ickestationäritet

Ickestationäritet is a term found in some German-language discussions to denote a form of non-stationarity in time-series data or dynamic systems. The term is not part of standard statistical nomenclature and is considered a neologism or colloquial label for phenomena where statistical properties change over time in complex ways.

Definition and scope: The concept describes time-varying behavior in the probabilistic structure of a process, including changes in mean, variance, and autocorrelation that are not explained by simple differencing or seasonal terms. It may encompass regime shifts, evolving seasonality, and conditional heteroskedasticity.

Classification: Ickestationäritet can be described as trend non-stationarity (changing mean), seasonal non-stationarity (changing seasonal patterns), structural-break non-stationarity (abrupt regime shifts), and stochastic non-stationarity (diffuse time-varying parameters).

Methodology: Analysts detect and model it with unit-root and stationarity tests (ADF, KPSS, PP), structural-break tests, time-varying parameter models, regime-switching models, and models addressing heteroskedasticity (GARCH). Differencing, de-trending, and seasonal adjustment are common preprocessing steps; more flexible approaches include local polynomial trends and state-space models.

Relation and reception: The term remains informal and is often used to contrast with strict stationarity. It relates to broader concepts such as non-stationarity, ergodicity, and cointegration in time series analysis.