We report market experiments in which human traders are replaced by "zero-intelligence" programs that submit random bids and offers. Imposing a budget constraint (i.e., not permitting traders to sell below their costs or buy above their values) is sufficient to raise the allocative efficiency of these auctions close to 100 percent. Allocative efficiency of a double auction derives largely from its structure, independent of traders' motivation, intelligence, or learning. Adam Smith's invisible hand may be more powerful than some may have thought; it can generate aggregate rationality not only from individual rationality but also from individual irrationality.

The study reports on the ability of competing models of market information integration and dissemination to explain the behavior of simple laboratory markets for a one-period security. Returns to the security depended upon a randomly drawn state of nature. Some agents (insiders), whose identity was unknown to other agents, knew the state before the markets opened. With replication of market conditions the predictions of a fully revealing rational-expectations model are relatively accurate. Prices adjusted immediately to near rational-expectations prices; profits of insiders were virtually indistinguishable from noninsiders; and efficiency levels converged to near 100 percent.

This paper reports market experiments in which human traders are replaced by "zero-intelligence" programs that submit random bids and offers. Imposing a budget constraint (i.e., n ot permitting traders to sell below their costs or buy above their valu es) is sufficient to raise the allocative efficiency of these auctions close to 100 percent. Allocative efficiency of a double auction deri ves largely from its structure, independent of traders' motivation, intelligence, or learning. Adam Smith's invisible hand may be more powerful than some may have thought; it can generate aggregate rationality not only from individual rationality but also from individual irrationality.

The study reports on the applicability of competing models of market information integration and dissemination in explaining the behavior of simple laboratory, one-period security markets. Returns to the security depended upon a randomly chosen state of nature. Some agents (insiders), whose identity was unknown to other agents, knew the state before the markets opened. With replication of market conditions a model based upon rational expectations principles is relatively accurate. Prices adjusted immediately to near rational expectations prices; profits of insiders were virtually indistinguishable from noninsiders; and efficiency levels converged to near 100 percent.