Forecasting Demand

1
Forecasting Demand

• WHY?
• to quess the future demand art science at
  the same time
• to set objectives and create plans
• forecasted demand is a foundation for
  operational, tactical and strategic decisions

2
Subjects of Forecasts

• Macro forecasts
• Gross domestic product
• Consumption expenditure
• Producer durable equipment expenditure
• Residential construction
• Industry forecasts
• Sales of an industry as a whole
• Sales of a particular product within an industry

3

• Firm-level forecasts
• Sales
• Costs and expenses
• Employment requirements
• Square feet of facilities utilized

4
Prerequisities of Good Forecast

• must be consistent with other parts of business
• should be based on adequate knowledge
• should take into consideration the economic and
  political environment

5
FORECASTING TECHNIQUES
There are many forecasting techniques. Choosing
the right technique depends on various factors.

• the item to be forecast
• the relation between value and cost
• the quantity of historical data available
• the time allowed to prepare the forecast

6
Forecast Techniques

• QUALITATIVE (not just an emergency exit)
• QUANTITATIVE (naive or casual)

7
Naive methods project past data without
explaining future trends. Causal (or
explanatory) forecasting attempts to explain the
functional relationships between the dependent
variable and the independent variables.
8
Forecast Techniques

• QUALITATIVE (not just an emergency exit)
• Expert opinion (e.g. Delphi)
• Opinion polls and market research
• Economic indicators
• QUANTITATIVE (naive or casual)
• Projections
• Econometric models

9
Time Series Analysis
Assumption behaviour in the future will be
similar to behavior in the past (BUT consider
environmental, political changes, govenmental
measures, etc) Forecasting of stock values is a
modern version of transforming lead into gold
10
Time Series ComponentsYtf(Tt,Ct,St,Rt)

• We can think of time series as consisting of
  several components
• Trend T (long-term moving of the average)
• Cyclical component C (regular pattern of sequence
  of points above and belove the trend line) . Ex
  cyclical movements in the economy
• Seasonal component S (regular pattern of
  variability in a shorter period of time)
• Irregular component R (caused by unanticipated
  and nonrecurring factors - unpredictable)

11
Forecasting methods

• LAST VALUE
• Forecast
• Can be a good estimate.
• TREND LINE
• straight line Qab(t)
• exponetial line Yabt
• quadratic line Yab(t)c(t2)

12
Linear Trend
13
(No Transcript)
14
STRAIGHT LINE Yabt
sales13.663.606time
15
EXPONENTIAL LINE Yabt or lnYlnatlnb
lna2.86 i.e a17.49 lnb0.107 i.e b1.1138
SALES17.491.1138time
16
QUADRATIC LINEYabtct2
SALES21.41-0.27time0.35time2
17
(No Transcript)
18
BEST FITSALES21.41-0.27time0.35time2
Prediction 21.41-0.27110.3511260.79
19
Moving Average Method
NOTE the larger I, the slower response
to changes, but more stable predictions.
20
Other Forecasting Methods

• Weighted moving average
• Exponential smoothing
• Decomposition (trend, seasonal effects, cyclical
  effects)
• ARIMA
• etc.

21
Econometric Models

• Regression analysis ? estimation of
  coefficients
• ASSUMPTION the relationship between variables
  doesnt change from past into future
• ? on the basis of independent variables the
  dependent variable is predicted

22
Forecasting Demand for Beer

• We have already estimated monthly demand function
  for beer
• Q 10.088,13 1.79 ? A 716,67 ? T
• For the month after we estimated
• average temperature T4
• advertising outlays A7.000
• therefore
• Qpredicted 10.088,13 1,79 ? 7.000 716,67 ?
  4 25.478

23
Why Study Modeling?

• Models generate insight which leads to better
  decisions
• Modeling improves thinking skills
• Break problems down into components
• Make assumptions explicit
• Modeling improves quantitative skills
• - Number sense, sensitivity analysis
• Modeling is widely used by business analysts
• Finance, marketing, operations

24
Types of Models

• One time use models (usually built by the
  decision maker)
• Decision support models
• Embedded models
• A computer makes the decision without the user
  being explicitly aware
• Models used in business education

25
Benefits of Modeling

• Provides timely information
• Saves costs
• Relative to alternatives (e.g., surveys)
• By avoiding expensive errors
• Allows exploration of the impossible
• Improves business intuition

26
A Problem Versus a Mess

• A mess is a morass of unsettling symptoms,
  causes, data, pressures, shortfalls,
  opportunities, etc.
• A problem is a well-defined situation that is
  capable of resolution
• Identifying a problem in the mess is the first
  step in the creative problem solving process

27
Problems Statements

• Statement of the form In what ways might?
• Focuses attention on problem definition
• Approach taken to resolve problem differs by
  form of problem statement
• Should
• Pay close attention to problem definition
• Take any problem definition as tentative
• Prepare to alter definition if evidence suggests
  a different statement would be more effective

28
Divergent and Convergent Thinking

• Divergent thinking
• Thinking in different directions
• Searching for a variety of answers to questions
  that may have many right answers
• Brainstorming
• Convergent thinking
• Directed toward achieving a goal or single
  solution
• Involves trying to find the one best answer
• Emphasis shifts from idea generation to
  evaluation
• A decision maker needs to be clear about which
  process they are using at the current time

29
The Creative Problem-Solving Process

• 1. Exploring the mess
• Divergent phase
• Search mess for problems and opportunities.
• Convergent phase
• Accept a challenge and undertake systematic
  efforts to respond to it.
• 2. Searching for information
• Divergent phase
• Gather data, impressions, feelings, observations
  examine situation from many different viewpoints.
• Convergent phase
• Identify most important information.
• 3. Identifying a problem
• Divergent phase
• Generate many different potential problem
  statements.
• Convergent phase
• Choose a working problem statement.

30

• 4. Searching for solutions
• Divergent phase
• Develop many different alternatives and
  possibilities for solutions.
• Convergent phase
• Select one or a few ideas that seem most
  promising.
• 5. Evaluating solutions
• Divergent phase
• Formulate criteria for reviewing and evaluating
  ideas.
• Convergent phase
• Select the most important criteria. Use criteria
  to evaluate, strengthen, and refine ideas.
•  6. Implementing a solution
• Divergent phase
• Consider possible sources of assistance and
  resistance to proposed solution. Identify
  implementation steps and required resources.
• Convergent phase
• Prepare most promising solution for
  implementation.

31
Aspects of the Modeling Activity

• Problem context
• Situation from which modelers problem arises
• Model structure
• Building the model
• Model realization
• Fitting model to available data and calculating
  results
• Model assessment
• Evaluating models correctness, feasibility, and
  acceptability
• Model implementation
• Working with client to derive value from the model

32
Tools of Successful Modelers

• Technical skills
• Lead to a single correct answer
• e.g., calculating present values
• Craft skills
• Do not lead to a single answer
• e.g., designing a prototype

33
Modeling Heuristics

• Simplify the problem
• Break the problem into modules
• Build a prototype and refine it
• Sketch graphs of key relationships
• Identify parameters and perform sensitivity
  analysis
• Separate the creation of ideas from their
  evaluation
• Work backward from the answer
• Focus on model structure, not data

34
Expert Modelers Attitudes Towards Data

• Treat data skeptically
• Realize that even good data may not be relevant
  for the model
• Realize that data collection can be distracting
  and limiting
• Build the model structure first and then use data
  to refine it

35
Focus on Model Structure, Not on Data Collection

• Novice modelers spend a high proportion of time
  on data
• Expert modelers spend most of their time on model
  structure

36
Summary Introduction to Business Modeling

• Modeling is a necessary skill for every business
  analyst
• Modeling involves
• Abstracting the essential features of a situation
• Building a logical structure that mimics some
  aspects of the real world
• Analyzing that structure to generate insight
• Creativity is an essential ingredient in
  successful problem-solving and modeling it can
  be enhanced with training
• Analysts can learn the required modeling skills
• Management science/statistics are important
  advanced tools

37
Demand for Gasoline

• Outline
• Modeling demand for gasoline
• Theoretical model
• Empirical implementation
• II. Data
• Sources
• Data manipulations
• ----------------------------------------
• III. Estimation
• IV. Results

38
Literature

• Baltagi, B,H, and J.M. Griffin, (1983), Gasoline
  demand in the OECD An application of pooling and
  testing procedures, European Economic Review 22,
  117--137.
• Baltagi BH, Griffin JM (1997), Pooled estimators
  v.s. their heterogeneous counterparts in the
  context of dynamic demand for gasoline. Journal
  of Econometrics 77303327
• Badi H. Baltagi, Georges Bresson, James M.
  Gri.n1, Alain Pirotte, (2003), Homogeneous,
  heterogeneous or shrinkage estimators? Some
  empirical evidence from French regional gasoline
  consumption. Empirical Economics 28795811
• Berry, W.,Feldman S., (1985), Multiple Regression
  in Pratice, Sage University Press
• Kennedy P., (2001), A Guide to Econometrics, MIT
  Press
• ecomoetric textbook (Greene, Maddala...

39
Theoretical model

• Gasoline consumption (Sweeney, Griffin)
• utlization of typical auto (U)
• Gasoline efficency (E)
• Stock of cars (CAR)

gasoline Gasoline km driven
consumption stock Consumption per car
per km of cars
U ( 1/ E )
CAR)
GAS / CAR U / E
40

• UTILIZATION (U) DEPENDS ON
• per capita income (Y/N)
• gasoline price (P)
• stock of cars per capita (CAR/N)

• GASOLINE EFFICENCY (E) DEPENDS ON
• distributed lags on per capita income
• distributed lags on gasoline price

41
Empirical implementation
42
Static model
Elasticities

• price

• income

43
Dynamic model
Elasticities short-run long-run

• price

• income

44
Data

• SOURCE
• http//www.euromonitor.com/
• OECD (www.oecd.org
• IEA (www.iea.org)
• ...

45
VARIABLES

• Passenger cars in use (CAR)
• Price per litre of premium leaded/lead
  replacement petrol
• Price per litre of premium unleaded petrol
• GDP measured at purchasing power parity

• Purchasing power parity conversion factor
• GDP deflator
• CPI
• Exchange rates
• Consumption of motor gasoline
• Population

46
Data Manipulations (using EXCEL)

• Transforming nominal variables into real
  variables
• CPI indexes or
• GDP deflator
• Translating different currencies to a common
  denominator
• Exchange rate conversion or
• Purchasing-parity adjustment factors

47
Estimation

• econometric estimator OLS
• robustness checks (adding TRUCK/CAR, using
  different prices of gasoline,using TRUCKCAR
  instead of CAR,...)

48
Results static model
49
Results static model 2
50
Results dynamic model