The detailed comments from the referee, Anat Admati, improved the article considerably. Address reprint requests to F. Douglas Foster, The

We analyze a multi-period model of trading with differentially informed traders, liquidity traders, and a market maker. Each informed trader's initial information is a noisy estimate of the long-term value of the asset, and the different signals received by informed traders can have a variety of correlation structures. With this setup, informed traders not only compete with each other for trading profits, they also learn about other traders' signals from the observed order flow. Our work suggests that the initial correlation among the informed traders' signals has a significant effect on the informed traders' profits and the informativeness of prices.MANY FINANCIAL AND COMMODITY markets can be characterized by a number of informed traders, each with different information. As an example, in agricultural commodity markets differing assessments of crop quality, weather conditions, and demand for products means investors will have many, distinct views about the future value of a harvest. Depending on the crop and the growing environment, differences in the opinions of traders can be significant and persistent. Such heterogeneity of private information is not limited to agricultural markets. In financial markets, information may be related to corporate acquisitions, Federal Reserve policy, patent filings, mineral finds, marketing policy, et cetera. Investors' beliefs about these values and their subsequent implications for corporate value can be varied. Because of the richness and diversity of information in actual markets, one would like to model speculative trading with private information where informed traders have disparate information, and where this disparate information takes on a variety of correlation structures. Further, one would like the model to be The Journal of Finance dynamic to address the endogenous evolution of beliefs, and to relate this evolution of beliefs to trading volume and the informativeness of prices.Modeling markets with heterogeneous information among the traders becomes complex very quickly because traders infer the value of an asset from not only their own private information, but also using any information revealed by other traders through trading. For example, an informed trader may increase her estimate of the value of a security if she notices unusually large purchases of the asset that are unlikely to come from liquidity (uninformed) motivated trades. Any model that incorporates differential information in this setting must also address the inferences that traders make once trading occurs. This richer, and more realistic, view of a market raises a number of interesting and fundamental economic questions. How is information incorporated into prices in a market with differentially informed traders? How quickly do prices reflect the information known jointly by all informed traders? How do informed traders learn about each others' information? How might inferences of the informed traders affect market maker behavior? What happens to trading volume and the volatility of prices? What ...

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.. Wiley and American Finance Association are collaborating with JSTOR to digitize, preserve and extend access ABSTRACT Patterns in stock market trading volume, trading costs, and return volatility are examined using New York Stock Exchange data from 1988. Intraday test results indicate that, for actively traded firms trading volume, adverse selection costs, and return volatility are higher in the first half-hour of the day. This evidence is inconsistent with the Admati and Pfleiderer (1988) model which predicts that trading costs are low when volume and return volatility are high. Interday test results show that, for actively traded firms, trading volume is low and adverse selection costs are high on Monday, which is consistent with the predictions of the Foster and Viswanathan (1990) model. ACADEMICS, INVESTORS, AND REGULATORS alike are now intensively focused upon understanding the volatility of asset returns and its relation to trading volume. This interest was undoubtedly piqued by the market break of October 1987-a time during which volatility and trading volume reached unprecedented levels. But, even beforehand, researchers observed regular differences in the return process for various hours of the day and days of the week.Research concerning temporal patterns in stock market volatility and volume falls in two groups-studies that document observed patterns and studies that develop models to predict patterns. Among the studies in the first group are Oldfield and Rogalski (1980), French and Roll (1986), Stoll and Whaley (1990), Harris (1986), and Wood, McInish, and Ord (1985), who report evidence on seasonalities in daily and weekly return variances. Among the regularities that have been documented using interday data is that volatility is higher when the market is open than when it is closed. Oldfield and Rogalski (1980), French and Roll (1986), and Stoll and Whaley (1990), for example, point out significant differences in return volatility between trading The Journal of Finance and nontrading intervals. Overnight volatility is proportionately less than volatility during the trading day and weekend volatility is proportionately lower than trading day volatility. Using intraday data, Wood, McInish, and Ord (1985) document a U-shaped pattern in return volatility during the trading day, that is, volatility is highest at the beginning and the end of the trading day. Similarly, Harris (1986) shows strong intraday patterns in return volatility and presents his results by firm size. Along a different but related dimension, Jain and Joh (1988) show that the trading volume is different within and across days. They provide evidence of an inverted U-shape in volume across days. Monday an...

No abstract

No abstract