Ambiguity and welfare losses
Ilut and Schneider review how ambiguity, defined as uncertainty when the odds are not known, and ambiguity aversion have been used by economic literature to understand aggregate fluctuations, asset pricing puzzles, decisions of heterogeneous agents in micro data, and optimal policy.
The basic premise is that averse agents are not confident enough to assign probabilities to events about which they have little information, as a result when they have to evaluate a plan, they act as if they are relatively pessimistic about it.
The authors argue that the difference between ambiguity and risk is that only the former can generate first order welfare effects:
Intuitively, an agent contemplating positive (negative) exposure to ambiguity evaluates plans based on low (high) mean payoffs and chooses zero exposure if both are sufficiently unfavorable. This feature leads to parsimonious explanations for a number of robust facts in micro data – for example, non-participation in asset markets that an investor is not familiar with, rigidity and memory in nominal prices, or lack of adoption of new technologies. We note that market frictions, technological rigidities or curvature in utility is not required for the argument. This is contrast to results on inaction due to risk: for example, ”wait-and-see” effects require irreversibility of investment, which induces curvature in the objective function.
Modeling uncertainty as ambiguity: a review
Authors: Cosmin Ilut, Martin Schneider
From: Duke University, Stanford University
Ambiguity and monetary policy
In this paper Donghai Zhang documents excess sensitivity i.e. the significant impact of monetary policy on long-term interest rates, as being more pronounced in response to monetary policy easing than monetary tightening. To understand the source of such asymmetric excess-sensitivity he introduces Knightian uncertainty (i.e., ambiguity) and ambiguity averse agents into a standard New Keynesian model. Policy actions signal the economy’s unobserved state because of the central bank’s private information. Knightian uncertainty about the precision of the policy signal and ambiguity-averse preferences give rise to the ambiguous signaling channel of monetary policy.
The private agent faces Knightian uncertainty and places a higher weight on the monetary policy signal. In such a framework, agents react more to bad (unexpected monetary easing) than to good (unexpected monetary tightening) news. For example, in the case of an expansionary monetary policy shock, although an unexpected reduction in the short-term interest rate stimulates consumption, it also leads to a downward forecast revision of the natural rate which reflects the long-term growth prospect. Such pessimistic belief updating leads to a reduction in the current consumption; therefore, the monetary policy signaling channel dampens the effects of the monetary policy.
The Ambiguous Signaling Channel of Monetary Policy
Authors: Donghai Zhang
From: University of Bonn
