Linear Regression

Description

The Linear Regression function performs a linear regression to calculate business figures. One (simple linear regression) or several (multiple linear regression) independent variables can be considered.

Using this Function

Syntax

In the Control File, the Linear Regression function is called up as follows:

CODE

[account] = Linear regression in relation to parameter
[account] = Linear regression in relation to parameter1, parameter2, ... and parameterN

Parameters

[account]

Business figure for which the value should be calculated.
This must be included in the following input data and must contain historical values for at least two reference dates in the past: Input Files | Financial-Data

{in relation to} parameter

Name of the independent variable in the case of simple linear regression.
This must correspond to an entry in the ‘Parameter’ column in both Sheet Input Files | HISTORICAL_VALUES and Sheet Input Files | FORECASTS.

{in relation to} parameter1, parameter2, {... and} parameterN

List of the names of the independent variables in the case of multiple linear regression.
Each independent variable must correspond to an entry in the ‘Parameter’ column in both Sheet Input Files | HISTORICAL_VALUES and Sheet Input Files | FORECASTS.

Examples

Example 1

CODE

[10006] = Linear regression in relation to reference rate

execution of a simple linear regression with the following values:
- historical values of the business figure [10006] as the dependent variable:
  (6,200.00, 6,700.00, 7,200.00, 7,700.00, 8,200.00)
- historical values of the reference rate as the independent variable:
  (3.5%, 3.7%, 3.8%, 4.1%, 3.9%)
result of the regression:
- intercept: -4,200.00
- coefficient of the independent variable: 300,000.00
calculation for 31/12/2024:
- forecast of the reference rate: 4.2%
- new value of the business figure [10006]: -4,200.00 + 300,000.00 * 4.2% = 8,400.00
calculation for 31/12/2025:
- forecast of the reference rate: 4,0%
- new value of the business figure [10006]: -4,200.00 + 300,000.00 * 4.0% = 7,800.00

Example 2

CODE

[20002] = Linear regression in relation to reference rate and unemployment rate

execution of a multiple linear regression with the following values:
- historical values of the business figure [20002] as the dependent variable:
  (30,000.00, 33,000.00, 35,000.00, 35,500.00, 36,200.00)
- historical values of the reference rate as the first independent variable:
  (3.5%, 3.7%, 3.8%, 4.1%, 3.9%)
- historical values of the unemployment rate as the second independent variable:
  (5.5%, 5.2%, 5.8%, 6.2%, 6.3%)
result of the regression:
- intercept: -3,210.00
- coefficient of the first independent variable: 825,000.00
- coefficient of the second independent variable: 100,000.00
calculation for 31/12/2024:
- forecast of the reference rate: 4.2%
- forecast of the unemployment rate: 6.4%
- new value of the business figure [20002]: -3,210.00 + 825,000.00 * 4.2% + 100,000.00 * 6.4% = 37,840.00
calculation for 31/12/2025:
- forecast of the reference rate: 4.0%
- forecast of the unemployment rate: 6.2%
- new value of the business figure [20002]: -3,210.00 + 825,000.00 * 4.0% + 100,000.00 * 6.2% = 35,990.00