Why a market can be efficient and wrong at once

A price that reflects everything currently known is described as efficient. This is a claim about the speed with which information travels, not about the accuracy of the resulting number. If what is known is thin or mistaken, the price can be simultaneously efficient in the technical sense and a poor estimate of the underlying situation.

What everyone sees

Efficiency is popularly read as correctness, so a market price is treated as a verdict. The technical definition is narrower: it says that publicly available information is already incorporated, so it cannot be used to obtain a predictable advantage. Correctness would additionally require that the available information be adequate, which is a separate question the definition does not address.

What is actually happening

Grossman and Stiglitz showed that a fully informationally efficient market is not coherent as an equilibrium, because if prices revealed everything, nobody would be paid for gathering information and the gathering would stop. Efficiency must therefore be partial, sustained by participants who are compensated for producing information not yet in the price. Shleifer and Vishny added the constraint on correction: even a participant who identifies a mispricing faces limits — finite capital, horizons and the risk that the gap widens first — so the mechanism that removes errors is bounded. Both results describe the same structure. Prices aggregate beliefs efficiently and beliefs are not facts. This is a description of an information mechanism, not guidance about any market.

Why it stays hidden

The confusion survives because the word efficient carries an ordinary meaning of working well. A term defined narrowly for information incorporation is heard as a general endorsement, and the endorsement then does argumentative work its definition cannot support. The gap only becomes visible after a revision, when it is attributed to new information rather than to information that had been absent all along.