By Ben Jarman*
Since the Canterbury region was hit with a brace of devastating earthquakes in September 2010 and February 2011, the forecasting community has framed New Zealand’s growth outlook around the timing and extent of reconstruction of dwellings and infrastructure in the region.
As recently as October, NZ Treasury was projecting a multi-year block of 3%-plus GDP growth kicking off in 2H12, with residential investment rising from its current nadir to the peak shares of GDP seen in the last housing boom.
However, the resumption of significant aftershocks just before year-end has thrown the timing (and perhaps even the feasibility) of that reconstruction into question.
Such is the “lumpiness” of the reconstruction boost, the assumption on the timing maps into the annual GDP growth forecasts in a very irregular fashion.
Indeed, removal of that positive impulse lays bare the underlying anemic growth profile, which is characterized by deleveraging in both the public and household sectors.
In our outlook for 2012, we are therefore stressing that while our forecasts assume the rebuilding does occur as currently planned, the distribution of risks around this modal outcome is very asymmetric.
Further, we expect policymakers and financial markets to give significant weight to these downside risks.
Forecasting the mode, watching the mean
Most of the time, the outlook for an economy can be adequately described by appealing to simple descriptions like expected GDP growth. If the range of outcomes for the economy evolve fairly continuously under different scenarios, the mean, median, and mode of the GDP forecast distribution can be treated as interchangeable concepts: when we talk about what we expect, there is little difference between the most likely single outcome (the mode) and the probability-weighted most likely outcome (the mean).
Occasionally though, when the possibilities are diverse, higher-order characteristics of the distribution matter. This is how we would describe the current situation in New Zealand.
Our GDP forecast of 2.5% for 2012 (and similarly, 2.9% in 2013) represents the most likely individual outcome, i.e., the mode. In that view, the national income gains from a boom in construction activity, which kicks off late this year, compensate for the drag from continued austerity from the household sector and government, lifting the economy to above-trend growth.
However, the distribution of other possible growth outcomes is fat-tailed and heavily skewed to the downside.
Financial markets, which have to price these skewed probabilities, and RBNZ officials, who would be concerned by possible atrophy in the labor market and slipping potential growth if the downside is borne out (GDP has averaged just over 1% annual growth since exiting a serious recession in 2009) will be highly attuned to these risks.
The year 2012 therefore looks like one in which we forecast the mode, but worry about the mean. In this context, financial markets and the RBNZ are unlikely to give the economy and the growth forecasts the benefit of the doubt until rebuilding becomes glaringly obvious in the data. Below, we discuss the breakdown of our forecasts, where the risks lie, and the implications for monetary policy.
The reconstruction arithmetic
Construction activity in New Zealand has been weak for several years due to the lingering affects of a housing market correction that began in 2008. Activity slipped even further through 2011 as the February earthquake brought what little building activity was under way to a grinding halt. Given the underlying trajectory, we need to differentiate the Canterbury rebuilding boost from the modest cyclical normalization we otherwise expect nationwide.
We define the rebuilding impulse as being established when the level of private fixed capital formation regains its prior peak (from 4Q10), which should occur in 4Q12. Under that assumption, GDP growth looks to be 2.5% in 2012 and 2.9% in 2013. The chart (previous page) shows the simple arithmetic of how the growth forecasts change if this impulse is delayed to 2Q13 and to 4Q13. The impact of the more modest delay is significant to 2012 GDP growth, which drops to 2%, while both delays are damaging to 2013 growth, which slips to 2.1% and 1.3%, respectively, under these scenarios.
Government seismologists have described the resumption of large aftershocks over the last few weeks as “close to forecast.” On face value then, the generally decaying, though irregular, pattern of aftershocks should not materially challenge the government’s planned timeline for reconstruction.
However, with rebuilding not allowed to occur while aftershocks persist, receipt of insurance claims linked to the timing of dwelling construction itself, and the threat of further significant earthquakes ever-present, the downside scenarios easily could be realized.
Deleveraging with help from disinflation
New Zealand households’ saving rate only just turned positive last year after more than a decade in the red. Similarly, household debt to disposable income is very high at over 150%. Redressing this imbalance has been a focus of both the Treasury and the RBNZ.
Clearly, there is still a large adjustment to the saving rate required in order to repair household balance sheets, which will cap consumption growth even if the labor market does benefit from rebuilding activity.
However, we do expect real household consumption to outperform somewhat relative to this headwind (second chart above), almost matching GDP growth over the next couple of years.
The dividend from falling inflation (first chart) will increase households’ real purchasing power, and non-wage income should remain solid from the structurally high terms of trade.
Risks, structural drags warrant caution
The Kiwi economy has struggled to grow for several years. With ongoing structural adjustments required, and reconstruction, the “hero” of the growth story, subject to significant uncertainty, there is every reason for the RBNZ to tread cautiously.
On our forecasts, the OCR stays on hold at a very low 2.5% until September, by which time the rebuilding surge should be fairly evident in the building permits data. Even if the investment numbers play to script from there, on our forecasts policy will stay persistently looser than indicated by a standard Taylor rule.
However, this reflects the risk bias that is inherently ignored in our central forecast. Much of the premium in the Taylor rule would be eliminated by assuming that the RBNZ reflects the possibility of disappointment in the reconstruction effort in its expectations for GDP and inflation.
Similarly, on the currency front, looming downside risks should keep a cap on NZD, as will persistently low policy rates.
And, even if rebuilding occurs as planned, with household and public savings low, such a scale of investment will draw significantly on the external sector.
For this reason, the current account deficit widens substantially in our forecasts, which will be a compensating drag on NZD.
In either case, it seems that with the underlying growth trajectory characterized by public and private deleveraging, the policy rate, and NZD are unlikely to rise much over the forecast horizon.
Ben Jarman is an analyst at JPMorgan Australia and is based in Sydney.