تقاضای اطلاعات و نوسانات بازار سهام

We study information demand and supply at the firm and market level using data for 30 of the largest stocks traded on NYSE and NASDAQ. Demand is approximated in a novel manner from weekly internet search volume time series drawn from the recently released Google Trends database. Our paper makes contributions in four main directions. First, although information demand and supply tend to be positively correlated, their dynamic interactions do not allow conclusive inferences about the information discovery process. Second, demand for information at the market level is significantly positively related to historical and implied measures of volatility and to trading volume, even after controlling for market return and information supply. Third, information demand increases significantly during periods of higher returns. Fourth, analysis of the expected variance risk premium confirms for the first time empirically the hypothesis that investors demand more information as their level of risk aversion increases.

Information is the most valuable and highly sought asset in financial markets. Unsurprisingly, a voluminous literature examines the intricacies between announcements, news and market activity. Probably due to data limitations, previous research efforts concentrate solely on exploring the role of supply of financial information. This is only one side of the story since the importance of demand for information is theoretically well established (see, for example, Kihlstrom, 1974; Grossman and Stiglitz, 1980; Radner and Stiglitz, 1984; Allen, 1990). In the present paper we exploit a new dataset to proxy intertemporal information demand on the basis of internet search volumes. This approach recognizes the fact that the internet has nowadays revolutionized the production, intermediation, dissemination and consumption of information in the financial industry (see Barber and Odean, 2001; Antweiler and Frank, 2004; Rubin and Rubin, 2010). Our dataset allows us to investigate for the first time the effects of information demand at the individual stock and at the overall market level, respectively. By controlling for information supply and by looking into both contemporaneous and dynamic relationships, we are able to shed light on issues related to the information formation and discovery process. Finally, by associating our novel proxy with the expected variance risk premium we can evaluate the validity of a longstanding hypothesis according to which demand for information is positively related to the level of risk aversion. Our empirical application focuses on 30 of the largest stocks traded on the New York Stock Exchange (NYSE) and NASDAQ. Proxies of idiosyncratic information demand for each stock are built on the basis of measures of the popularity of the company name as a keyword in the most popular internet search engine. Accordingly, information demand for the overall market is proxied using S&P 500 as the search keyword. In order to control for the known effects of information supply, we also take into account the number of firm-specific news stories along with the total number of news stories reported by the Reuters NewsScope service. Information demand for the stocks in our sample is shown to vary stochastically along a deterministic trend with seasonal regularities. Demand and supply tend to be positively correlated at the contemporaneous level while their dynamic relationship is characterized by causalities running in both directions. Overall, our results demonstrate that variations in information demand have a significant effect at the individual stock and overall market level in terms of historical volatility and trading volume. This effect is robust even after controlling for variations in the supply of information and in the market return. Market information demand has a consistently positive effect on all measures of market activity. However, the effect of idiosyncratic information on risk is mixed in direction and strength and diminishes if an implied measure of volatility derived from options market data is used to represent market activity. Our analysis also suggests that the effect of information demand on market activity becomes stronger in “high return” market states. Finally, we analyze expected variance risk premia for the S&P 500 and confirm the hypothesis which postulates that information demand increases along with the level of risk aversion in the market. The next section reviews the relevant literature and outlines the theoretical and methodological background. Section 3 discusses the datasets for measuring information demand and supply and offers a preliminary descriptive analysis. This section also presents the results of our empirical application. The final section concludes the paper.

This paper studies the supply and demand for idiosyncratic and market-related information and its relationship to stock market activity and risk aversion. The analysis employs a novel proxy for information demand which is derived on the basis of internet search volume for keywords related to 30 of the largest stocks traded on NYSE and NASDAQ and to the S&P 500 index, respectively. Our correlation and causality analysis indicates that information demand and supply are linked both contemporaneously and dynamically. Market information demand has a significant positive association with historical volatility, implied volatility and trading volume. Variations in information demand appear to have a significant effect at the individual stock and overall market level in terms of historical volatility and trading volume. This effect is robust even after controlling for variations in the supply of information and in the market return. The effect of idiosyncratic information on risk is shown to diminish if an implied measure of volatility derived from options data is employed. The analysis also suggests that the relationship between information demand and market activity becomes stronger during ‘‘high return’’ market states. Using the expected variance risk premium for the S&P 500 as a proxy for time-varying risk aversion, we confirm the hypothesis which wants information demand to increase along with the level of risk aversion in the market.