2014年-IMF国际货币组织全球_Introducing_a_Semi_37页_988kb

Summary of "Introducing a Semi-Structural Macroeconomic Model for Rwanda"

Core Content

This paper introduces a semi-structural macroeconomic model for Rwanda, aimed at improving the understanding of the monetary policy transmission mechanism. The model is based on a rational expectations New-Keynesian framework, which links monetary policy to economic activity and inflation. It is calibrated to reflect the stylized facts of Rwanda's economy and is used to filter observed data and produce forecasts.

Main Objectives

• To analyze the factors influencing inflation dynamics in Rwanda.
• To assess the impact of domestic and external shocks on inflation.
• To evaluate the effectiveness of monetary policy in maintaining inflation within target levels.
• To provide insights into the conduct of monetary policy by the National Bank of Rwanda (NBR) in response to these shocks.

Key Features of the Model

• The model includes four blocks: aggregate demand, aggregate supply, exchange rate determination, and interest rate policy.
• A modified uncovered interest parity (UIP) condition is introduced to capture the limited capital mobility and dual nominal anchors (inflation and exchange rate) in Rwanda.
• The model incorporates autoregressive components to better match the properties of the data.
• It distinguishes between core inflation, food inflation, and oil inflation, and defines headline inflation as a weighted average of these components.
• The model is used to perform shock decomposition and forecasting exercises to understand the behavior of key macroeconomic variables.

Main Findings

• Food and oil price shocks account for the majority of inflation fluctuations in Rwanda, especially during 2008 and 2011.
• Food prices have a greater impact on inflation than oil prices, due to partial pass-through from international to domestic prices and administrative controls.
• Exchange rate movements significantly affect inflation, particularly during periods of foreign aid suspension in 2012, leading to a depreciation of over 12%.
• The NBR's monetary policy stance was too accommodative in 2008 and 2011, contributing to inflation deviating from its 5% target.
• The sacrifice ratio for headline inflation is estimated at 2.0, while the model predicts a sacrifice ratio of 1.8, indicating a reasonable fit.
• The model is robust even in the presence of noisy or scarce data and underdeveloped financial markets.

Structure of the Model

I. Aggregate Demand

• The output gap is modeled as a function of the real monetary conditions index (rmci), the output gap from the US, and a domestic demand shock.
• The real monetary conditions index is a weighted average of the real interest rate gap and the real exchange rate gap.

II. Aggregate Supply

• Core inflation is influenced by past inflation, inflation expectations, and real marginal costs.
• Food and oil inflation are also modeled as functions of past inflation, expectations, and external price shocks.
• The real marginal costs are a function of the real exchange rate gap and the output gap.

III. Exchange Rate Determination

• A modified UIP condition is used to model the nominal exchange rate dynamics, taking into account capital mobility and monetary policy.
• The target exchange rate is determined by a long-run relative purchasing power parity adjusted for real exchange rate trends.
• The exchange rate corridor system introduced in 2010 reflects the NBR's active management of the exchange rate.

IV. Interest Rate Policy Rule

• The central bank sets the interest rate in response to inflation deviations from the target and the output gap.
• The neutral real interest rate and the neutral nominal interest rate are modeled based on long-run equilibrium values.
• A stochastic process is used to model the inflation target, allowing for different disinflation paths.

V. Long-Run Trends

• The long-run values of real interest rates, output growth, and real exchange rate changes are modeled as first-order autoregressive processes.
• These trends are based on steady-state values derived from the last six years of data.

Calibration and Data

• The model is calibrated using data from 2003Q1 to 2013Q3.
• The CPI basket is divided into core, food, and oil components, with food accounting for 35.4% and oil (transport) for 11.9%.
• The Hodrick-Prescott (HP) filter is used to smooth GDP data, and the X12-ARIMA filter is used for CPI data.
• The model is tested for goodness of fit and forecast accuracy, both in-sample and out-of-sample.

Policy Implications

• The model helps in identifying the sources of inflation and assessing the effectiveness of monetary policy.
• It highlights the importance of exchange rate management in the context of limited capital mobility and dual policy targets.
• The findings support the need for greater flexibility in the monetary policy framework and improved communication with market participants to enhance policy credibility and transmission efficiency.

Conclusion

The paper contributes to the understanding of monetary policy transmission in Rwanda by introducing a semi-structural model that accounts for domestic and external factors. The model provides valuable insights into the impact of shocks on inflation and the conduct of monetary policy, which is essential for policy formulation and economic forecasting.