Skip to content
English
Back to the main page
  • There are no suggestions because the search field is empty.

ARIMA with Drift (Autoregressive Integrated Moving Average with Drift)

ARIMA with drift is a time-series forecasting method that estimates future values based on past patterns while also accounting for a steady average change over time.

The model combines three key elements:

  • past values of the series,
  • changes between periods (e.g. year-over-year differences),
  • and typical short-term fluctuations observed in the data.

In addition, it includes a drift term, which represents a long-term tendency such as gradual growth or decline. In simple terms:
→ the next value is estimated as the current value plus an average change and short-term variations.

ARIMA with drift is particularly useful for data that evolves gradually over time, such as the number of active companies, new registrations, or closures.

In North Data Company Intelligence, this method is used to generate stable and realistic forecasts, capturing both long-term trends and short-term movements while avoiding overly optimistic or unrealistic projections.