We assess gains from parallel computation on Backlight supercomputer. We find that information transfers are expensive. To make parallel computation efficient, a task per core must be sufficiently large, ranging from few seconds to one minute depending on the number of cores employed. For small...

In a Q-learning model-free setting, the mortgage servicer can learn the default incentive of the borrower from soft information and responsiveness during communication; and in turn undertake appropriate actions to maximize her reward. This reward maximization incentivizes the effort of the...

This paper provides an under-the-hatch look at the monetary-policy transmission mechanism at work in a two-agent new Keynesian (TANK) model. We compare the transmission of anticipated, future monetary-policy shocks with contemporaneous shocks by decomposing consumption responses into direct and...

We present a systematic treatment of dynamic e⁄ects of Odyssean forward guidance explicit announcements about future policy rates. We focus on normal times unlike the mainstream of the literature that focuses on episodes of an e⁄ective (zero) lower bound on nominal interest rates. We present...

We assess gains from parallel computation on Backlight supercomputer. The information transfers are expensive. We find that to make parallel computation efficient, a task per core must be sufficiently large, ranging from few seconds to one minute depending on the number of cores employed. For...