1. Adeyl Khan, Faculty, BBA, NSU Car buyer- Models & Option Does the dealer know! Basic Managerial function- Planning

2. Adeyl Khan, Faculty, BBA, NSU Forecast  A statement about the future value of a variable of interest such as demand.  Forecasting is used to make informed decisions.  Long-range (Plan system)  Short-range (Plan use of system) 3-2 Traffic Weather Traffic Weather I see that you will get an A this semester.

3. Adeyl Khan, Faculty, BBA, NSU Uses of Forecasts 3-3 AccountingCost/profit estimates FinanceCash flow and funding Human ResourcesHiring/recruiting/training MarketingPricing, promotion, strategy MISIT/IS systems, services OperationsSchedules, MRP, workloads Product/service designNew products and services Example!

4. Adeyl Khan, Faculty, BBA, NSU Features of Forecasts 1. Assumes causal system. past ==> future 2. Forecasts rarely perfect because of randomness 3. Forecasts more accurate for groups vs. individuals 4. Forecast accuracy decreases as time horizon increases 3-4

5. Adeyl Khan, Faculty, BBA, NSU Elements of a Good Forecast 3-5 TimelyReliable AccurateMeaningful WrittenEasy to use

6. Adeyl Khan, Faculty, BBA, NSU Steps in the Forecasting Process 3-6 Step 1 Determine purpose of forecast Step 2 Establish a time horizon Step 3 Select a forecasting technique Step 4 Obtain, clean and analyze data Step 5 Make the forecast Step 6 Monitor the forecast “The forecast” Traffic Weather Traffic Weather

7. Adeyl Khan, Faculty, BBA, NSU 3-7 Types of Forecasts Judgmental Time series Associative models Quantitative Qualitative

8. Adeyl Khan, Faculty, BBA, NSU 1. Judgmental Forecasts Uses subjective inputs  Executive opinions  Sales force opinions  Consumer surveys  Outside opinion  Delphi method  Opinions of managers and staff (Experts)  Anonymous, encourage honesty,  Questionnaire sequence  Achieves a consensus forecast  Other usage of this method 3-8

9. Adeyl Khan, Faculty, BBA, NSU 2. Time Series Forecasts Uses historical data  Trend - long-term movement in data  Seasonality - short-term regular variations in data  Specific dates, days, times  Cycle – wavelike variations of more than one year’s duration  Economic, political, GDP …  Irregular variations - caused by unusual circumstances  Atypical- remove from analysis  Random variations - caused by chance  Include? 3-9

10. Adeyl Khan, Faculty, BBA, NSU Forecast Variations 3-10 Trend Irregular variatio n Seasonal variations 90 89 88 Figure 3.1 Cycles

11. Adeyl Khan, Faculty, BBA, NSU 2. Time Series Forecasts Naive Forecasts 3-11 Uh, give me a minute.... We sold 250 wheels last week.... Now, next week we should sell.... The forecast for any period equals the previous period’s actual value.

12. Adeyl Khan, Faculty, BBA, NSU Naïve Forecasts  Simple to use  Virtually no cost  Quick and easy to prepare  Data analysis is nonexistent  Easily understandable  Cannot provide high accuracy  Can be a standard for accuracy 3-12

13. Adeyl Khan, Faculty, BBA, NSU Uses for Naïve Forecasts  Stable time series data  F(t) = A(t-1)  Seasonal variations  F(t) = A(t-n)  Data with trends  F(t) = A(t-1) + (A(t-1) – A(t-2)) 3-13

14. Adeyl Khan, Faculty, BBA, NSU Techniques for Averaging  Moving average  Weighted moving average  Exponential smoothing 3-14

15. Adeyl Khan, Faculty, BBA, NSU 2. Time Series Forecasts Moving Averages  Moving average – A technique that averages a number of recent actual values, updated as new values become available.  Weighted moving average – More recent values in a series are given more weight in computing the forecast. 3-15 F t = MA n = n A t-n + … A t-2 + A t-1 F t = WMA n = n w n A t-n + … w n-1 A t-2 + w 1 A t-1

16. Adeyl Khan, Faculty, BBA, NSU Simple Moving Average 3-16 Actual MA3 MA5 F t = MA n = n A t-n + … A t-2 + A t-1

17. Adeyl Khan, Faculty, BBA, NSU F t = MA n = n A t-n + … A t-2 + A t-1 F t = MA 3 = 3 A t-3 + A t-2 + A t-1 F t = MA 5 = 5 A t-5 + A t-4 + A t-3 + A t-2 + A t-1

18. Adeyl Khan, Faculty, BBA, NSU Exponential Smoothing  Premise--The most recent observations might have the highest predictive value.  Therefore, we should give more weight to the more recent time periods when forecasting.  Weighted averaging method based on previous forecast plus a percentage of the forecast error  A-F is the error term, α is the % feedback 3-18 F t = F t-1 +  ( A t-1 - F t-1 )

19. Adeyl Khan, Faculty, BBA, NSU Example 3 - Exponential Smoothing 3-19

20. Adeyl Khan, Faculty, BBA, NSU Picking a Smoothing Constant 3-20 .1 .4 Actual

21. Adeyl Khan, Faculty, BBA, NSU Time series forecasting Common Nonlinear Trends 3-21 Parabolic Exponential Growth Figure 3.5

22. Adeyl Khan, Faculty, BBA, NSU Linear Trend Equation  Ft = Forecast for period t  t = Specified number of time periods  a = Value of Ft at t = 0  b = Slope of the line 3-22 F t = a + bt 0 1 2 3 4 5 t FtFt

23. Adeyl Khan, Faculty, BBA, NSU Calculating a and b 3-23 b = n(ty) - ty nt 2 - ( t) 2 a = y - bt n    F t = a + bt 0 1 2 3 4 5 t FtFt

24. Adeyl Khan, Faculty, BBA, NSU Linear Trend Equation Example 3-24 t Week t2t2 y Sales ty 11150 24157314 39162486 416166664 525177885 S t = 15S t 2 = 55S y = 812S ty = 2499 (S t) 2 = 225

25. Adeyl Khan, Faculty, BBA, NSU Linear Trend Calculation 3-25 y = 143.5 + 6.3t a= 812- 6.3(15) 5 = b= 5 (2499)- 15(812) 5(55)- 225 = 12495-12180 275-225 = 6.3 143.5

26. Adeyl Khan, Faculty, BBA, NSU Techniques for Seasonality  Seasonal variations  Regularly repeating movements in series values that can be tied to recurring events.  Seasonal relative  Percentage of average or trend  Centered moving average  A moving average positioned at the center of the data that were used to compute it. 3-26

27. Adeyl Khan, Faculty, BBA, NSU Types of Forecasts 3. Associative Forecasting  Predictor variables - used to predict values of variable interest  Regression - technique for fitting a line to a set of points  Least squares line - minimizes sum of squared deviations around the line 3-27

28. Adeyl Khan, Faculty, BBA, NSU Linear Model Seems Reasonable 3-28 A straight line is fitted to a set of sample points. Computed relationship

29. Adeyl Khan, Faculty, BBA, NSU Linear Regression Assumptions  Variations around the line are random  Deviations around the line normally distributed  Predictions are being made only within the range of observed values  For best results:  Always plot the data to verify linearity  Check for data being time-dependent  Small correlation may imply that other variables are important 3-29

30. Adeyl Khan, Faculty, BBA, NSU Forecast Accuracy  Error - difference between actual value and predicted value  Mean Absolute Deviation (MAD)  Average absolute error  Mean Squared Error (MSE)  Average of squared error  Mean Absolute Percent Error (MAPE)  Average absolute percent error 3-30

31. Adeyl Khan, Faculty, BBA, NSU MAD, MSE, and MAPE 3-31 MAD = Actualforecast   n MSE = Actualforecast ) - 1 2   n ( MAPE = Actualforecas t  n / Actual*100) 

32. Adeyl Khan, Faculty, BBA, NSU MAD, MSE and MAPE  MAD  Easy to compute  Weights errors linearly  MSE  Squares error  More weight to large errors  MAPE  Puts errors in perspective 3-32

33. Adeyl Khan, Faculty, BBA, NSU Example 10 3-33

34. Adeyl Khan, Faculty, BBA, NSU Controlling the Forecast  Control chart  A visual tool for monitoring forecast errors  Used to detect non- randomness in errors  Forecasting errors are in control if  All errors are within the control limits  No patterns, such as trends or cycles, are present 3-34 Sources of Forecast errors Model may be inadequate Irregular variations Incorrect use of forecasting technique

35. Adeyl Khan, Faculty, BBA, NSU Tracking Signal 3-35 Tracking signal = ( Actual - forecast ) MAD  Tracking signal –Ratio of cumulative error to MAD Bias – Persistent tendency for forecasts to be Greater or less than actual values.

36. Adeyl Khan, Faculty, BBA, NSU Choosing a Forecasting Technique  No single technique works in every situation  Two most important factors  Cost  Accuracy  Other factors include the availability of:  Historical data  Computers  Time needed to gather and analyze the data  Forecast horizon 3-36

37. Adeyl Khan, Faculty, BBA, NSU Operations Strategy  Forecasts are the basis for many decisions  Work to improve short-term forecasts  Accurate short-term forecasts improve  Profits  Lower inventory levels  Reduce inventory shortages  Improve customer service levels  Enhance forecasting credibility 3-37

38. Adeyl Khan, Faculty, BBA, NSU Supply Chain Forecasts  Sharing forecasts with supply can  Improve forecast quality in the supply chain  Lower costs  Shorter lead times  Gazing at the Crystal Ball (reading in text) 3-38

39. Adeyl Khan, Faculty, BBA, NSU Username: NSU ID Password: alpine 39

40. Adeyl Khan, Faculty, BBA, NSU Problems P109-117 1, 7, 15, 24, 27 3-40

41. Adeyl Khan, Faculty, BBA, NSU Learning Objectives  List the elements of a good forecast.  Outline the steps in the forecasting process.  Describe at least three qualitative forecasting techniques and the advantages and disadvantages of each.  Compare and contrast qualitative and quantitative approaches to forecasting. 3-41

42. Adeyl Khan, Faculty, BBA, NSU Learning Objectives  Briefly describe averaging techniques, trend and seasonal techniques, and regression analysis, and solve typical problems.  Describe two measures of forecast accuracy.  Describe two ways of evaluating and controlling forecasts.  Identify the major factors to consider when choosing a forecasting technique. 3-42

43. Adeyl Khan, Faculty, BBA, NSU 3-43 Exponential Smoothing

44. Adeyl Khan, Faculty, BBA, NSU 3-44 Linear Trend Equation

45. Adeyl Khan, Faculty, BBA, NSU 3-45 Simple Linear Regression