Australian Bureau of Statistics
Celebrating the International Year of Statistics 2013

• Summary
• Downloads
• Explanatory Notes
• Related Information
• Past & Future Releases

What If...? Revisions to Trend Estimates


Effect of new seasonally adjusted estimates on trend estimates


SENSITIVITY ANALYSIS

Readers should exercise care when interpreting the trend estimates of recent months because they will be revised when next month’s seasonally adjusted estimates become available. For further information, see paragraphs 21 to 22 in the Explanatory Notes.

The graph below presents the effect of two possible scenarios on the previous trend estimates:

1 The March 2004 seasonally adjusted estimate of number of dwellings financed is higher than the February 2004 seasonally adjusted estimate by 4.0%.

2 The March 2004 seasonally adjusted estimate of number of dwellings financed is lower than the February 2004 seasonally adjusted estimate by 4.0%.

The percentage change of 4.0% was chosen because the average absolute percentage change of the seasonally adjusted series, based on the last 10 years of data, has been 4.0%.

Number of Dwellings Financed

What if next month's seasonally adjusted estimate: Trend as published (1) rises by 4% on this month (2) falls by 4% on this month no. % change no. % change no. % change September 2003 59,678 0.7 59,751 0.8 59,810 0.9 October 2003 59,094 -1.0 59,203 -0.9 59,307 -0.8 November 2003 57,739 -2.3 57,780 -2.4 57,832 -2.5 December 2003 56,008 -3.0 55,941 -3.2 55,804 -3.5 January 2004 54,200 -3.2 54,108 -3.3 53,657 -3.8 February 2004 52,431 -3.3 52,485 -3.0 51,636 -3.8 March 2004 - - 51,107 -2.6 49,826 -3.5 - nil or rounded to zero (including null cells)