Optimal v. simple financial policy rules in a production economy with “liability dollarization”

Abstract

We evaluate the effectiveness of financial policy rules in a small open economy with production, liability dollarization and "unconventional shocks" (global liquidity shifts and news about future fundamentals). Tradable and nontradable final goods are produced with tradable inputs. Debt is denominated in units of tradables and cannot exceed a fraction of the market value of total income. Optimal policy has a macroprudential or ex-ante component (a debt tax levied at date t only when the credit constraint may bind at t + 1), and ex-post components (sectoral production taxes/subsidies used when the constraint binds). The optimal policy reduces sharply the frequency and severity of financial crises but is also very complex. Simple policies are less effective and can be welfare reducing.

Macroprudential Policy: Promise and Challenges

“…policy that applies when μ * t >0). For instance, in a liability dollarization model with production, the planner would like to reallocate inputs from nontradables to tradables production when μ * t >0, because this props up the value of collateral and makes the constraint less binding (see Hernandez and Mendoza (2016)). Fourth, if collateral values at date t are determined jointly by date-t and date-t+1 allocations, the optimal plans of the social planner can be time-inconsistent under commitment (see ).…”

“…Second, these models yield very favorable results about the effectiveness of optimal macroprudential policy, because of large pecuniary externalities. The specific formulation of the liability dollarization model is based on Hernandez and Mendoza (2016), which in turn follows from .…”

“…Quantitative Findings Hernandez and Mendoza (2016) calibrate the model using data for Colombia (previous applications of the liability dollarization model have used data for Mexico and Argentina). The parameter values are listed in Table 1, and the details of the calibration are available in their paper.…”

“…Debt taxes are used more frequently, with a long-run probability of almost 12 percent, while the long-run probability of using the production tax and subsidy, which are states in which μ t >0, is roughly 5 percent. Hernandez and Mendoza (2016) and document in detail the complexity of the optimal policies by studying the variation of the optimal taxes both in the time-series dimension and in the schedules of taxes across values of debt and realizations of the various shocks. Figure 2 illustrates some of this complexity by showing the optimal policy schedules as functions of the value of b for regimes with high (blue, continuous curves) and low (green, dashed curves) world interest rates in pairs of panels for each tax that correspond to bad news and good news in s, both coinciding with bad realizations of z.…”

Positive and Normative Implications of Liability Dollarization for Sudden Stops Models of Macroprudential Policy

“…33, Bank for International Settlements, January 2009.2 This framework originated in the seminal articles bySalter (1959),Swan (1960), and Díaz-Alejandro (1965).3 This setup originates in the work ofMendoza (2002). Studies that explore the models' normative implications, and in particular the implications for macroprudential policy, include Bianchi (2011),Benigno et al (2016),Korinek (2011), Schmitt-Grohé and Uribe (2017),Bianchi et al (2016), andHernández and Mendoza (2017).…”