Forecast Bias

What is this formula?

Forecast Bias measures the systematic tendency of a forecasting model to consistently overestimate or underestimate actual values.

A positive bias indicates that forecasts tend to be higher than actual observations, while a negative bias indicates that forecasts tend to be lower than actual observations.

Forecast Bias is one of the most important indicators used to evaluate forecast quality and detect systematic forecasting errors.

When to use it

Use this formula when evaluating forecasting performance and determining whether a forecasting model consistently overpredicts or underpredicts observed values.

Example

Suppose:

Forecast Value = 125

Actual Value = 115

Formula:

B=F-A

Substitution:

B=125-115

B=10

Result:

Forecast Bias = 10

Interpretation:

The forecast overestimated the actual value by 10 units.

Applications

Forecast evaluation

Demand forecasting

Sales forecasting

Inventory planning

Economic forecasting

Business analytics

Operations management

Note

Forecast Bias is a performance indicator rather than a forecasting model. In practice, bias is often evaluated across many periods using the Mean Forecast Error (MFE), which averages forecast errors over a dataset. The present formula represents the forecast error for a single observation and serves as the fundamental building block of more advanced bias metrics.