Seokil Kang

I am an Economist at the Bank of Korea. I received my Ph.D. in Economics from the Indiana University Bloomington in 2022.

Field of Interest: Macroeconomics, Monetary and fiscal policy, Bayesian Econometrics

Disclaimer: This site does not express the views of the Bank of Korea.

Curriculum Vitae (PDF)

♦ Click title to see abstract

Working Papers

Quantifying the Fiscal Backing for Monetary Policy, Paper(pdf), Codes

Abstract
I ask to what extent can data reveal whether fiscal policy responses to monetary policy shock are consistent with the theoretical adjustments necessary for successful inflation-targeting monetary policy. I employ a DSGE model to estimate the fiscal response to a monetary policy shock under the active monetary and passive fiscal policy regime. A monetary contraction raising interest rate by 25 basis points reduces the market value of government debt by 0.8% because the bond price devaluation outweighs the fall in inflation. This reduction splits into a 1.7% decline due to higher discount rates and a 0.9% increase in expected primary surpluses. I also estimate a VAR that takes an agnostic view on the policy regime to examine how closely the data conforms to the theory. I find that the data accounts for 90% of the primary surplus response dictated by theory, suggesting that the data reveals the presence of fiscal backing for monetary policy.

Abstract

I ask to what extent can data reveal whether fiscal policy responses to monetary policy shock are consistent with the theoretical adjustments necessary for successful inflation-targeting monetary policy. I employ a DSGE model to estimate the fiscal response to a monetary policy shock under the active monetary and passive fiscal policy regime. A monetary contraction raising interest rate by 25 basis points reduces the market value of government debt by 0.8% because the bond price devaluation outweighs the fall in inflation. This reduction splits into a 1.7% decline due to higher discount rates and a 0.9% increase in expected primary surpluses. I also estimate a VAR that takes an agnostic view on the policy regime to examine how closely the data conforms to the theory. I find that the data accounts for 90% of the primary surplus response dictated by theory, suggesting that the data reveals the presence of fiscal backing for monetary policy.

Simulated Annealing Multiplicative Weights Algorithm for Solving a DSGE Model, Paper(pdf), Codes

Abstract
This paper introduces a simulation-based adaptive algorithm to solve a DSGE model with a large state space, namely the curse of dimensionality. It aims to generate a stationary distribution over policy space which is concentrated on the optimal policy. The key strategy is to construct a finite policy space of heuristic policies. To update the distribution over policy space, the method adopts on-line computation via iterative simulation with emphasis on rolling-horizon control to foster the speed of algorithm. Subsequently, I deliver that the algorithm achieves theoretical convergence to the optimal value function and the stationary distribution over policy space is concentrated on the optimal policy. Application to solve the simple two-period RBC model follows as a sample exercise. The result shows the performance is desirable within the feasible number of iterations and size of restricted policy space respectively.

Abstract

This paper introduces a simulation-based adaptive algorithm to solve a DSGE model with a large state space, namely the curse of dimensionality. It aims to generate a stationary distribution over policy space which is concentrated on the optimal policy. The key strategy is to construct a finite policy space of heuristic policies. To update the distribution over policy space, the method adopts on-line computation via iterative simulation with emphasis on rolling-horizon control to foster the speed of algorithm. Subsequently, I deliver that the algorithm achieves theoretical convergence to the optimal value function and the stationary distribution over policy space is concentrated on the optimal policy. Application to solve the simple two-period RBC model follows as a sample exercise. The result shows the performance is desirable within the feasible number of iterations and size of restricted policy space respectively.

♦ Click course title to see syllabus

Teaching

Associate Instructor (full teaching responsibilities)	Macroeconomics I(Master Course)	Spring 2021
	Statistical Analysis for Business and Economics	Spring 2019, 2020
	Intermediate Macroeconomics Theory	Fall 2018
	Method of Economic Analysis	Spring 2018
Teaching Assistant	Macroeconomics I(Ph.D. Course)	Fall 2019, 2020, 2021

Miscellaneous

Julia Replication for Arellano(AER 2008)