Search

Search results

    Bordalo, Pedro, Nicola Gennaioli, Andrei Shleifer, and Stephen J. Terry. Working Paper. “Real Credit Cycles”. Abstract
    Recent empirical work has revived the Minsky hypothesis of boom-bust credit cycles driven by uctuations in investor optimism. To quantitatively assess this hypothesis, we incorporate diagnostic expectations into an otherwise standard business cycle model with heterogeneous firms and risky debt. Diagnostic expectations are a psychologically founded, forward-looking model of belief formation that captures over-reaction to news. We calibrate the diagnosticity parameter using microdata on the forecast errors of managers of listed firms in the US. The model generates countercyclical credit spreads and default rates, while the rational expectations version generates the opposite pattern. Diagnostic expectations also offer a good fit of three patterns that have been empirically documented: systematic reversals of credit spreads, systematic reversals of aggregate investment, and predictability of future bond returns. Crucially, diagnostic expectations also generate a strong fragility or sensitivity to small bad news after steady expansions. The rational expectations version of the model can account for the rst pattern but not the others. Diagnostic expectations offer a parsimonious account of major credit cycles facts, underscoring the promise of realistic expectation formation for applied business cycle modeling.
    Bordalo, Pedro, Katherine Coffman, Nicola Gennaioli, Frederik Schwerter, and Andrei Shleifer. Working Paper. “Memory and Representativeness”. Abstract

    We explore the idea that judgment by representativeness reflects the workings of episodic memory, especially interference. In a new laboratory experiment on cued recall, participants are shown two groups of images with different distributions of colors. We find that i) decreasing the frequency of a given color in one group significantly increases the recalled frequency of that color in the other group, ii) for a fixed set of images, different cues for the same objective distribution entail different interference patterns and different probabilistic assessments. Selective retrieval and interference may offer a foundation for the representativeness heuristic, but more generally for understanding the formation of probability judgments from experienced statistical associations.