Presentation is loading. Please wait.

Presentation is loading. Please wait.

1 Forecasting BA 339 Mellie Pullman. What is a Forecast? What and why might we wish to forecast?What and why might we wish to forecast?

Similar presentations


Presentation on theme: "1 Forecasting BA 339 Mellie Pullman. What is a Forecast? What and why might we wish to forecast?What and why might we wish to forecast?"— Presentation transcript:

1 1 Forecasting BA 339 Mellie Pullman

2 What is a Forecast? What and why might we wish to forecast?What and why might we wish to forecast?

3 3 Forecasting Independent vs. Dependent DemandIndependent vs. Dependent Demand Qualitative Forecasting MethodsQualitative Forecasting Methods Simple & Weighted Moving Average ForecastsSimple & Weighted Moving Average Forecasts Exponential Smoothing ForecastExponential Smoothing Forecast Causal Forecast (Regression)Causal Forecast (Regression)

4 4 Independent vs. Dependent Demand A Independent Demand: Finished Goods B(4) C(2) D(2)E(1) D(3)F(2) Dependent Demand: Raw Materials, Component parts, Sub-assemblies, etc.

5 5 Independent Demand: What a firm can do to manage it. Can take an active role to influence demand.Can take an active role to influence demand. Can take a passive role and simply respond to demand.Can take a passive role and simply respond to demand. Forecasting Independent DemandForecasting Independent Demand

6 6 Types of Forecasts Qualitative (Judgmental)Qualitative (Judgmental) QuantitativeQuantitative – Time Series Analysis

7 7 Qualitative Methods Grass Roots Market Research Panel Consensus Executive Judgment Delphi Method Qualitative Methods

8 8 Quantitative Method: Time Series Analysis Uses historical data Uses historical data Many types of models available Many types of models available Pick a model based on: Pick a model based on: 1. Fits previous data best 2. Time horizon to forecast 3. Data availability 4. Accuracy required

9 9 Patterns of Demand Quantity Time (a) Horizontal (Random): Data cluster about a horizontal line. Quantity Time (b) Trend: Data consistently increase or decrease.

10 10 Patterns of Demand Quantity |||||| 12345 years (d) Cyclical: Data reveal gradual increases and decreases over extended periods. Quantity ||||||||||||JFMAMJJASOND||||||||||||JFMAMJJASOND Year 1 Year 2 (c) Seasonal: Data consistently show peaks and valleys.

11 11 Finding Components of Demand 1234 x x x x x x xx x x x xxx x x x x x xx x x x xxx x x x x x x x x x x x x x x x x x x x x Year Sales Seasonal variation Linear Trend

12 12 Simple Moving Average D t = actual demand from period tD t = actual demand from period t F t+1 = forecast of demand for period t+1 (next period that has not occurred yet)F t+1 = forecast of demand for period t+1 (next period that has not occurred yet) Forecast for the next period t+1 = average from the last n periods of actual demand.Forecast for the next period t+1 = average from the last n periods of actual demand.

13 13 Simple Moving Average Let’s develop 3-week and 6-week moving average forecasts for demand. Assume you only have 3 weeks and 6 weeks of actual demand data for the respective forecasts

14 14

15 15

16 16 In-Class Exercise Develop 3-week and 5-week moving average forecasts for demand for week 8Develop 3-week and 5-week moving average forecasts for demand for week 8

17 17 Weighted Moving Average Determine the 3-period weighted moving average forecast for period 4. Weights: t.5 t-1.3 t-2.2

18 18 Solution

19 19 In-Class Exercise Determine the 3-period weighted moving average forecast for period 5. Weights: t.7 t-1.2 t-2.1

20 20 Exponential Smoothing (  is the smoothing parameter) Premise — we should determine how much weight to put on recent information versus older information. 0 <  < 1 0 <  < 1 High  such as.7 puts weight on recent demand Low  such as.2 puts weight on many previous periods F t+1 =  D t + (1-  )F t

21 21 Exponential Smoothing Example Determine exponential smoothing forecasts for periods 2-10 using  =.10 and  =.60. Let F 1 =D 1

22 22

23 23 Effect of  on Forecast

24 24 In-Class Exercise Determine exponential smoothing forecasts for periods 2-3 using  =.50 Let F 1 =D 1

25 25 Forecasting with Causal Relationships

26 26 Potential Relationships Temperature and SalesTemperature and Sales Interest rate and number of loansInterest rate and number of loans Average daily temperature or rainfall with acre- feet of water usedAverage daily temperature or rainfall with acre- feet of water used Others?Others?

27 27 34 What do you notice?

28 28 35 Simple Linear Regression Model b represents? a represents? Y t = a + bx 0 1 2 3 4 5 x (weeks) Y

29 29 37 Regression Equation Example Develop a regression equation to predict sales based on these five points.

30 30 y = 143.5 + 6.3t 135 140 145 150 155 160 165 170 175 180 12345 Period Sales Forecast 39

31 31 Choosing a Method: Depends on Forecast Error

32 32 Forecast Accuracy Forecasts Consist of 2 Numbers 1. The projection of actual demand (D), called the forecast (F) which projects historical patterns or relationships 2. The error (E) which defines deviation between the forecast and the actual demand Measures of Forecast Error E t = D t - F t

33 33 31 Example- Error Calculation MonthSalesForecast 1220n/a 2250255 3210205 4300320 5325315 Determine the Error for the four forecast periods

34 34 30 Forecast Errors Study the formula for a moment. Now, what does each calculation tell you?Study the formula for a moment. Now, what does each calculation tell you? –MFA: mean forecast error –MAD: mean absolute deviation

35 35 31 Example--MAD MonthSalesForecast 1220n/a 2250255 3210205 4300320 5325315 Determine the MAD for the four forecast periods

36 36 32 Solution MonthSalesForecastAbs Error 1220n/a 22502555 32102055 430032020 532531510 40


Download ppt "1 Forecasting BA 339 Mellie Pullman. What is a Forecast? What and why might we wish to forecast?What and why might we wish to forecast?"

Similar presentations


Ads by Google