"Robust Control and Filtering under Rational Inattention in a Permanent Income Model" with Yulei Luo. In this paper we consider robust control and filtering under limited information-processing capacity (rational inattention or RI) in an otherwise standard permanent income model. We first solve the robustness (RB) versions of the permanent income model with RI explicitly and develop three main results. First, the observational equivalence between robustness (or risk-sensitivity) and the discount factor obtained by Hansen, Sargent, and Tallarini (1999) in the perfect-state-observations permanent income model holds exactly in the RI version of the model when the agent distrusts the Kalman filtering equation governing the evolution of the perceived state. Second, both a stronger preference for robustness in the Kalman filtering gain and higher channel capacity increase the Kalman gain. Third, concerns about the shocks to the perceived state and concerns about the Kalman gain have different effects on consumption dynamics, precautionary savings, and the welfare costs of uncertainty.

"Crowding Out or Crowding In? UI and Private Insurance." This paper reconsiders the results in Young (2004) under the assumption that debt limits are endogenously determined in order to prevent default, as in Kehoe and Levine (1993). I find that eliminating UI is a net welfare loss for the economy because it eliminates all borrowing, although for smaller changes in the replacement rate the effect of changing debt limits is minor. General equilibrium effects dominate. Preliminary and relatively dormant.

"Optimal Policy for Macro-Financial Stability" with Gianluca Benigno, Huigang Chen, Christopher Otrok, and Alessandro Rebucci. We study optimal policy in an small open economy in which a foreign borrowing constraint binds only occasionally and a financial crisis is an endogenous event. In this environment, the scope for policy arises because of a pecuniary externality stemming from the presence of a key relative price in the borrowing constraint. We study how optimal policy should be set. We first examine the merit of alternative policy tools. In the endowment case, we compare exchange rate policy and controls on capital flows and find that exchange rate policy dominates capital controls since it replicates the unconstrained first best allocation while capital controls achieve only the constrained efficient one. In this special endowment case, optimal policy (in the Ramsey sense) is time consistent and replicates the social planner allocation. We then examine the case of a production economy and compute the time-consistent optimal policy for alternative instruments confirming that, even in this more realistic case, exchange rate policy dominates capital controls but with the important difference that, when used individually, neither of these policies can replicate the constrained efficient outcome. For this reason, we examine the optimal use of the two policy instrument. A methodological contribution of the paper is the development of computational algorithms to solve optimal policy problems in environments with constraints that bind only occasionally.

"Portfolio Choice with Information-Processing Limits" with Alta Batchuluun and Yulei Luo. This paper studies the portfolio choice of an agent with limited information processing ability. Very preliminary.