Using a comprehensive database of headlines about individual companies, I examine monthly returns following public news. I compare them to stocks with similar returns, but no identifiable public news. There is a di#erence between the two sets. I find strong drift after bad news. Investors seem to react slowly to this information. I also find reversal after extreme price movements unaccompanied by public news. The separate patterns appear even after adjustments for risk exposure and other e#ects. They are, however, mainly seen in smaller, more illiquid stocks. These findings support some integrated theories of investor over- and underreaction.

Text is not unadulterated fact. A text can make you laugh or cry but can it also make you short sell your stocks in company A and buy up options in company B? Research in the domain of finance strongly suggests that it can. Studies have shown that both the informational and affective aspects of news text affect the markets in profound ways, impacting on volumes of trades, stock prices, volatility and even future firm earnings. This paper aims to explore a computable metric of positive or negative polarity in financial news text which is consistent with human judgments and can be used in a quantitative analysis of news sentiment impact on financial markets. Results from a preliminary evaluation are presented and discussed. 1

This paper reexamines frozen concentrated orange juice (FCOJ) futures returns as they relate to fundamentals, in particular, temperature. We show that when theory clearly identifies the fundamental, i.e., at temperatures close to or below freezing, there is a close link between FCOJ prices and that fundamental. Using a simple theoretical nonlinear model of the relation between FCOJ returns and temperature, we can explain approximately 50 % of the return variation. This is important because while only 4.5 % of the days in winter coincide with freezing temperatures, two-thirds of the entire winter return variability occurs on these days. Moreover, when theory suggests no such relation, i.e., at most temperature levels, we show empirically that none exists. The fact that there is no relation the majority of the time is good news for the theory and market efficiency, not bad news. In terms of other FCOJ return volatility, we also show that other fundamental information about supply, such as USDA production forecasts and news about Brazil production, generate significant return variation that is consistent with theoretical predictions. The evidence in this paper suggests that the literature’s conclusion about irrationality drawn from the FCOJ market have more to do with econometricians ’ lack of modeling ability than with the empirical facts. 1

This paper shows how to extract information from equilibrium trading volume. The analysis is first carried out in a market-clearing framework with symmetrically (and later normally) distributed demands, and is then extended to include market-making models. The conclusions of this paper hence apply to the majority of the models developed in the noisy rational expectations literature. If a random variable is symmetrically distributed with the traders' demands around zero and the asset market clears, the volume-based conditional distribution of this variable is symmetric, and consequently its conditional expectation based on volume is zero. The random variable under consideration may be the true value of the asset, the price or, in a dynamic model, the price difference. The paper further proves that the covariance between the absolute value of any variable jointly normally distributed with the traders' demands and the equilibrium volume is positive, which agrees with the empirical evidenc...

Numerous empirical studies have demonstrated that asset prices react rapidly, if at all, to news published in the mass media. In many cases, the information has been discounted and prices have already moved upon primary publication through news wires, press releases or firm announcements. Any remaining information is usually quickly priced in after dissemination through the mass media. But not always: Often enough delayed price adjustments, underreactions as well as overreactions, can be observed after particular news reports have been published. This points to inadequacies in the efficient markets hypothesis as well as in Behavioral Finance theories: Delayed reactions appear too often to be explained away as "anomalies" within models of rational pricing. But they appear too eratically to be explained as "normalities" such as in newer models of systematically irrational pricing. In other words: Asset prices frequently do not react to news published in the media. Sometimes they do. The evidence leads to the following conclusion: That markets can be efficient and inefficient at the same time.

Stock market event studies are often used to estimate the impact of an unanticipated event on stock returns of a company. Traditionally, these analyses focus on developing estimates of abnormal returns attributed to the event or some measure of post-event loss in shareholder value. In 1989 Mark Mitchell used an event study to estimate the impact of the 1982 Tylenol poisonings on Johnson & Johnsons share returns. Mark Mitchell was able to demonstrate that (1) Johnson & Johnson share returns were significantly impacted by the poisonings, and (2) such an impact translated, at least in part, to a depreciation of brand name capital. This study sets forth the basic framework of Mark Mitchells 1989 analysis and wherever appropriate, provides possible alternatives to his methodologies. Using several alternative approaches including, but not necessarily limited to, consideration of the incremental values associated with the Tylenol brand name, cost to develop the brand, alternative market factors, and changes in income streams I compare changes in brand value to brand name capital depreciation estimated by Mitchell. In some instances the aforementioned approaches are used in conjunction with aspects of Mitchells methodology. The results tend to provide more accurate estimates of the loss in brand value possibly associated with the 1982 poisonings. III Thanks and Dedication The author would like to thank Roger Waud, Ph.D., Dr. Thomas Lutton Ph.D., and Richard Theroux, Ph.D. for their valuable insights and helpful comments throughout these past months. I would like to dedicate this work to my wife Eileen for putting up with mood swings, long hours, distractions and seemingly endless economic discussions. It ha...

Abstract. The efficient market hypothesis states that investors immediately incorporate all available information into the price of an asset to accurately reflect its value at any given time. The sheer volume of information immediately available electronically makes it difficult for a single investor to keep abreast of all information for a single stock, let alone multiple. We aim to determine how quickly investors tend to react to asset specific news by analysing the accuracy of classifiers which take the content of news to predict the short-term market reaction. The faster the market reacts to news the more cost-effective it becomes to employ content analysis techniques to aid the decisions of traders. We find that the best results are achieved by allowing investors in the US 90 minutes to react to news. In the UK and Australia the best results are achieved by allowing investors 5 minutes to react to news. Keywords. Document Classification, Stock Market, News, SVM, C4.5. 1.

Assessing the impact of macroeconomic news announcements

Walter Krämer and Volker Wolff for their friendly support and valuable advice. News Events and Price Movements Numerous empirical studies have demonstrated that asset prices react rapidly, if at all, to news published in the mass media. In many cases, the information has been discounted and prices have already moved upon primary publication through news wires, press releases or firm announcements. Any remaining information is usually quickly priced in after dissemination through the mass media. But not always: Often enough delayed price adjustments, underreactions as well as overreactions, can be observed after particular news reports have been published. This points to inadequacies in the efficient markets hypothesis as well as in Behavioral Finance theories: Delayed reactions appear too often to be explained away as “anomalies ” within models of rational pricing. But they appear too eratically to be explained as “normalities ” such as in newer models of systematically irrational pricing. In other words: Asset prices frequently do not react to news published in the media. Sometimes they do. The evidence leads to the following conclusion: That markets can be efficient and inefficient at the same time.

The impact of news on the exchange rate of the lira and long-term interest rates