Essays on Inflation Dynamics: A New Keynesian Perspective

從新凱恩斯學派角度的通貨膨脹論文

Student thesis: Master's Thesis

View graph of relations

Author(s)

Related Research Unit(s)

Detail(s)

Awarding Institution
Supervisors/Advisors
Award date22 May 2019

Abstract

This thesis focuses on the explanation of inflation dynamics with the help of different models and methods. Specifically, the validity of the New Keynesian Phillips Curve will be investigated based on two major arguments with reference to the driver of inflation dynamics: the forward-looking and the backward-looking price adjustment behavior. This study attempts to compare across models to investigate which of these behaviors should be the most important determinant of inflation. Drawing on the Hong Kong Quarterly Data 1982Q1-2017Q2, it is found: (1) The pure-backward looking model explains the inflation dynamics better than the pure-forward looking one; (2) Firms take longer time to complete the price adjustment in the backward-looking model (3) The backward-looking feature of the model underestimates inflation fluctuation during the period 1994Q2 to 2017Q2, while the forward-looking feature overestimates it. To the best of our knowledge, our findings are different from most previous studies, which favour the forward-looking model. The Backward-looking model outperforms the forward-looking one in terms of lower measurement errors, thus making it more important to consider it in explaining the inflation dynamics in Hong Kong. In addition, this study also provides greater research possibilities in terms of extending it to the Bayesian dynamics stochastics general equilibrium (DSGE) New Keynesian Phillips Curve. The empirical results match with the macroeconomics theory. Some other research possibilities are also provided in the last section of this thesis such as extending to Asian regions, inflation expectation pass through, zero lower bound, the Behavioural New Keynesian Model and comparing the menu cost model with the Sticky Price model and the Sticky Information model.