance, or a business that cannot be valued reliably. Complexity can create opportunity and also reflect a risk the market understands better than the analyst.
Forced selling, redemptions, benchmark changes, tax deadlines, and regulation can create price pressure unrelated to new information. They can also create a reason that the pressure persists. A patient investor still needs financing, custody, a legal ability to trade, and a return sufficient to cover the time and risk.
Efficiency and behavioural explanations can coexist
Anchoring, overreaction, underreaction, and herding are plausible mechanisms, but a behavioural label does not establish a profitable trade. A negative announcement may trigger a large price move because it reveals lower future cash flow, not because investors are irrational. A small price response may reflect information already anticipated.
The correct test compares competing explanations. Define the information event, estimate the expected cash flows and risks, include transaction costs and taxes, and examine whether a repeatable strategy survives different periods and markets. The more flexible the backtest, the more evidence is needed against data-mining and selection bias.
What a responsible investor can infer
- Large, liquid markets are a demanding benchmark. A simple public fact is unlikely to be an overlooked source of excess return when many participants can act on it.
- Less coverage is not free alpha. Scarcity of analysts may reflect poor disclosure, illiquidity, governance risk, or expensive research.
- A mispricing needs a path. Identify the catalyst, holding period, funding, exit, and risk that can prevent convergence.
- Price is an observation, not a verdict. It reflects beliefs, constraints, market design, and information; it need not equal intrinsic value in a philosophical sense.
- Outperformance is evidence with a denominator. Compare against the right risk, liquidity, leverage, and benchmark, and include the cost of implementing the strategy.
Market efficiency is most useful as a disciplined starting assumption: prices are usually difficult to improve upon, but the degree and mechanism can be investigated. The investor's edge, if one exists, lies not in declaring the market wrong but in showing which participant is constrained, what information is missing, and why the trade can survive until the information is reflected.