by Dr Shiqi Chen, Research Associate, Cambridge Centre for Finance and Cambridge Endowment for Research in Finance
We are entering into an era of Big Data, in which the demand and supply of data and information are growing exponentially. Such unprecedented demand and supply have triggered reforms in many sectors, where the financial industry is at the forefront of the trend. According to the reports by IBISWorld, the revenue of the US financial data providers industry will reach $14.6 billion in 2021, and the expected annual growth rate is 8.2%. Financial data providers refer to those who collect and pack data from various sources such as brokers, regulatory filings, stock exchange feeds, and supply to financial institutions, investors, analysts, corporate executives etc.
The increasing importance of data information rests on the common questions faced by every participant in the financial market: How to make financial decisions under uncertainty. The growing complexity and surging uncertainties in the market induce participants to engage in a broader scale of information search, especially through different information intermediaries, in order to make better decisions and improve financial return. Examples are abundant: companies perform due diligence before investment, mergers and acquisition decisions; producers purchase consumer data for production and marketing purpose; fund managers rely on analysts’ report to adjust portfolios; retail investors invest in mutual fund hoping to draw on their information advantage.
The substantial demand gives rise to many interesting questions. Do agents bounded by incomplete information behave differently compared with those with perfect information? What is the influence of information acquisition on financial behaviour? In particular, given the dynamic nature of information flow and learning, how does the influence of information vary across states? What is the relation between public information and private information? Is more public information going to crowd out private information production, or the opposite?
Existing literature has shed light on some of the questions raised above. For example, earlier papers by Gennotte (1986), Brennan (1998), and others show that investors with incomplete information have different portfolio allocations compared with those with perfect information and introduce the concept of “estimation risks”. Barberies (2000) and Xia (2001) study if stock returns are predictable from a signal such as dividend yield, does the traditional argument – investors with long investment horizon (eg young investors) should invest more into equity than those with relatively short investment horizon (eg retired age investors) – still hold. Even though using a very different modelling approach, both papers show that this is not necessarily the case.
More recently, Gao and Huang (2018), and Goldstein, Yang and Zuo (2020) use the implementation of the Electronic Data Gathering, Analysis, and Retrieval (EDGAR) system to study how modern information dissemination technology affects the relation between public and private information production. Gao and Huang (2018) find that after the full implementation of the EDGAR system, trades by investors with internet access become more informative, and the quality of sell-side analyst report increases significantly. Their findings suggest that public information does encourage private information production. Goldstein, Yang and Zuo (2020) focus on the staggered implementation process and discover a decrease in the investment-to-price sensitivity. They suggest that the propagation of public information can crowd out private information production incentives.
These intriguing results demonstrate that information does affect agents’ behaviour along various dimensions. However, some interesting questions are still underexplored, especially from a theory perspective. For example, how does the demand for information varies over time? Do investors conduct more information acquisition in good or bad times? How do different factors such as risk preferences, beliefs, volatility affect the demand? Answer to these questions can help to explain agents’ financial behaviour, for example, the trend-chasing phenomena, and passive and active investment choices as active investment can be considered as one way of information acquisition. It can also improve our understanding of the value created by information intermediaries. Bhattacharya et al. (2009) show that although there is a positive relation between news coverage and internet IPOs in the 1990s, the news frenzy fails to explain the interest bubble. The news information factor can only explain 2.9% of the astonishing 1646% return difference between internet and non-internet firms. This, to some extent, implies that preference or demand for information is state-dependent, so is the value of information.
Therefore, exploring how investors demand for information vary over different states can help to understand many market phenomena that otherwise remain puzzling. Let me stop here by referring to the interview of Robert Shillers in the New York Times: “The fundamental problem is that the information obtained by any individual – even one as well-placed as the chairman of the federal reserve – is bound to be incomplete.”
Barberis, N. (2000) “Investing for the long run when returns are predictable.” Journal of Finance, 55(1): 225-264
Bhattacharya, U., Galpin, N., Ray, R. and Yu, X. (2009) “The role of the media in the internet IPO bubble.” Journal of Financial and Quantitative Analysis, 44(3): 657-682
Brennan, M.J (1998) “The role of learning in dynamic portfolio decisions.” Review of Finance, 1(3): 295-306
Gao, M. and Huang, J. (2020) “Informing the market: the effect of modern information technologies on information production.” Review of Financial Studies, 33(4): 1367-1411
Gennotte, G. (1986) “Optimal portfolio choice under incomplete information.” Journal of Finance, 41(3): 733-746
Goldstein, I., Yang, S. and Zuo, L. (2020) The real effects of modern information technologies. National Bureau of Economic Research, No. w27529.