Research

Geopolitical risk cannot be measured in a universal way. We develop new geopolitical risk indicators relying on local newspaper coverage to account for different perceptions. Using Russia as a case study, we demonstrate that local geopolitical risk shocks have significant adverse effects on the Russian economy, whereas geopolitical risk shocks identified from English-language news sources do not. We control for restricted press freedom by analyzing state-controlled and independent media separately. Employing a novel Russian sanctions index, we illustrate that geopolitical risk shocks propagate beyond the sanctions channel. Still, sanctions worsen the inflationary impact of geopolitical risk shocks substantially. 

Could a monetary policy loosening in a low interest rate environment have unintended recessionary effects?  Using a non-linear macroeconomic model fitted to the euro area economy, we show that the effectiveness of monetary policy can decline in negative territory until it reaches a turning point, where monetary policy becomes contractionary. The framework demonstrates that the risk of hitting the rate at which the effect reverses depends on the capitalization of the banking sector. The possibility of the reversal interest rate gives rise to a novel motive for macroprudential policy. We show that macroprudential policy in the form of a countercyclical capital buffer, which prescribes the build-up of buffers in good times, substantially mitigates the probability of encountering the reversal rate and increases the effectiveness of negative interest rate policies. This new motive emphasizes the strategic complementarities between monetary policy and macroprudential policy.

Motivated by the build-up of shadow bank leverage prior to the financial crisis of 2007-2008, I develop a nonlinear macroeconomic model featuring excessive leverage accumulation and   endogenous runs to capture the dynamics and quantify the build-up of instability. Incorporating monetary policy, I demonstrate that the zero lower bound increases the crises frequency and lowers welfare. The model is taken to U.S. data to estimate the run probability around the financial crisis of 2007-2008. The estimated run risk was already considerable in 2005 and kept increasing. Counterfactual simulations evaluate whether monetary interventions boost welfare and could have averted the financial crisis.

We study contact tracing in a new macro-epidemiological model with asymptomatic transmission and limited testing capacity. Contact tracing is a testing strategy that aims to reconstruct the infection chain of newly symptomatic agents. This strategy may be unsuccessful because of an externality leading agents to expand their interactions at rates exceeding policymakers' ability to test all the traced contacts. Complementing contact tracing with timely-deployed containment measures (e.g., social distancing or a tighter quarantine policy) corrects this externality and delivers outcomes that are remarkably similar to the benchmark case where tests are unlimited. We provide theoretical underpinnings to the risk of becoming infected in macro-epidemiological models. Our methodology to reconstruct infection chains is not affected by curse-of-dimensionality problems.

Since the 2001 recession, average core inflation has been below the Federal Reserve’s 2% target. This deflationary bias is a predictable consequence of a symmetric monetary policy strategy that fails to recognize the risk of encountering the zero-lower-bound. An asymmetric rule according to which the central bank responds less aggressively to above-target inflation corrects the bias, improves welfare, and reduces the risk of deflationary spirals – a pathological situation in which inflation keeps falling indefinitely. This approach does not entail any history dependence or commitment to overshoot the inflation target and can be implemented with an asymmetric target range. A counterfactual simulation shows that a modest level of asymmetry would have removed the deflationary bias observed in the United States. 

We leverage recent advancements in machine learning to develop an integrated method to solve globally and estimate models featuring agent heterogeneity, nonlinear constraints, and aggregate uncertainty. Using simulated data, we show that the proposed method accurately estimates the parameters of a nonlinear Heterogeneous Agent New Keynesian (HANK) model with a zero lower bound (ZLB) constraint. We further apply our method to estimate this HANK model using U.S. data. In the estimated model, the interaction between the ZLB constraint and idiosyncratic income risks emerges as a key source of aggregate output volatility.

How does a shift to ambitious climate policy affect financial stability? We develop a quantitative macroeconomic model with carbon taxes and endogenous financial crises to study so-called ``Climate Minsky Moments''. By reducing asset returns, an accelerated transition to net zero initially elevates the crisis probability substantially. However, carbon taxes enhance long-run financial stability by diminishing the relative size of the financial sector. Quantitatively, the net financial stability effect is only negative for higher social discount rates. Even then, the welfare effects of ``Climate Minsky Moments'' are, at most, second-order relative to the real costs and benefits of an accelerated transition.

This paper studies the distributional effects of monetary policy and its dependence on inflation. We document a novel dependency in the earnings heterogeneity channel of monetary policy using high-frequency, administrative tax data from eurozone member Estonia. Monetary policy shocks substantially influence earnings inequality during high-inflation periods, with weaker effects during low-inflation periods. Extending our dataset with granular MPC estimates, we show that earnings heterogeneity amplifies the aggregate MPC and consumption response. In high-inflation periods, consumption and inequality respond more, even though the aggregate MPC may be lower. We rationalise our findings with a nonlinear tractable HANK model featuring inflation dependencies.

We examine the impact of a central bank digital currency (CBDC) on the banking system, combining survey evidence from German households with a macroeconomic model featuring endogenous bank runs. The survey reveals non-trivial demand for CBDC as a substitute for bank deposits in normal times (“slow disintermediation”) and increased withdrawal risks during financial distress (“fast disintermediation”). Informed by the survey, the model indicates that naively introducing a CBDC might reduce financial stability because CBDC offers storage at scale - making it attractive to run to. We estimate an optimal holding limit which chokes off fast disintermediation and actually enhances financial stability by shrinking a fragile banking system.

We examine how private sector agents might learn a new monetary strategy in a stochastic environment in which large inflationary and deflationary shocks may occur unexpectedly. We consider a central bank that adopts a new asymmetric average inflation targeting rule aimed at countering the disinflationary bias imparted by the effective lower bound (ELB). The new rule is announced while at the ELB.  The most crucial time for learning runs from when rates would be near liftoff under the old strategy through early liftoff under the new rule.  Recessionary shocks during this time could delay learning while large inflationary shocks could outright stop it, inhibiting the ability of the new strategy to address the costs associated with the ELB.  A large inflation surprise may thus give rise to a novel policy dilemma: The central bank could offset some of the inflation-induced learning loss by being even more accommodative than its new strategy, but this decision comes at the cost of higher near-term inflation and greater uncertainty about monetary policy.