Causal Analysis

Use

One of three categories of models used for forecasting. The other two are time series models and judgmental models. The basic premise of a causal model is that the future sales of a particular product or service are closely associated with changes in some other variable(s). Therefore, once the nature of that association or relationship is quantified, information about that other variable(s) can be used to develop a demand forecast. For example, you can gauge what price point you need to hit in order to reach a particular sales volume.

The causal model used in Demand Planning is multiple linear regression.

Causal analysis cannot determine which independent variables are relevant. You decide which variables to examine from the business process. Causal analysis however does give you information on how well a selected independent variable explains changes in the independent variable.

Features

Multiple linear regression (MLR) is a statistical technique that can be used to analyze the relationship between a single dependent variable and several independent variables. The APO system uses the ordinary least squares method to do MLR.

The objective of multiple regression analysis is to use the independent (or explanatory) variables whose values are known in the past and can be projected into the future to predict the future values of the single dependent variable. Each predictor variable (Xi) is weighted, the weights (b i) denoting their relative contribution to the overall prediction. In calculating the weights, the regression analysis procedure ensures maximal prediction from the set of independent variables. These weights also facilitate interpretation as to the influence of each variable making the prediction, although correlation among the independent variables can complicate the interpretative process.

The general notation for MLR is:

Y = b 0 + b 1X1 + b 2X2 + b 3X3...b nXn + ei

Where:

Y = Dependent variable

b 0 = Y intercept or constant

b i = Coefficients or weights

Xi = Independent variables

ei = Residual or prediction error

The assumptions of an MLR model are as follows:

• The Xs are non-stochastic.
• No exact linear relationship exists between two or more of the explanatory variables.
• Errors corresponding to different observations are independent and therefore uncorrelated.
• The error variable is normally distributed or has a Poison distribution.
• The error variable has 0 expected value.

Activities

You set up your causal model in an MLR profile. For instance here you decide which distribution to use and whether the variance is constant for all observations or whether it is variable. For more details see Creation of an MLR Profile