QUANTITATIVE METHODS FOR MANAGEMENT

Transcription

1 3rd Term MBA-2016 QUANTITATIVE METHODS FOR MANAGEMENT COURSE OUTLINE 1. Introduction The main task of a manager is to make decisions. To make decisions managers need information. Information is spread in a great variety of forms: experience of the managers, judgments and opinions, and mainly information is stored in form of data. The goal of this course is to learn to extract information contained in data that can help managers to make better decisions. The course is a half credit course with 12 sessions. But for the first one, all the sessions are case discussions. 2. Objectives In this course, the student is expected to get experience in business-oriented quantitative analysis and familiarity with the elementary techniques of data analysis. The course takes advantage of the technical notes, cases and assignments to present a range of situations which illustrate how the quantitative analysis allows to extract information from the data in order to describe and predict demand, customer behaviour and other issues helpful to make decisions. 3. Learning Outcomes i. Understand the informational needs a company has and to use the information companies have in order to use that information for making decisions. ii. iii. Learn how to produce sales forecasts using past sales data using curve fitting techniques Learn regression techniques in order to relate sales to other variables external to the company. 1

2 iv. Learn classification techniques in order to classify groups of elements (good clients vs. default clients) v. To use common sense and business experience to understand what are the most important techniques in each situation 4. Competences General Competences: Decision Making Analytic reasoning Judgment Communication Specific Competences: To learn the most basics statistics techniques in order to analyze the information a company has in databases. To learn how to apply the outcomes of the statistics techniques to make decisions. 5. Content The main topics covered by this course are: Forecasting with time series data. Forecasting with simple and multiple linear regression. Classification of clients and products. 6. Methodology The sessions are based on cases which show various applications of quantitative analysis in business. Each case has a companion Excel file with the corresponding data set. The students are expected to prepare each case before the class. To help them, a set of several technical notes, covering the topics mentioned above, is made available. The notes come with examples which are similar to the cases and assignments. 2

3 The discussion of the cases in the classroom is supported by statistical analyses performed with Excel. The regression and correlation analyses involved in the discussion of the cases are performed with Excel s Analysis ToolPak. 7. Evaluation The evaluation of the course is based on class participation, individual assignments, some possible quizz, team assignments and the final team project. 8. Course Outline & Bibliography # Title of Session & Material 1 Introduction to Technical notes: -ADN-248-E Forecasting in companies -ADN-251-E The normal distribution -ADN-258 Measuring uncertainties: Probability Functions This first session will be a lecture. After the class you can read the technical note ADN-248-E Forecasting in companies. At some moment during this first week of the course, it would be useful to read and work through the document ADN-251-E The normal distribution (only pages 1 to 6). To understand better some statistical concepts, read the note ADN-258 Measuring uncertainties: Probability Functions. 2 s Sales Forecasting (I) Time Series Sales Forecasting Prepare a forecast of the sales of The GAP Co. for the whole year 1996 with an explanation of how you got that forecast. Individual Assignment submission (I): Follow the instructions given in the file QMM 1st Assignment 2015D.xls 3 Sales Forecasting (II) Technical note: ADN-279-E Time Series Sales Forecasting Prepare a forecast of the sales of The GAP Co. for each quarter of 1996 with an explanation of how you got that forecast. S Individual assignment submission (II): Present a forecast of each quarter of 1996 of the GAP Company following the Holt Winters Method explained in the technical note. 3

4 4 Sales forecasting (III) Technical note: Holt Winters Forecast Technique Prepare a forecast of the sales of each quarter of 1996 using the Holt Winters method described in the Technical Note "Holt Winters Forecast Technique" 5 Simple linear regression (I) Case: [AD-342-E] Forecasting at the Gap Company (B) Data: AD-342-E GAP (B) data.xlsx Technical note: [QMM-TN-02] The regression line Data: qmm-tn-02.xls - Is there any relationship between the sales of the GAP Company and the growth rate of the US GDP? - Is there any way to predict Sales from other variables? - What level of sales would you expect for GAP for 2013 if the growth rate of the US GDP were 3%?? 6 Simple linear regression (II) Case: AD-235-E Return, volatility and beta. Data: AD-235-E Return, volatility and beta.xls This session will be devoted to applying the linear regression concept to a financial problem. Work the document AD-235-E. The section The CAPM may help you to understand what you are doing. In Finance courses you will learn (and are already learning) more about this. - Calculate expected (mean) returns, standard deviations and betas for the IGBM and each Industry. - How do the expected returns of these sectors compare with the risk-free asset yearly return of 14%? - State the linear relationship between the return of some industry (say, Metals, Banks and Food) and the return of the General Index. - What conclusions do you draw from the analysis about the soundness of investing in each industry? Hint: To get the annualized return of an investment one has to multiply the average daily return by the number of days of the year. To get the volatility, one has to multiply the standard deviation of the daily returns times the square root of the number of days of the year 7 Multiple linear regression (I) Technical note: [QMM-TN-03] Multiple regression Case: [QMM-13] Windsor housing prices Data: qmm-tn-03.xls, qmm-13.xls - Are the houses located in a preferred neighborhood more expensive than the others? For how much? 4

5 - What is the relationship between price and size of different houses? What is the interpretation of this relationship? - What is the relationship between price and number of bedrooms? What is the interpretation of this relationship? - What is the relationship between price and number of bathrooms? s Individual assignment (III): Present an analysis of the Hot Dog Pricing case 8 Multiple linear regression (II) Case: [QMM-02] Hot dog pricing Data: qmm-02.xls Answer the questions proposed at the end of the exercise Regression with lagged variables Technical note: [QMM-TN-06] Regression with lagged variables 9 Case: HBS Harmon Foods, Inc. Data: HBS Harmon Foods.xlsx, qmm-tn-06.xls - Which model do you think makes more sense? - What do you learn from the coefficients of model selected? - What improvements would you recommend to the product manager in terms of timing and amount of promotions? 10 Classification (I) Technical note: [QMM-TN-04] Classification Case: [QMM-15] The churn model Data: qmm-tn-04.xls, qmm-15.xls In this case we pretend to anticipate what clients that will churn in the future and to think about what we can offer them for them not to churn. Develop a profile of the potential churner, i.e. identify the characteristics of the customers most likely to churn. Develop a regression model for predict churn S Team assignment (I): Follow the instructions that will have been given in the previous session 11 Classification (II) Case: [QMM-15] The churn model (second round) In this session we will continue with The churn model case. Follow the instructions given at the end of the previous session. 12 Final session Case: [QMM-11] Tracking churners at movtel Data: qmm-11.xls Answer the questions proposed in the case Team Project 5