Browsing by Subject "content bias"

This dissertation comprises two chapters studying how information transmitted in the economy and the financial markets becomes compressed in communications and its consequences. In chapter one, the compression is due to limited communication bandwidth, and in chapter two, it is due to limited attention on the receiver’s end.

Chapter one discusses how information intermediaries selectively present evidence to serve financial decision-makers. Faced with a space limit for their communication reports, the information intermediaries present information selectively. I solve the model for the optimal messages in the intermediaries' communication with the decision-makers and investigate the relationship between the apparent messages and the inferred economic fundamentals. The two main findings are: (i) the model matches many stylized facts about content biases such as prior or extreme biases; (ii) I derive an analytical relationship between the messages and the inferred fundamentals in the asymptotics. This relationship can be conveniently used to interpret observed content biases and quantitatively analyze the effects of the context on the interpretation of contents. The theory also shows that content biases may improve rather than decrease welfare. The model relates to empirical content analysis using frequency-based proxies and can be used to analyze contextual effects on contents.

In chapter two, I develop and analyze a theoretical model that shows how investors allocate their limited attention resources to monitor a wide selection of target firms. An investor with limited attention demands information about the types of her portfolio firms before investing. The firms strategically supply good news and withhold bad news. The investor may press companies to reveal more information by allocating more costly attention to them. Because the benefit to attention is convex, the investor will optimally focus on a subset of firms and acquire complete information while giving up learning anything about the other firms. Firms in the scrutinized subset have low investigation costs and a high Expected Value of Perfect Information (EVPI), and they always receive an efficient amount of capital. The other firms are provided with an inefficient level of capital and suffer from extreme asymmetry in information transparency. The result rationalizes convertible debt as a socially optimal financing instrument for private firms. It can be applied to a venture capital context to analyze entrepreneurial investment relationships.