Working Papers

“Banking Crises as Self-Defeating Prophecies”, Job Market Paper, (link)

Banking crises are modeled as fire sales in a calibrated endowment economy. Rare and slowly evolving disasters, together with heterogeneous risk aversion, explain the existence of high banking leverage in normal times, simultaneous with a high risk premium, while accounting for a counter-cyclical Sharpe ratio. In the model, bankers are distinguished from depositors and real estate speculators only by their intermediate level of risk aversion. Contrary to the view that banking crises are the result of self-fulfilling prophecies, it is argued that they are in fact partially due to self-defeating prophecies. It is shown that if agents underestimate the potential for a drop in real estate prices, the realized drop will be larger than if their estimate had been accurate. This dynamic is due to the increase in ex-ante leverage and the resulting increase in ex-post fire sales. Increased leverage can account for up to two-thirds of the realized error. Moreover, the increase in leverage generates losses on mortgages that were considered, and would otherwise be, perfectly safe.


“Walrasian Economies with Debt: Leverage Causes Multiplicity”, joint with John Geanakoplos (link coming soon)

Since the seminal paper of Diamond and Dybvig (1983), bank runs as multiple Nash equilibria have been associated with a joint ownership problem. We study the conditions required for bank-run-like multiple equilibria in Walrasian economies with debt, in the absence of a joint ownership problem. We show that for multiple goods and multiple agents, enough leverage creates multiplicity. The intuition is simple:  Increasing bank leverage along the budget line does not change the original equilibrium, but it does change the derivative of excess demand for risky assets with respect to their price. Eventually the wealth effect overcomes the substitution effect, and the derivative of excess demand at equilibrium becomes positive, which implies multiplicity.


“Doom Triangles: Sovereign-Banking-Corporate Crises as Endogenous Leverage Disasters”, joint with John Geanakoplos (link coming soon)

The proper functioning of national banking systems is often thought to be predicated on the credit worthiness of their sovereigns. At the same time, the credit worthiness of a sovereign depends on tax revenues that are generated by an economy predicated on a functioning national banking system. We present a model of sovereign debt crises whose key features are: endogenously pro-cyclical banking leverage, government infrastructure expenditures that are determined by the demand for its debt, and banks as crossover investors in risky and safe debt markets. We show how the crisis is driven by the cross-amplification of four fundamental forces: banks' balance sheet losses, their endogenously forced deleveraging, a contraction in government expenditures, and banks' preference for legacy non-performing loans over new issues.



“Tracking Decision Makers under Uncertainty”, joint with Amos Arieli and Ariel Rubinstein, American Economic Journal: Microeconomics 3.4 (2011): 68-76. (link, >50 citations)

Eye tracking is used to investigate the procedures that participants employ in choosing between two lotteries. Eye movement patterns in problems where the deliberation process is clearly identified are used to substantiate an interpretation of the results. The data provide little support for the hypothesis that decision makers rely exclusively upon an expected utility type of calculation. Instead eye patterns indicate that decision makers often compare prizes and probabilities separately. This is particularly true when the multiplication of sums and probabilities is laborious to compute.