Amplification of Uncertainty in Illiquid Markets
Central Bank of Chile
November 3, 2011
AFA 2012 Chicago Meetings Paper
This paper argues that the capacity of financial markets to aggregate dispersed information about economic conditions is diminished in times of distress, resulting in countercyclical uncertainty. Building on a rational expectations equilibrium dynamic environment, I model informed traders as financial intermediaries facing countercyclical fund outflows. As conditions deteriorate, households pull out their funding from intermediaries and force premature liquidations, exposing intermediaries to lower expected returns and non-fundamental price fluctuations. In anticipation, risk-averse intermediaries trade less aggressively to private information and the informational content of equilibrium asset prices is reduced. The model highlights a dynamic interdependence between price informativeness and the endogenous severity of liquidations when both intermediaries and traders who absorb their asset liquidations learn from prices. This mutual reinforcement creates a strong internal amplification mechanism that delivers sharp spikes in economic uncertainty as funding becomes tighter. The mechanism can explain the time-variation in the risk premium, Sharpe ratio and volatility even when risk aversion and the variance of fundamental shocks remain constant. In contrast to theories stressing exogenous variations in preferences or heteroskedastic volatility of fundamentals, the mechanism highlighted suggests fluctuations in the risk premium can be sub-optimal to the extent they arise from the endogenous disaggregation of information in funding-constrained asset markets.
Number of Pages in PDF File: 44
Keywords: Business cycle, funding constraints, information aggregation, uncertainty
JEL Classification: E23, E32, G12, G14, G23working papers series
Date posted: March 21, 2011 ; Last revised: November 4, 2011
© 2014 Social Science Electronic Publishing, Inc. All Rights Reserved.
This page was processed by apollo8 in 0.250 seconds