|Page tools: Print Page|
SEASONALLY ADJUSTED AND TREND ESTIMATES
7 Series in this publication include original, seasonally adjusted and trend series. Special care should always be taken in interpreting data for the most recent months and quarters. Some of the original and all of the seasonally adjusted and trend series are subject to revision.
8 It is not uncommon for movements in original time series data and those provided from seasonally adjusted and trend series to differ significantly. Movements in a time series of original data may reflect several factors, including:
9 Seasonal adjustment and trend estimates help the user to identify the effect of these influences on the time series. Seasonal adjustment removes the effect of the last three listed influences from the data, leaving only the trend and short-term irregular movements. Trend estimates are then obtained by removing the effects of the short-term irregularities, which in many series can be a major contributor to movements in the original data. By comparing the historical trend series with the seasonally adjusted series, the user can identify the short-term irregularities which have influenced the original series.
10 Trend estimates produced by the ABS are based on Henderson Moving Averages, and are therefore subject to revision especially for the latest few months or quarters.
11 The general methods used in the ABS for making seasonal adjustments are described in Information paper: An Introductory Course on Time Series Analysis - Electronic Delivery ( cat. no. 1346.0.55.001). The method used to estimate trends is described in Information Paper: A Guide to Interpreting Time Series - Monitoring Trends ( cat. no. 1349.0).
12 The ABS uses concurrent seasonal adjustment to derive the combined adjustment factors. This means that data from the current month are used in estimating seasonal and trading day factors for the current and previous months. Concurrent adjustment can result in revisions each month to estimates for earlier periods. However, in most instances, the only noticeable revisions will be to the combined adjustment factors for the current month, the previous month and the same month a year ago. As a result of the concurrent seasonal adjustment the trend estimates are also concurrently estimated. Further information on concurrent seasonal adjustment can be found in Information Paper: Introduction of Concurrent Seasonal Adjustment into the Retail Trade Series, 1999 ( cat. no. 8514.0).
CHAIN VOLUME MEASURES
13 The chain volume measures appearing in this publication are annually-reweighted chain Laspeyres indexes referenced to the current price values in a chosen reference year (i.e. the year when the quarterly chain volume measures sum to the current price annual values). Chain Laspeyres volume measures are compiled by linking together (compounding) movements in volumes, calculated using the average prices of the previous financial year, and applying the compounded movements to the current price estimates of the reference year. Quarterly chain volume estimates are benchmarked to annual chain volume estimates, so that the quarterly estimates for a financial year sum to the corresponding annual estimate.
14 Chain volume measures are not generally additive. In other words, in general, component chain volume measures do not sum to a total in the way original current price components do. In order to minimise the impact of this property, the ABS uses the latest base year as the reference year (i.e. the year prior to the latest complete financial year). By adopting this approach, non-additivity does not exist for the quarters following the reference year (currently 2009-10) and is relatively small for the quarters in the reference year and the quarters immediately preceding it. A new base year and the reference year for volume measures are updated annually in the September quarter Australian National Accounts (Australian Economic Indicators, January issue). In the September quarter 2011 Australian National Accounts issue, the new reference year became 2009-10 for volume estimates. This had two effects, first, a new base year, 2009-10, was introduced which has resulted in revisions to growth rates in subsequent periods. Second, the volume estimates have been re-referenced to 2009-10 thus preserving additivity in the quarters after the reference year. Re-referencing changes levels but not growth rates, although some revision to recent growth rates can be expected because of the introduction of a more recent base year (and revisions to the current price estimates underlying the chain volume measures).
15 Columns in bold in the tables indicate a total. Columns usually sum from the left, with the total appearing on the right. A major exception is the balance of payments and international investment position data in Chapter 2. These data sum from the right with the total appearing on the left - in keeping with the international standard for these statistics. Some tables include subtotal columns and columns which need to be subtracted to obtain the total. Where this is not clear a footnote explains the relationship between the columns.
16 Columns in italics indicate that seasonally adjusted or trend estimates are not available for that column, and original data are provided instead.
17 Yearly periods shown as, e.g. 2009-10, refer to the fiscal year ended 30 June 2010.
18 Where figures have been rounded, discrepancies may occur between sums of the component items and totals. Percentage changes may not be exactly comparable to tabulated source series due to rounding of the source series, the percentage change series or both.
19 Deca (da) = ten
Hecto (h) = hundred
Kilo (k) = thousand
Mega (M) = million
Giga (G) = thousand million
Tera (T) = million million
These documents will be presented in a new window.