• Document: Section A. Index. Section A. Planning, Budgeting and Forecasting Section A.2 Forecasting techniques Page 1 of 11. EduPristine CMA - Part I
  • Size: 966.08 KB
  • Uploaded: 2019-03-24 07:33:56
  • Status: Successfully converted


Some snippets from your converted document:

Section A Index Section A. Planning, Budgeting and Forecasting Section A.2 – Forecasting techniques .................................................................................... 1 © EduPristine | CMA - Part I Page 1 of 11 Section A Section A.2 – Forecasting Techniques Section A. Planning, Budgeting and Forecasting Section A.2 – Forecasting techniques Prelude Planning involves forecasting key revenue and cost drivers. Forecasting can be done using qualitative (market research, anecdotes, opinions from experts, historical analogy) and quantitative methods. Here we are focusing on following quantitative methods of forecasting: 1) Linear Regression Analysis a) Single variable and b) Multi variable 2) Learning Curve Analysis a) Cumulative Average Time Learning Model and b) Incremental Unit Time Learning Model 3) Moving Averages a) Weighted Average and b) Exponential Smoothing 4) Time Series Analysis 5) Expected Value Techniques and Sensitivity Analysis A. Understanding of a simple regression equation and the measures associated with it What is a Linear Regression Analysis? 1. Statistical tool / model to establish relationship between one variable (called dependent variable, Y) with another (or a group of another) variable (called independent variable(s), Xi) 2. The relationship is then translated into a linear regression equation and used to predict / forecast the value of dependent variable (Y) given the values of independent variable(s) 1) Types of Regression Analysis: a) A simple regression analysis uses / involves only one independent variable b) A multiple regression analysis uses / involves multiple independent variables 2) A simple linear regression line has an equation of the form Y = a + bX, where a) X is the explanatory variable b) Y is the dependent variable c) b is the slope of the line and measures change in y w.r.t unit change in x Section A Section A.2 – Forecasting Techniques d) as is the intercept (the value of y when x = 0) 3) A linear regression is fitting a straight line to data and explaining the change in one variable through changes in other variables. This is based on following assumptions: a) Linearity: Linear relationship between X and Y (Y varies directly with first power of X) b) Constant Process: Process relating the variables is constant or stationary c) No auto correlation: Dependent variable is not auto-correlated – this implies the errors measured by Y(actual) – Y(predicted) are normally distributed with zero mean and a constant standard deviation d) No multi-co linearity: The independent variables are independent of each other. They are not correlated with each other 4) In real life, we hardly come across a situation where these assumptions are met. We nevertheless perform regression analysis. So, it’s likely that analysis doesn’t yield efficient results. We therefore have various measures to test the efficiency of a regression analysis or model: a) R Squared: Also known as coefficient of determination; takes a value between 0 and 1; explains the extent to which changes in dependent variable can be explained by changes in dependent variables; a statistical measure of how close the data are to the fitted regression line; i) 0 indicates that the model explains none of the variability of the response data around its mean. ii) 1 indicates that the model explains all the variability of the response data around its mean. b) T value: Measure of strength of relationship between the independent and dependent variable: i) A value of 0 means no significant relationship between the two and hence the independent variable should be removed from the regression analysis ii) should be more than 2 to indicate a strong relationship between the dependent and independent variables c) Standard Error (SE): A measure of the accuracy of predictions; a measure of dispersion around the regression line i) ~68% of observations should fall within ± 1 x SE ii) ~95% of observations should be within ± 2 x SE Regression Fit with Low R Regression Fit with High R Graphical Representation of squared value squared Value Standard Error Section A Section A.2 – Forecasting Techniques B. A multiple regression equation and when it is an appropriate tool to use for forecasting Multiple Linear Regression Equation: y = α + β1x1 + β2x2 + β3x3 + β4x4 Interpretation of the variables and measure of efficiency of the regression model remain as per single linear regression equation (intercept, slope, R squared, T value, standard error) When Multiple Regression is an appropriate tool to use for forecasting: 1. When a si

Recently converted files (publicly available):