Forecasting techniques generally assume:

accuracy that is better when individual items, rather than groups of items, are being considered.

continuity of some underlying causal system.

accuracy that increases the farther out in time the forecast projects.

a linear relationship between time and demand.

the absence of randomness.