Essays on Information-processing Constraints and Monetary Policy Transmission = 정보처리 제약과 통화정책 전달 메커니즘에 관한 연구

다국어 초록 (Multilingual Abstract)

This dissertation consists of two articles on information-processing constraints in macroeconomic decision-making: one on firm price-setting under heterogeneous information capacity, and one on household consumption and portfolio choice under dynamic Rational Inattention.

Chapter 1 extends the standard Rational Inattention (RI) framework by introducing two types of firms—those with high information-processing capacity (High-type) and those with low capacity (Low-type)—to reexamine how macroeconomic shocks influence price-setting behavior and inflation dynamics. High-capacity firms adjust prices even under low shock volatility, whereas low-capacity firms only respond when volatility surpasses a certain threshold. This generates a discontinuous kink between smooth and sharp inflation responses, a feature amplified by strategic complementarity across firms. The model reveals that under low uncertainty, inflation volatility remains limited, but once the threshold is crossed, it expands explosively. Only in this high-volatility regime does central bank information provision meaningfully contribute to inflation stabilization. These findings imply that in environments characterized by heterogeneous information capacity and strategic interaction, price rigidity and the effectiveness of monetary policy become nonlinear and asymmetric. Consequently, central banks should design differentiated communication strategies that account for the distribution of firms’ information-processing capacities.

Chapter 2 develops a dynamic model of consumption and portfolio choice under limited attention in a finite-horizon setting. The consumer selects, at each period, the consumption share, portfolio allocation, and information signals endogenously, subject to an information constraint formalized via Rational Inattention based on mutual information. The model is approximated around a deterministic path using second-order expansion, reducing the optimization problem to a linear-quadratic (LQ) structure. Within this framework, we analyze the optimal response functions and the endogenous structure of signals. In particular, we show theoretically that when the timing of policy announcements—which occur with endogenous probability—is aligned with consumers' attention, inefficiencies in market reactions are minimized. By identifying the interaction between information selection and policy timing, the paper offers a novel analytical framework for modeling consumption and asset allocation behavior grounded in dynamic Rational Inattention theory.

국문 초록 (Abstract)

본 논문은 거시경제 환경에서의 정보처리 제약이 경제주체의 의사결정에 미치는 영향을 분석한 두 편의 연구로 구성되어 있다. 첫 번째 논문은 정보처리 능력이 이질적인 기업들의 가격 설정 행태를 이론적으로 분석하며, 두 번째 논문은 소비자들의 소비 및 포트폴리오 선택이 동태적 합리적 부주의(Rational Inattention) 하에서 어떻게 이루어지는지를 탐구한다.

제1장에서는 표준적인 합리적 부주의(RI) 모형을 확장하여, 정보처리 능력이 높은 기업(High-type)과 낮은 기업(Low-type)의 이질적 구조를 도입한다. 정보처리 능력이 높은 기업은 낮은 충격 변동성 하에서도 가격을 조정하지만, 낮은 능력을 가진 기업은 변동성이 특정 임계값을 넘어서야만 반응한다. 이로 인해 물가 반응은 연속적이지 않고, 완만한 반응과 급격한 반응 사이에 불연속적인 전환이 나타난다. 이러한 비선형성은 기업 간 전략적 상호작용에 의해 더욱 증폭된다. 본 모형은 변동성이 낮을 때는 인플레이션 변동성도 제한되지만, 임계값을 초과하면 급격히 확산됨을 보여준다. 이와 같은 고변동성 구간에서만 중앙은행의 정보 제공이 실질적으로 인플레이션 안정화에 기여할 수 있다. 따라서 정보처리 능력이 이질적인 기업 간 전략적 상호작용이 존재할 경우, 통화정책의 효과와 가격 경직성은 비선형적이고 비대칭적으로 작동하며, 이에 따라 중앙은행은 정보처리 능력의 분포를 고려한 차별적 커뮤니케이션 전략을 설계해야 한다.

제2장에서는 유한기간 환경에서 소비자가 제한된 주의 하에 소비 비중, 자산 포트폴리오 구성, 정보 신호를 내생적으로 선택하는 동태적 모형을 구축한다. 정보 제약은 상호정보량(mutual information)을 기반으로 한 합리적 부주의 제약을 통해 표현되며, 모형은 결정론적 경로 주변에서 2차 근사를 적용하여 선형-제곱형(LQ) 구조로 단순화된다. 이 틀 내에서 최적 반응 함수와 내생적 정보 구조를 이론적으로 도출한다. 특히 정책 발표 시점이 내생적 확률로 발생하고, 그 시점이 소비자의 주의와 정렬될 경우 시장 반응의 비효율성이 최소화됨을 보인다. 본 논문은 정보 선택과 정책 시점 간의 상호작용을 명시적으로 모델링함으로써, 정보 제약 하의 소비 및 자산배분 의사결정을 분석하는 새로운 이론적 틀을 제시한다.

목차 (Table of Contents)

• Contents
• Abstract 1
• Chapter 1: Information Heterogeneity and Strategic Complementarity in Non-Linear Price Dynamics: A Two-Agent Rational Inattention Model 3
• 1.1 Introduction 3
• 1.2 Overview 5
• 1.3 Basic Rational Inattention Framework 8
• 1.3.1 Framework of the Standard Rational Inattention Literature 8
• 1.3.2 Basic Information Theory for RI Literature 9
• 1.3.3 Two Modelling Approaches to Information Constraints 12
• 1.3.4 Connections Between the Two Models 14
• 1.4 Model with Heterogeneous Information Capacity 15
• 1.4.1 Basic Model Setup 15
• 1.4.2 Quadratic Approximation of the Objective Function and Optimal Behavior 18
• 1.4.3 Significance of Mackowiak and Wiederholt (2009) as a Benchmark Model 18
• 1.4.4 Extension to a Two-Group Model with Heterogeneous Information Capacities 21
• 1.4.5 Definition of Equilibrium 21
• 1.4.6 Mechanism Induced by the Information Constraint 28
• 1.4.7 Nonlinear Response with Respect to the Relative Variance Ratio 30
• 1.4.8 Kink Points and Nonlinearity from Strategic Interaction 32
• 1.5 The Central Bank’s Optimization Problem under Public Information Provision 42
• 1.5.1 Objective Function and Policy Instruments 42
• 1.5.2 Piecewise Definition of the Price Responsiveness Function and Reduction of the Optimization Domain 43
• 1.5.3 Simplification of the Objective Function and Conditions for Existence of Solution 44
• 1.5.4 Verification of Optimal Intervention Conditions and Updating of the Response Function 45
• 1.5.5 Optimization and Solution Derivation in the RLH Region 46
• 1.5.6 Verification of Validity and Update of the Reaction Function in the RLH Region 47
• 1.5.7 Visualization of the Reaction Function Before and After Policy Intervention 48
• 1.5.8 Limitations of Information Interventions and the Role of Information Costs 48
• 1.5.9 Structural Limits of Information Intervention and the Lower Bound of Price Responsiveness 50
• 1.5.10 Structural Limits of Policy Intervention under Strategic Complementarity 51
• 1.5.11 Implications of Nonlinear Policy Effects and the Endogeneity of Information Structure 52
• 1.5.12 Structural Regime Shift in Dynamic Responses under Varying Information Constraints 53
• 1.6 Conclusion 55
• A Chapter 1. Appendix 56
• A.1 Proof of Information Flow Using Szegő’s Theorem 56
• A.2 Proof of the Equivalence Between Two Static Models 58
• A.3 Derivation 60
• A.4 Boundary Condition Differentiation 67
• Chapter 2: Endogenous Shock Bundling and Optimal Announcement Timing in a Rational Inattention Framework 73
• 2.1 Introduction 73
• 2.2 Finite-Horizon Consumption and Portfolio Share Optimization under Information Constraints 74
• 2.2.1 Basic Model Setup 74
• 2.2.2 Construction of Information Sets and the Full-Information Benchmark 76
• 2.2.3 Benchmark for LQ Approximation 77
• 2.2.4 Implementing the Benchmark with Embedded Uncertainty 81
• 2.3 Model Extension under Rational Inattention 84
• 2.3.1 LQ Approximation and Incorporation of Information Constraints under Rational Inattention 84
• 2.3.2 Quantifying Uncertainty and Information Using Mutual Information 85
• 2.3.3 Solving the Finite-Horizon Consumption–Portfolio Choice Model 86
• 2.3.4 Construction of the LQG Model Under Gaussian Assumptions 91
• 2.3.5 Uniqueness of Impulse Response Functions (IRFs) 97
• 2.4 Application to Variations in Information Variables 101
• 2.4.1 Effect of Uncertainty about Announcement Timing on IRFs 101
• 2.4.2 Optimal Information Acquisition under Uncertain States 102
• 2.4.3 Policy Effects at Announcement Timing 106
• 2.4.4 Optimal Probability and Timing for Treasury Yield Variance 107
• 2.4.5 Interpretation via a Simple Example 108
• 2.4.6 Optimal Government Announcement Timing 109
• 2.5 Conclusion 111
• B Chapter 2. Appendix 112
• B.1 2.1 Proof of the Myopic Expectation Solution (MES) 112
• B.2 2.2 Equivalence of Optimal Consumption–Portfolio Strategies under Perfect Foresight and Partial Information 113
• B.3 2.3 Fixed–Dimensional Signal Selection via Perfect Wealth Information 115
• B.3.1 Shock-Variable Dimensionality Increases Proportionally over Time 115
• B.3.2 Assumption for Time-Invariant Dimensionality of Shock Variables in the Signal 115
• B.3.3 Separation of the Optimization Problem by Signal 116
• B.4 2.4 Eigenstructure-Based Signal Design under Information Constraints 120
• B.5 2.5 Threshold for Single- versus Dual-Signal Acquisition under Two-Dimensional Information Constraints 124
• B.6 2.6 Equivalent IRF Condition: Uniform Posterior Gain in Optimal Signal Means 126
• B.7 2.7 Closed-Form Characterization of Optimal Rotation Angle and Information Allocation 126
• B.8 2.8 Rotation Angle for Diagonalizing a 2×2 Symmetric Matrix 129
• B.9 2.9 Characterization of Optimal Rotation Angle and Objective via Prior Covariance Elements 130
• B.10 2.10 Optimal Rotation Angle and Information Allocation in the Fixed–Cost Model 131
• B.11 2.11 Equivalence of the Interior Solution to a Single Weighted Prior Covariance Matrix 132
• B.12 2.12 Equivalence of the Interior Solution to a Composite Weighted Prior Covariance 134
• B.13 2.13 Optimal Announcement Probability and Delay Timing for Treasury-Yield Variance 135
• 국문초록 141