Exponential Smoothing with Trend Forecast

What is this formula?

Exponential Smoothing with Trend Forecast, also known as Holt's Method, extends simple exponential smoothing by incorporating an estimate of the trend present in the data.

The forecast is obtained by combining the current smoothed level and the current smoothed trend.

When to use it

Use this method when historical data exhibits a relatively stable upward or downward trend but does not show strong seasonality.

It generally provides more accurate forecasts than simple exponential smoothing when a trend is present.

Example

Suppose:

Smoothed Level = 150

Smoothed Trend = 8

Formula:

Ft+1=Lt+Tt

Substitution:

Ft+1=150+8

Ft+1=158

Result:

Forecast for the next period = 158

Applications

Economic forecasting

Sales forecasting

Demand planning

Inventory management

Production planning

Business analytics

Operations management

Note

This formula represents the forecast stage of Holt's exponential smoothing model. The level and trend estimates are themselves obtained through additional smoothing equations. Forecast quality depends on the selection of smoothing parameters and the stability of the underlying trend. When seasonality exists, Holt-Winters methods are usually preferred.