3 Data and Preliminary Analysis

All data are quarterly and the sample period used in our SVAR analysis is 1982Q2 to 2011Q4, or a slightly shorter period when commodity price data are employed. As usual, economic growth is measured in terms of real GDP (seasonally adjusted) for each country. The GDP series, together with New Zealand's short-term (90 day) interest rate and the consumer price index (seasonally adjusted), are mainly downloaded from the Haver Database[6]; New Zealand's trade weighted real exchange rate series is obtained from the database of the New Zealand Treasury. Commodity prices relevant to New Zealand are captured through the nominal ANZ world commodity price series, which are also downloaded from Haver Analytics. These represent weighted average world prices for New Zealand’s major commodity exports, with both the overall index and the dairy, forestry, meat and wool, and aluminium sub-indices employed. The commodity price data are available from 1986. The nominal commodity price indices (expressed in US dollars) are seasonally adjusted using the X-11 method in EViews and deflated by the US consumer price index to obtain real commodity prices. For model estimation, all variables are expressed as quarterly growth rates (100 times the quarterly differences of log values) with the exception of interest rates which are in levels.

Appendix Table 1 provides information on the dataset employed in the analysis, including some descriptive statistics. The rapid expansion of China's economy is evident, as is the volatility of quarterly growth in China relative to the US, from the sample means and standard deviations, while the volatility of the various real commodity price inflation series is also notable.

3.1  Growth rate correlations

To provide initial insights into the international spillovers to New Zealand GDP and how these change over time, Figure 1 examines simple correlations of growth rates in a time-varying framework. More specifically, each correlation is calculated using ten years of data, starting at the indicated date; panel (a) shows the contemporaneous correlations for New Zealand's GDP growth with that in each of the other three countries, while panel (b) shows the corresponding correlations but now calculated in relation to foreign growth one quarter earlier.

Over the period 1992-1998, panel (a) indicates that growth in New Zealand is more strongly contemporaneously correlated with that of China and Australia than with the US and the relationships are rather stable. From around 1998, all correlations are relatively higher, which may reflect the effects of the global financial crisis internationally on GDP growth.

Figure 1: Correlations of New Zealand GDP growth with China, USA and Australia

Note: All correlations are computed using ten years of quarterly data, starting at the date indicated on the horizontal axis. Source: Statistics Haver, authors' calculations.

Interestingly, the correlations with lagged growth in panel (b) generally imply a less strong role for China on New Zealand than the contemporaneous ones of panel (a), indicating that the spillover effects from China growth to New Zealand appear quickly. On the other hand, spillovers from the US and Australia may take time to feed through to New Zealand, particularly later in the period. Nevertheless, lagged correlations with the US and China are relatively low before the global financial crisis.

In terms of our later modelling, the contemporaneous correlations, in particular, indicate that China's importance for GDP growth in New Zealand is not purely a recent phenomenon. The different temporal patterns of spillover effects from China and the US to New Zealand are also informative. Although these correlation patterns provide a useful first look at the data, they tell only part of the story, because they are not conditional on any other variables, either foreign or key domestic variables (namely inflation, interest rates and the exchange rate).

3.2  Commodity prices

As a small open economy and a commodity exporter, it can be anticipated that New Zealand will be influenced by changes in world commodity prices. This is supported by Figure 2, which shows the evolution of the weighted average world price for New Zealand’s major commodity exports, expressed in US dollars and deflated by the US consumer price index, alongside New Zealand’s terms of trade during the period 1986Q1 to 2011Q4. Four real commodity price series are shown: the aggregate ANZ world price index and the indices for dairy, forestry and meat and wool products. It is evident that real commodity prices have risen dramatically over the past decade (except for forestry products), with a relatively brief interruption due to the global financial crisis. Largely in line with these commodity prices, New Zealand’s terms of trade have also risen since 2000.

Figure 2: Real commodity prices and New Zealand terms of trade

Note: Data are seasonally adjusted and expressed as indices with 1986=100. Nominal commodity price indices are deflated by the US consumer price index to form real commodity prices. Source: Statistics Haver, authors' calculations.

Commodity price increases since 2000 are driven by growing demand from developing countries, including China and India, and also reflect the impact of constrained supply conditions in various parts of the world due to factors such as adverse weather conditions. The marked peak in the dairy price index for about a year from the middle of 2007 pre-dates the melamine scandal in China, but nevertheless may reflect growth in demand from China for dairy products; see Bowman and Conway (2013a, 2013b) for further discussion.