The Treasury

Global Navigation

Personal tools

1  Introduction

1.1  Motivation

This paper compares the dynamic properties of the New Zealand Treasury model (NZTM) and the current version of the Reserve Bank of New Zealand’s (RBNZ) Forecasting and Policy System (FPS) model. The main use of the two models is for producing macroeconomic forecasts. The NZTM model is used to produce forecasts that are used as an input into the final forecast numbers presented in the Economic and Fiscal Updates. The FPS model is used to produce the published forecasts in the Reserve Bank’s quarterly Monetary Policy Statement. Both models involve a number of differing judgements concerning the structure of the economy, the important shocks that impact on the economy, and how the economy responds to these shocks. These judgements provide us with a basis for comparison between the two models.

The New Zealand Treasury has been through a period of considerable macroeconomic model development over the last few years. Following an internal review of the existing New Zealand Model (NZM) in 2000, the Treasury decided to make a number of improvements on the existing model structure. The outcome was the development of a new model structure called the New Zealand Treasury Model (NZTM), which is documented in Szeto (2001 and 2002).

The RBNZ which first started using the FPS model in 1997, is documented in Black, Cassino, Drew, Hansen, Hunt, Rose and Scott (1997). This model has been used to produce the Reserve Bank’s forecasts since that time. During the process of using the model to produce forecasts and research, a number of changes have also been made to FPS.

An early attempt at comparing these models (Drew and Hunt, 2000) was undertaken before the extensive redevelopment of the Treasury model. The current comparison changes their approach in two key areas. First, the response functions for monetary and fiscal policy are not calibrated to be the same for the two models. Second, the shocks investigated are all temporary. We have not attempted to compare the steady states of the two models, which are very similar. Rather, we focussed on the way the business cycle evolves following a temporary shock.

The results indicate that while the NZTM has some similarities with the FPS, there are important differences. These result in different adjustment paths following various shocks.

The remainder of the paper is organised as follows: Section 2 contains a brief overview of the two models. In section 3, the results from the shocks are presented. Section 4 contains a summary and future work is described in section 5.

Page top