5.2  Modelling national saving scenarios

The SWG commissioned the New Zealand Institute of Economic Research (NZIER) to investigate the effects of a significant increase in New Zealand's national saving rate. It was interested in seeing the effects of a simulated saving “shock” in a realistic dynamic general-equilibrium model of the New Zealand economy. The model traces the impacts over time on variables such as GDP, NFL, the exchange rate, and national income. Given that a significant reduction in NFL would also be likely to reduce the cost of capital to New Zealand businesses, the NZIER also looked at the additional effects of such a reduction on investment, the capital stock and output.

The modelling showed changes in the key variables relative to a baseline. The baseline is the counterfactual of what would have happened (according to NZIER's Quarterly Predictions macro-economic forecasts) if the shocks to national saving and the cost of capital did not occur. For example, the model's baseline, with no increase in New Zealanders' propensity to save, would see the level of NFL increasing from its current level of around 85% of GDP to around 115% by 2025.

5.2.1  Results of an increase in national saving 

The following questions were asked of the model: what increase in the national propensity to save would be required (i) to stabilise NFL at around their current level (90% of GDP) or (ii) to reduce them to 60% of GDP (similar to Australia's current level) within around 10 years?

According to the model, the first goal could be achieved by increasing the national saving rate by 3 percentage points of gross national disposable income (GNDI) over the three years from 2010 to 2012 and keeping it steady thereafter. Reducing NFL to 60% of GDP by 2020 would be an altogether more ambitious target - needing approximately a 3-percentage-point reduction in the NFL-to-GDP ratio each year between 2010 and 2020. In turn, this would require a reduction in consumption to GNDI of 8 percentage points (i.e., from its baseline level of 86% to 78%). Once the 60% NFL-to-GDP target is reached, the consumption ratio is able to increase but only back to 82% in order to maintain NFL at 60% of GDP thereafter.

5.2.2  Results of increased national saving plus a lower cost of capital

Increased saving will reduce domestic demand pressures and, over time, the risk premium that foreign lenders apply to their loans to New Zealand. Both effects will help reduce the cost of capital (interest rates and the cost of equity) that New Zealand businesses have to pay to finance investment in new plant and equipment, buildings, research and development and business expansion more generally. To gain insight into this effect, a second shock - a reduced cost of capital - was applied to the model. Specifically, the NZIER modellers used IMF estimates and other evidence as a basis for assuming that a fall in New Zealand's NFL-to-GDP ratio of 25 percentage points relative to baseline will result in a fall in the cost of capital of 10% below its baseline.

Three scenarios were tested using different combinations of the two shocks, as described in Table 3 below:

Source: NZIER

Scenario 2 comprises the largest ‘shocks' and therefore has the largest effects. The other scenarios demonstrate similar effects in milder form. The key conclusions that the SWG draws from NZIER's modelling are:

• Debt-servicing costs fall. Reducing the NFL-to-GDP ratio to 60% in itself yields a significant financial gain to New Zealanders. This comes from savings in servicing foreign liabilities – less interest and/or profits paid to foreigners. It lifts New Zealand's real income and living standards over time. For example, under scenario 2, the offshore payments that New Zealand makes reduce by $1.5 billion per year by 2025 compared to baseline as the debt to be serviced drops, and national income correspondingly rises.
• Exchange rate falls, exports rise. Higher saving leads to a lower real exchange rate over the medium term, with higher exports and lower imports. The current account improves. These happen because New Zealanders borrow less from abroad (so lower demand for NZ dollars) and exports and import substitution expand to use available resources and make up for the fall in consumption.

Figure 5.4 illustrates the changes in volumes of exports and imports from the baseline under scenario 1. Volumes revert towards the baseline after an initial surge because the initial increases in saving partly reverse and, as the foreign liabilities decline, fewer exports are needed to service them.

Figure 5.4: Trade volumes – scenario 1 (percentage changes from baseline)

Source:  NZIER

• Lower cost of capital boosts investment. Scenario 1 has NFL to GDP falling 25 percentage points relative to baseline by 2025^[14] and under scenario 2 the fall is 55 percentage points. These falls generate falls in the cost of capital of 10% and 22% respectively and the model estimates the effects of these on investment and capital stock relative to baseline.

Figure 5.5: Cost of capital - scenario 1 vs scenario 2

Source:  NZIER

Figure 5.6: Capital stock - scenario 1 vs scenario 2

Source:  NZIER

• GDP and GNDI initially drop but then grow strongly. The initial falls come from the short-term output and employment losses in the face of the fall in consumption. GDP and income growth in the medium term come from a higher capital stock while GNDI benefits additionally from the reduced cost of debt servicing. Under scenario 2, consumption and GDP increase by 8% -10% compared to baseline and GNDI by around 15% by 2025. Figure 5.7 illustrates these changes from baseline.

Figure 5.7: GDP, income and consumption - scenario 2

Source:  NZIER

• Rebalancing – tradable industries expand, non-tradable ones contract. The effects of lower consumption, a fall in the exchange rate, higher exports and lower imports show up at the industry level - tradable industries expand, non-tradable ones contract. Of course resources (labour and capital) need to move to support these changes and this takes time and is not costless.

Figure 5.8 shows a 2014 snapshot of shifts in the outputs of 16 industries relative to baseline in scenario 1. The shifts in scenario 2 are larger. The shifts are decomposed into changes arising from local demand, imports (domestic share of importables) and exports.

Figure 5.8: Decomposition of industry output shift in 2014 – scenario 1

Source:  NZIER

Assessment of the model results

Models are not reality and indeed deliberately abstract from reality. They are only as good as the quality of the relationships and empirics built into them. We wanted to test our intuition and simple macro analysis and to outline the likely economic outcomes if New Zealand manages significantly to improve its national saving. The NZIER's dynamic computable general equilibrium model of the New Zealand economy offered a good way to do this.

The results from the modelling do indeed confirm intuition and economic analysis. They strongly suggest that a significant step up in national saving would cut NFL as a percentage of GDP, and lower the exchange rate (at least for a period), thus boosting exports and reducing imports. Under the reasonable view of an NFL cut leading to falls in the cost of capital, investment, output and incomes all rise over time, and national income is further boosted by the reduced cost of servicing foreign liabilities.

These are all desirable outcomes for the New Zealand economy – even if the sizes of the changes predicted by the model are subject to considerable uncertainty. The modelling results also confirm the merits of the much-talked-about rebalancing of the economy - resulting in significant shifts from consumption and debt towards exporting and investment following a significant increase in saving.