Skip to Content
May 16 at 04:38 PM

Second-Order Exponential Smoothing forecasting model


Hi SAP users,

I am new here, so I do appologize if this post is posted in the wrong place.

I have a short question regarding to one of your forecast models: Second-Order Exponential Smoothing.

I am currently doing my Master Thesis. In short, we are writing about food waste and its impact on sustainability through CO2e emissisons - we are trying to improve a current forecasting model for a firm which is a constant simple expoential smoothing model.

I have analyzed the second-order exponential smoothing model to see if it is a better fit, however, I am not quite sure how to interpret the formula of the model (

I can see that it contains two parts, G(t)(1) and G(t)(2). The first part is quite similar to the initial first-order smoothing model, but the second part confuses me a bit. Does it use t and t-1 vaues from the first smoothing and does it not include any beta or gamma values? I have read other theories on the second smoothing approach which often relates to a Holt Winter approach/model which includes beta and gamma values.

Does anyone have any mathematical derivations of the model, examples or simply want to explain my how to use the second-order exponential smoothing model? I have all data available including alpha values (no beta- and gamma values).

Thank you a lot!
Best regards, Oscar