A negative interest rate on reserves is effectively a tax on the banking system. The rationale for adopting such policy in the context of a classic liquidity trap is its expansionary effect, via the reduction in the incentive to save. However, critics argue that the distortion in the banking system may well have adverse effects which outweigh the welfare gains from economic stimulus. This paper contributes to the debate by modelling the banking sector in accordance with the maturity-transformation framework of Diamond and Dybvig (1983). I show that, on the one hand, negative interest on reserves leads to financial disintermediation, which reduces the extent of maturity transformation below the first-best level; while, on the other hand, it is expansionary, as consumers save less overall. The paper's main finding is that, while in normal times taxing banks is inefficient, optimal monetary policy in the liquidity trap prescribes a strictly negative interest on reserves.
IMF Working Paper, February 2016 Read online
Structural reforms in the liquidity trap need not be deflationary. This paper develops a simple framework to study the role that key characteristics of Japan’s labor and product markets—labor-market duality and weak corporate governance—play
in generating unfavorable wage-price dynamics. The model allows a discussion of whether and in what form structural reforms may contribute to Japan’s short-run goal of reflating the economy. It finds that boosting inflation with structural
reforms implies an unusual trade-off with employment, that is an inverted Phillips curve. Simultaneous implementation of labor-market and product-market reforms is most effective in terms of reflating the economy.
A well-known result of the Diamond-Dybvig model is that the existence of financial markets curtails provision of liquidity-risk insurance by the banking system. The model’s equilibrium allocation has inefficiently low investment in short-term assets. Farhi et al. (2009) propose imposing a reserve requirement on banks in order to increase investment in short-term assets. In a model where consumers are restricted to investing their wealth in bank deposits, they show that a reserve requirement can implement the efficient allocation. First, I show that the fully efficient allocation can equivalently be implemented by payment of positive interest on reserves. Then, I extend the model to allow consumers to invest their wealth directly in illiquid capital assets as well as in bank deposits. Such generalisation breaks down the equivalence between reserve requirement and interest on reserves. A reserve requirement does not lead to more aggregate investment in short-term assets, because it is effectively a tax on bank deposits and consumers respond to it by reducing their holdings of deposits. The paper’s main result is that, in the generalised setting, the efficient level of liquidity-risk insurance can only be implemented by a policy of paying strictly positive interest on bank reserves. Interest on reserves provides an incentive for banks to hold more short-term assets and does not lead to disintermediation.
Work in Progress
“Balance-Sheet Policies and Financial Stability”
In a standard model of financial intermediation, I show that the size of the central bank's balance sheet has an effect on the term premium. An increase in the supply of bank reserves reduces the term premium, as it pushes up the relative supply of short-term assets. Following the global-games literature with equilibrium bank runs, I show that a large central-bank balance sheet reduces the probabilty of runs. This is because a lower term premium reduces the incentive for financial intermediaries to have a maturity mismatch in their balance sheets. Ultimately, I show that optimal monetary policy prescribes a large balance sheet for the central bank.