September 2017 seasonal adjustment and trend changes to building consents issued

We have improved the way we calculate the seasonally adjusted and trend series in building consents issued. These changes were introduced in the September 2017 release (published on 31 October 2017). Changes to seasonally adjusted series All seasonally adjusted series now include an adjustment for the timing of Easter. This accounts for when Easter moves between March and April. This change also helps the seasonal adjustment algorithm to better calculate the seasonal pattern for other parts of the year. This means that seasonally adjusted movements throughout the time series have been affected by the new methodology. This change reduces the volatility of most of our seasonally adjusted series. The monthly seasonally adjusted series affected by this change are: number of new dwellings consented (shown in the graph below) number of new houses consented value of consents for all buildings value of consents for residential buildings. The quarterly seasonally adjusted series affected by this change are: number of new dwellings consented number of new houses consented value of consents for all buildings value of consents for residential buildings value of consents for non-residential buildings (shown in the graph below). Changes to non-residential trends We have changed the way we treat outliers in the trends for the value of building consents for non-residential buildings. Previously the monthly trend excluded consents with a value of $25 million or more before 2006, and a value of $50 million or more from 2006 onwards. Now it excludes consents with a value of $100 million or more from 2006 onwards. The methodology before 2006 is unchanged. Previously, outliers were not excluded from the quarterly trend. For consistency, we are now excluding outliers from the quarterly trend. Before 2006, consents with a value of $25 million or more are excluded. From 2006 onwards, consents with a value of $100 million or more are excluded. en-NZ