SEASONALLY ADJUSTED AND TREND ESTIMATES
REINSTATEMENT OF CONCURRENT SEASONAL ADJUSTMENT AND TREND ESTIMATES
From this issue concurrent seasonal adjustment and current price trend estimates from November 2008 are reinstated to the retail trade series. The trend series were suspended as at November 2008 as it was not possible to determine the trend in retail turnover through the period affected by the Government's stimulus packages and other influences associated with global economic conditions. Without being able to determine the trend series, concurrent seasonal adjustment factors would have been unduly influenced by the spending resulting from these influences. As a result, forward seasonal factors were used. It is now possible to assess, for the purposes of trend estimation and concurrent seasonal adjustment, the impact on the Retail trade series of these influences.
These influences have been assessed as having both short term transitional and long term ongoing impacts on the Retail Trade original time series. The transitional impact is not part of the long term underlying behaviour of the series and should not be included in trend estimates. It is discounted in the calculation of seasonal factors. The assessment of the transitional impact is based on evidence that the level of the seasonally adjusted series have stabilised over the past three months - July, August and September 2009. The assessment that the series have stabilised is based on the estimated irregular components for these months being within two standard deviations of normal observed irregulars over the past 15 years.
Each industry subgroup has shown independent increased levels of spending between December 2008 and June 2009. Therefore each industry subgroup was modelled separately in estimating the transitional impact. The estimated impacts at the industry subgroup level were then reconciled to the estimated transitional impact at the Australian level. Each State by industry subgroup series received the same respective Australian industry subgroup correction. The trend estimates were produced by excluding the modelled transitional impacts but they contain the long term ongoing impacts of the various influences.
Note that this assessment of the impacts may be revised with the availability of additional time series data and further evidence of new spending levels.
The assessment that the level of the trend series has stabilised over the past three months and the calculation of modelled transitional impacts were made for the purposes of trend estimation and concurrent seasonal adjustment. This information should not be used as a measure of the amount of the stimulus packages spent on Retail trade. The increased levels of spending estimated may result from the Government's stimulus packages but could also reflect other irregular impacts like changes in interest rates and other influences associated with global economic conditions. This series does not measure how the stimulus packages were used - they may have been spent on retail activities (possibly bringing forward some expenditure or creating additional expenditure), non-retail activities, used to reduce debt or contributed to savings.
Forward seasonal factors based on data to the November 2008 reference month have been used since the November 2008 issue. From this issue concurrent seasonal adjustment is being used. This results in revisions to the recent seasonally adjusted estimates. The corrections for trend estimates referred to above are also used in the creation of the seasonal factors. This ensures that the factors for December 2008 to June 2009 are not unduly affected by the unusual spending patterns. As a result, the reintroduction of concurrent seasonal adjustment has not significantly revised the seasonally adjusted series for these months. The revisions to August 2008, September 2008 and August 2009 are slightly larger.