We propose a parsimonious model of over-the-counter trading under asymmetric information to study the presence of intermediation chains that stand between heterogeneously informed market participants. We show that moderately informed intermediaries can reduce trading inefficiencies due to asymmetric information by layering an adverse selection problem over multiple transactions. Informed market
participants may prefer to trade through one or more of these intermediaries as they improve trade efficiency but also reduce the surplus accruing to uninformed traders. Our model makes novel predictions about optimal network formation when adverse selection problems impede the efficiency of trade.
Episodes of boom-bust cycles tend to occur in sectors with recent arrivals of new technologies and are often related to excessive funding by the financial sector. In this paper, I develop a dynamic general equilibrium model consistent with a role for the financial sector in propagation during such episodes. I extend a standard Schumpeterian growth model by incorporating (a) a monopolistically competitive financial sector and (b) time-varying technological conditions in real sectors. I identify two propagation channels. The first operates through financial firms’ acquisition of sector-specific knowledge (skill channel); financial firms chase "hot sectors" and thereby amplify fluctuations. The second channel originates in an interaction between competition in the financial sector and patent races in product markets (competition channel). Financial firms’ temporary competitive advantages in access to new ventures imply market segmentation: financial firms maximize the surplus generated by the client firms they can currently attract, anticipating competing financial firms’ future screening and funding decisions. Relative to the Pareto optimum, the competition channel generates over- investment in sectors with temporarily improved technological conditions; excessively high growth in these sectors comes at the cost of lower growth in the economy as a whole. The model links financial propagation to time variation in the cross section of asset prices. Exposures to aggregate risk dampen amplification effects.
I analyze the links between intertemporal information acquisition and the dynamics of asset markets. In my model, investors are Bayesian learners that optimally choose how much to consume, how much to invest, and how much information to acquire. The model predicts that investors acquire more information in times when future capital productivity is expected to be high, the cost of capital is low, new technologies are expected to have a persistent impact on productivity, and the scalability of investments is expected to be high. My results shed light on the economic mechanisms behind various dynamic aspects of information production by the financial sector, such as the sources of variation in returns on information acquisition for investment banks or private equity funds.