Essays on financial externalities

This thesis explores some of the trade-offs faced by policymakers in trying to prevent or moderate the impact of financial externalities generating instability in the macroeconomy. The first chapter explores the role of cash flow constraints combined with lower equilibrium interest rates in inducing less productive firms (zombies) to invest and produce. Zombie firms generate a negative spillover on the borrowing capacity of more productive firms: by demanding capital they contribute to raising wages, reducing the value of profits for all firms and further tightening the borrowing constraint of productive firms. If the interest rate hits the effective lower bound however, aggregate demand is low, fewer low-productivity firms invest and liquidating zombie firms can be counterproductive. At the lower bound, these firms are not zombies but make use of idle resources, boosting output and welfare. The second chapter focuses on the key role played by collateral assets in determining the distribution of productive capital across heterogeneous producers, as well as in inducing business cycle amplifications. From a prudential point of view, moderating firms' access to credit is helpful in avoiding fire sale externalities in a financial crisis; however, this impairs the distribution of productive capital. From a normative perspective, the use of capital requirements can help implementing the constrained effcient allocation, provided that the regulator has the ability to commit to future policies. The last chapter is a co-authored work that explores the link between aggregate demand externalities and housing tenure choices. Shocks that induce households to deleverage sharply can push the policy rate at its lower bound, at which point output is demand-determined. Restricting access to mortgages is therefore beneficial in preventing this externality; however, it distorts housing choices when households have a preference for owning as opposed to renting. Macroprudential interventions have important distributional consequences, which are explored in a version of the model calibrated to match the UK data