91Ó°ÊÓ

In casual relationship forecasting, to be worthy of forecasting, any variable must be a number one indicator.

For example, we will expect that an extended period of rain will increase sales of umbrellas and raincoats. The rain causes the sale of rain gear. This can be a causal relationship, where one occurrence causes another.

All forecast contains some errorswhether the model is simple or sophisticated because the forecast is a prediction of the long run supported by past data. Forecast errors are often caused by changes in conditions that generated the past data.

As an example, an economic recession could change the demand certainly of unnecessary products.

Fact that every forecast models have some error, regardless of what a forecaster does, they will not predict all events within the future which can cause demand to fluctuate.