Time series analyses of inflation in New Zealand : a thesis presented in partial fulfilment of the requirements for the degree of Master of Applied Statistics at Massey University

Thumbnail Image
Open Access Location
Journal Title
Journal ISSN
Volume Title
Massey University
The Author
Modelling of the economy has become increasingly important over the years. It serves two main purposes. It enables forecasts and it can be used for the evaluation of various economic policies. Economic models come with various degrees of size and statistical complexity. Models can be of a qualitative or of a quantitative nature. The soundness of the statistical techniques that are used for quantitative models is critical. In recent years a number of such techniques have been developed. This thesis will evaluate some on existing economic New Zealand time series. Inflation plays a main role in everyday life and it has been of major ongoing concern to the New Zealand governments in recent times. These governments have instructed the Reserve Bank of New Zealand (RBNZ) to set monetary policies to ensure certain targets are met. The RBNZ achieves this to a large degree by setting the Official Cash Rate which is the major determinant of the interest rates that are used by the banks. This thesis will consider some theoretical aspects of time series analysis. In particular the Dickey-Fuller tests and cointegration analysis are considered. Also some theoretical aspects of inflation are considered. Examples are given of aspects of New Zealand life other than the interest rates that may also affect the current inflation rates. The time series that were analysed could be categorised as inflation indices, monetary aggregates, interest rates and gross domestic product. The thesis attempted to evaluate the time series in such a manner that there was little room for an analyst's bias. This was mainly achieved by developing a standardised approach to the analysis of these series. A number of interactions between the time series were evaluated and some were identified as being suitable for further research with the ultimate aim of developing a small model of the New Zealand economy. Another aim was to evaluate some aspects of economic policy where possible given the small number of time series that were used. Granger Causality tests seemed to show the effect of economic policy, where the interest rates affect the inflation rates. However, this was not further supported by cointegration analyses. There are various possible explanations for this. It was surmised that the standardised way of analysis may not have identified this relationship where it existed. The analyses showed that at times the results of the statistical tests were inconsistent. This applied to the Dickey-Fuller tests as well as the cointegration analyses. In some cases unit root models with significant coefficients for the deterministic components were identified. Further analysis would show that the deterministic components were not significant after all. However, the resulting models without these components did not have a unit root. The cointegration analyses invariably showed a number of Vector Error Correction Models with significant cointegration equations. Since their economic implications would be quite different at times there was a reason for concern. In conclusion there are some worrying problems when the methodology is used for existing short New Zealand data series. However, at times some plausible results were shown as well. Suggestions for further research were made.
Econometric models, Inflation (Finance), New Zealand