スーパーグローバル大学創成支援実証政治経済学拠点では、Junior Maih 教授（ノルウェー中央銀行、ノルウェービジネススクール）をお招きしてマクロ経済モデルのシミュレーションに関する全7回の特別ワークショップシリーズを開催します。参加条件、各回の詳細は下記をご参照ください。
- 講演者: Junior Maih 教授 (Norges Bank, Norwegian Business School)
- 日時: 2023年1月10日～18日, 14:45～16:15（90分 x 7回）
- 場所: Webinar と対面による開催（早稲田大学3号館)
- 日よって教室が変わりますのでご注意ください。
1月10日 早稲田大学3号館 3－909 1月11日 早稲田大学3号館 3－915 1月12日 早稲田大学3号館 3－305 1月13日 早稲田大学3号館 3－801 1月16日 早稲田大学3号館 3－302 1月17日 早稲田大学3号館 3－301 1月18日 早稲田大学3号館 3－302 - 参加条件: 大学院またはそれ以上の知識、動学的一般均衡モデル (DSGEモデル）またはそれを取り扱うRISEツールボックスの知識があれば望ましい。
- 推奨環境: Matlab (core matlab, optimization toolbox, statistics toolbox)
MikTex, dropbox, oxedit (optional) - 申込: 参加希望の旨を下記までメール連絡
mail to: [email protected] - 使用言語: English
- 参加料無料
Outline and Objectives:
This workshop is intended for participants with some considerable background knowledge of both DSGE modeling and RISE. The goal is to perform quantitative analysis within the confines of a theoretically coherent microfounded DSGE model and derive some policy implications. We will use some of the advanced features of RISE to parameterize the model efficiently and carry out various simulation exercises.
Much of the discussion will revolve around the modeling of regime switches and nonlinearities. Those features are ubiquitous in economic data and frequently discussed in policy circles. In a world with sometimes abrupt changes in uncertainty, breakdowns of economic structures, and policy shifts (e.g., unconventional policies), agents must consider the possibility of such events reoccurring.
The workshop is organized over 7 lectures, with each lecture consisting of both lectures and hands-on training. A particular emphasis will be put on research methodology. In other words, we will be conducting research, addressing the issue of” How effective is monetary policy” as if we were writing a research paper. At the end of the workshop, the course participants are expected not just to have an improved command of RISE but also an improved ability to organize and conduct effective research.
Planning of the Workshop (7 sessions x 90 minutes)
Lecture 1 (1/10,) and Lecture 2 (1/11)
The purpose of these lectures is to introduce the problem, discuss its relevance, overview some of the relevant literature on the subject, map out a plan for how we will tackle the problem, and construct a simple model to guide the analysis. At each step of the construction of the model, we will strive to understand the background theory, the purpose, and implications of the assumptions made as well as the intuition behind each simplification. Finally, we will review the basics of the RISE language.
Lecture 3 (1/12) and Lecture 4 (1/13 @Rm801)
On these days, we will review solution methods, and implement the model accounting for both heteroskedasticity and switches in the deep structural parameters. We will then go ahead and solve the regime-switching DSGE system using perturbation. The solution of the model will give us the opportunity to conduct some basic simulation exercises.
Lecture 5 (1/16) and Lecture 6 (1/17)
In these lectures, we will discuss various techniques for parameterizing1 our model. These include calibration, indirect inference, and likelihood-based estimation. Much of the emphasis will be on the choice of priors and Bayesian estimation, possibly with property priors2.
We will then use the acquired knowledge to calibrate some of the parameters of the model and estimate the remaining ones using Bayesian estimation techniques. After that, we will analyze the economic implications.
Lecture 7 (1/18)
In this lecture, we will focus on robustness analysis. We will, in particular, compute various measures of marginal data density and conduct model comparisons.
