Retail Sales - July
Data Flash (New Zealand)
Retail Sales - July
Total retail sales rose 0.7% mom in July, 0.4 percentage points weaker than the provisional estimate of 1.1% released on 21 August. The level of seasonally adjusted sales was 6.4% higher than a year earlier. Growth in June and May was revised to 0.7% and 0.6% respectively from 0.5% and 0.8% previously.
A 4.1% increase in motor vehicle services made the largest positive contribution to growth during the month. This largely reflects higher petrol prices - we estimate that the price of petrol rose 3.6% in July.
A 3.2% decrease in motor vehicle retailing made the largest negative contribution to growth during the month, consistent with a sharp fall in motor vehicle registrations and imports. Excluding motor vehicle sales and services, retail sales rose by 0.8% mom in July - significantly weaker than the 1.7% mom growth estimated in the provisional release. Half of that increase was accounted for by a 1.2% rise in food retailing, perhaps reflecting a strong rise in food prices over the course of June and very strong tourist inflows in July.
The appliance retailing, hardware and furniture storetypes recorded reasonable growth in July. This outcome is surprising given the very soft construction and house sales data in recent months. This may partly reflect the passing through of price rises as a result of currency weakness or the bring forward of spending in anticipation of price rises. Department store sales fell 3.1%.
Over the past year, sales growth has been strongest in the Auckland region and weakest in the Wellington region.
Readers will recall that we had some difficulty reconciling Statistics NZ's strong provisional estimate of July retail sales with a range of indicators pointing to more subdued activity in the household sector (including confidence surveys, credit card billings, and movements in disposable income). Today's revisions bring the July data more into line with our sense of underlying sales growth.
Nonetheless, we think that there is some risk that underlying growth in retail volumes is still weaker than the July data portray.
We believe that at least half of the rise in sales in July can be accounted for by higher prices. We expect the CPI to rise by over 1% in Q3.
Net tourist flows were very strong in July, with arrivals some 17% ahead of the July 1999. This is likely to have been an important factor boosting food sales. It is unclear whether such growth has been maintained in August.
It is unclear whether purchases of large durable items have been pulled forward to some extent as households anticipate widely announced increases in prices as a result of the weakening exchange rate.
Statistics NZ's estimate of retail trade in July is boosted by a so-called `trading day' adjustment that tries to account for the differing number of each day of the week that make up each month. July contained 5 Saturdays and 5 Sundays - days that, on average, over the 1990s were associated with a lower level of recorded sales (reflecting, amongst other factors, shopping patterns and bookkeeping conventions). While in theory such adjustments are desirable, given possible changes in shopping patterns over the 90s (with the possibility of more weekend shopping in particular), there is some risk that the size of the adjustment - based on average patterns over the last eight years - is too great. Based on industry contacts, Statistics NZ do not believe this risk to be large but conclusive evidence is not available. However, if this issue is important, retail sales in August may print correspondingly weak relative to underlying growth as the trading day adjustment for August will work in the opposite direction.
Today's revision to the provisional estimate was the largest recorded since Statistics NZ began publishing an advance estimate in March. However, the size of the revision is consistent with previous comments by Statistics NZ that revisions of up to +/- 0.5 percentage points should be expected.
Darren Gibbs, Senior Economist (64) 9 351 1376
This, along with an extensive range of other publications, is available on our web site http://research.gm.db.com
In order to read our research you will require the Adobe Acrobat Reader which can be obtained from their website http://www.adobe.com for free.