4  Macroeconomic responses to trade, financial and climate shocks

Structural VAR models have typically been used to identify dynamic responses of an economy to particular shocks. This serves two purposes. It provides a means of analysing an estimated structural VAR and it reveals information about the dynamic properties of the economy investigated. The results can be used to inform policy makers and economic forecasters how economic variables such as real output and prices respond over time to changes in policy or other events. This type of analysis can also be used to inform the development of calibrated macroeconomic models (such as those developed for New Zealand by Black, Cassino, Drew, Hanson, Hunt, Rose and Scott, 1997 and Szeto, 2002) about the dynamic structure those models should be capable of replicating.

As with all empirical work, the information value of dynamic simulations depends on the validity of the structure of the simulated empirical model. For this reason, the analysis is restricted to dynamic properties in response to shocks that appear to have been identified. These include shocks to international variables, domestic financial variables and domestic climate. We do not attempt in this paper to separate out the impact of domestic fiscal, trade and industrial policy changes that have taken place in New Zealand during the past two decades. These are left for future development work.

The traditional means of analysing an estimated structural VAR model is through impulse response functions (Hamilton, 1994). Impulse response functions (IPS) represent the dynamic response of a variable in the model to an error term (referred to as a shock or innovation) in one of the structural equations. The transmission of the shock will depend on the form of the structural equations. Using Tables 2 and 3, a shock to , for example, will have a contemporaneous impact on domestic GNE, the nominal exchange rate, domestic asset returns and domestic prices, and an impact on these and other variables one period into the future, two periods into the future, ….etc. These reactions represent the impulse responses.

Each variable in our model can be expressed as a combination of current and all past errors in the structural equations. That is, from equations (2) and (4), the SVAR can be written in moving average representation as follows:

(7)    

where contains the dynamic multipliers used to map out the impulse response functions following innovations to the structural error terms. The impulse response function represents the dynamic path for from the ith equation following an innovation to the structural error term from the jth equation, holding all other structural error terms constant.

The evolution of domestic output for example is therefore determined by all shocks that enter the system of equations (13 in this system). However, as explained in Section 2.1, while some of these shocks are in principle capable of being identified and interpreted (such as climate shocks), others are more difficult to interpret. For example, although it is common practice to classify the shocks from GDP and GNE as supply and demand shocks respectively, this is not always justified and would not be appropriate for this model. One obvious reason is that some components of expenditure, such as investment and government infrastructure spending, simultaneously affect aggregate demand and supply and these components of demand cannot be identified in this model. The approach adopted here is therefore to label shocks on the basis of the structural equation from which they emanate.

Table 4: Impulse response shocks: One standard deviation of the structural error

	0.006		0.011		0.004
	0.470		0.007		0.024
	0.063		0.019		4.241
	0.024		1.21
	0.020		0.063

Note: The size of each shock is equal to one standard deviation of the structural error term. For example, the size of the foreign output shock corresponding to the impulse response functions presented below is equal to 0.6%.

The sizes of shocks applied to structural VAR systems are traditionally measured as either one unit or one standard deviation shocks of the structural error. The metric applied in this paper is a one-standard deviation shock of the structural error. The impulse response functions have been normalised by dividing by one standard deviation of the structural form shocks. This normalisation enables all the impulse responses to be plotted on a single scale. The sizes of each shock are shown Table 4.

As is commonplace in the VAR literature, sixty-eight percent confidence bands have been estimated for the impulse response functions using the Monte Carlo bootstrapping approach of Runkle (1987). This approach simulates the SVAR model to generate error distributions. The actual residuals are randomly sampled (with replacement), and the sampled residuals are then used to create artificial data to re-estimate and simulate impulse responses. That procedure is repeated 1000 times to calculate standard errors for the parameter estimates and the impulse response functions. The programming required for model estimation, simulating impulse responses and historical error variance decompositions discussed in Section 5 was undertaken using the RATS software (Estima, 2000).

4.1  Foreign output shock

The first shock examined is the impact of a positive innovation to world output (). Figure 2 illustrates an immediate response by export prices () and import prices (). Their peak responses both occur in the first quarter after the shock. However, the impact of the rise in on domestic GDP is greater than the impact of the rise in , as reflected in the eventual decline in domestic output (). The exchange rate () appreciates over the first two quarters, which will also dampen domestic output. The exchange rate eventually depreciates in response to relatively higher world output and higher world interest rates.

Figure 2: Responses to foreign output shock

4.2  Foreign financial shocks

4.2.1  Foreign interest rate shock

The responses to a rise in the world interest rate () are illustrated in Figure 3. This shock is transmitted immediately into higher domestic interest rates (). Higher domestic interest rates imply higher demand for domestic bonds and therefore lower demand for domestic equities (), reducing real returns on domestic equities. Domestic demand eventually falls in response to higher domestic interest rates and the fall in domestic equities.

Although not shown in Figure 3, higher world interest rates reduce the growth of import () and export () prices for up to eleven quarters. The impact on exports () (not shown) is initially small but exports increase after three quarters. The net effect of higher foreign interest rates is a fall in domestic output.

Figure 3: Responses to foreign interest rate shock

4.2.2  Foreign equity shock

Figure 4 shows that an increase in returns on world equities provokes an immediate increase in returns on New Zealand equities and a rise in domestic demand and domestic output. Domestic interest rates rise, which results in an eventual, fall in domestic demand and domestic output.

Figure 4: Responses to foreign equity shock