what important assumption is made about statistical $($quantitive$)$ forecasting methods?

A$)$ the past is valid indicator of the future

B$)$ demand trend is seldom linear

C$)$ seasonal variations are small

D$)$ random variations are small

E$)$ all of above