Berlin process

from Wikipedia, the free encyclopedia

The Berlin procedure (BV) is a mathematical method for component analysis and seasonal adjustment monthly and quarterly economic time series . The mathematical principles were developed at the end of the 1960s by the Technical University of Berlin and the German Institute for Economic Research (DIW). The most important user of the procedure is the Federal Statistical Office .

The current version 4.1 of the Berlin procedure is described in detail on the website of the Federal Statistical Office. In addition, the Federal Statistical Office provides user-friendly BV4.1 software for PC as freeware to carry out analyzes for non-commercial purposes .

Special features of the process

The current version 4.1 of the Berlin method differs from other common component breakdown and seasonal adjustment methods (such as X-12-ARIMA ) by the following characteristic properties:

  • Seasonal adjustment succeeds without any problems even with rapidly changing annual seasonal patterns.
  • The courses of the determined trend-economic components are economically plausible.
  • The cost / benefit ratio is low, because the user does not need any special training or even expert knowledge or extensive experience in handling the process to carry out high-quality analyzes.
  • The analysis results are basically user-independent, since no time-series-specific process parameters need to be defined.
  • Due to the use of linear regression models, there are basically no deviations between the directly and indirectly determined analysis results of aggregate series.

Web links