I have worked on methods of seasonal adjustment of time series for many years. I am still trying to do further research on the techniques of adjustment, and particularly on ways of making the standard adjustment software (such as the Census Bureau X-12-Arima program) easier to use. This page will be used to provide links to conference papers I have written, work in progress and experimental software.
This is a major project to convert the Bank's seasonal adjustment to use the latest US Census Bureau software. I acted as adviser and mentor to the Bank's own staff and carried out a good deal of the analysis. The attached paper was written, jointly with a member of the Bank of England staff, for presentation at the 2004 Joint Statistical Meetings in Toronto.
Download PDF file (222KB)The attached paper describes what I think is a new development in presentation of revision limits for recent seasonally adjusted and trend data. At present this is simply an open question for discussion; more comments may be added in the light of discussion at the Conference.
Download PDF file (145KB)Joint Statistical Meetings, Washington DC, August 2009. The attached paper describes an analysis of data on temperature anomalies from Goddard Institute of Space Studies.
Download PDF file (3.8MB)