DEMAND FORECASTING METHODS 2(QUANTITATIVE METHODS) OPERATION MANAGEMENT

·         QUANTITATIVE METHODS DEMAND FORECASTING/DEMAND FORECASTING 2

·         OPERATION MANAGEMENT

·         MANAGERIAL( MICRO) ECONOMICS

•       TIME SERIES ANALYSIS

1.       DEMAND IS FORECASTING ON THE BASIS OF ANALYSIS OF HISTORICAL DEMAND DATA OVER A NO OF YEARS.

2.       WHEN THIS DATA IS ARRANGED IN A CHRONOLOGICAL ORDER WITH APPROPRIATE TIME INTERVAL IT IS CALLED TIME SERIES.

•       COMPONENT OF TIME SERIES

•       THE TIME SERIES OF DATA IS GENERALLY EFFECTED B VARIOUS MOVEMENTS OR FLUCTUATIONS CALLED COMPONENTS OF THE TIME SERIES

1.       LONG TERM TREND: THE CHANGES IN ACTUAL DEMAND MAY BE UPWARD OR DOWNWARD OVER A PERIOD OF FORMING A TREND.

2.       CYCLICAL : CHANGES DUE TO DEPRESSION OR BOOM

3.       SEASONAL :-CHANGES DUE TO CLIMATE OR FESTIVE

4.       RANDOM:-CHANGES DUE TO I=UNCONTROLLABLE EVENT WHICH CAN NOT BE PREDICTED

•       TREND PROJECTION METHOD

•       LONG RUN TENDENCY OF A TIME SERIES TO INCREASE OR DECREASE OVER A PERIOD OF TIME IS KNOWN AS TREND. PAST TREND IS USED TO PREDICT FUTURE TREND. TREND CAN BE MEASURED BY USING THE FOLLOWING TECHNIQUES:-

1.       GRAPHIC METHODS

2.       LEAST SQUARE METHODS

3.       METHODS OF MOVING AVERAGE

4.       EXPONENTIAL SMOOTHING

•       GRAPHIC METHOD

1.       SIMPLEST METHOD AS FREE FROM ANY MATHEMATICAL CALCULATIONS

2.       TIME SERIES DATA IS PLOTTED ON GRAPH BY TAKING TIME ON X AXIS AND OTHER VARIABLE ON Y AXIS

3.       PLOT THE DATA ON THE GRAPH

4.       IN ORDER TO REPRESENT THESE PLOTTED POINTS FREE HAND

•       GRAPHIC METHODS

•       MOVING AVERAGE METHOD

1.       FUTURE DEMAND ARE CALCULATED ON THE BASIS OF AVERAGE DEMAND OF PREDETERMINED NO OF PREVIOUS YEAR CALLED PREDETERMINED WINDOW.

2.       IT MAY BE 3,4,5,6,7 YEAR AVERAGE AND USED FOR SHORT TERM FORECASTING

3.       MOVING AVERAGES CONSISTS OF A SERIES OF ARITHMETIC MEANS CALCULATED FROM OVERLAPPING GROUPS OF SUCCESSIVE VALUES OF A TIME SERIES.

4.       MOVING AVERAGE METHOD

5.       MOVING AVERAGE METHODS: FIRST VALUE OF MOVING AVERAGE =1/N(A+B+C)

6.       SECOND VALUE OF MOVING AVERAGE=

=1/N( B+C+D)

7.       THIRD VALUE OF MOVING AVERAGE

=1/N( C+D+E)

8.       FOR CALCULATING THE DEMAND FORECAST OF ANY PERIOD,THE SUMMATION OF THE LAST ACTUAL DEMAND WILL BE DIVIDED BY NO OF YEARS

•       EXAMPLE

•       REGRESSION OR METHOD OF LEAST SQUARES

1.       REGRESSION ANALYSIS IS A STATISTICAL TOOL TO INVESTIGATE THE RELATIONSHIP BETWEEN TWO VARIABLES

2.       TECHNIQUE TO ESTIMATE THE UNKNOWN VALUE OF ONE VARIABLE CALLED DEPENDENT VARIABLE FROM INDEPENDENT VARIABLE.

3.       LEAST SQUARE METHODS

4.       WITH THE HELP OF TREND LINE IS FITTED TO THE DATA. KNOWN AS THE LINE OF BEST FIT. IT DOES NOT EXPLAIN THE REASONS OF CHANGE.

5.       LINEAR TREND:- Y =a + b X

6.        A AND B ARE INTERCEPT AND SLOPE AND Y IS THE NUMBER OF YEARS THE FOLLOWING TWO NORMAL EQUATION ARE TO BE SOLVED TO FIND OUT THE VALUE OF A AND B

7.       ∑Y =Na + b∑ X--------I

8.       ∑ XY = a ∑ X+ b ∑  X²  -------II

•       EXAMPLE

•       SOLUTION

1.       y =a + b x

2.     a=Y‾-bX‾

3.     X‾=∑X/N=260/5=52

4.     Y‾=∑Y/N=830/5=166

5.       b=(b∑XY-∑X∑Y)/n(∑X² )- (∑X)²

6.     =5X48600-(260x830)/(14250X5)-67600)

7.     (243000-215800)/(71250-67600)

8.     =27200/3650=7.45

9.     a=166-7.45(52)=166-387.5=-221.5

•       EXPONENTIAL SMOOTHING

1. POPULAR APPROACH FOR SHORT TERM FORECASTING.  THIS METHOD DETERMINES THE VALUES BY COMPUTING EXPONENTIALLY WEIGHTED SYSTEM. THE WEIGHT ASSIGNED TO EACH VALUE REFLECT THE DEGREE OF IMPORTANCE OF VALUE.
2. IN THE SIMPLE MOVING AVERAGE THE PAST DEMAND IS WEIGHTED EQUALLY.ASSIGNS WEIGHT TO DEMAND AS PER THEIR OCCURRENCE.THE MOST RECENT DATA IS GIVEN MORE WEIGHT AS COMPARED TO OLD DATA

3. F_T+1 =F_T +Α(A_T –F_T  ) = A_T +( 1- Α    ) F_T
4. WHERE F_{T +1} = NEXT PERIOD’S FORECAST ED DEMAND
5. A_T = ACTUAL DEMAND
6.  F_T=FORECAST ED DEMAND FOR CURRENT PERIOD WITH SIMPLE AVERAGE

7.       Α =SMOOTHING CONSTANT HAVING VALUE BETWEEN 0 AND 1. HIGHER VALUE LEADING TO GREATER RESONSIVENESS AND LOWER VALUE GREATER STABILITY

•       CASUAL METHOD

•       ESTIMATING TECHNIQUES BASED ON THE ASSUMPTION THAT THE DEMAND TO BE FORECASTED HAS CAUSE AND EFFECT RELATION.

•       CORRELATION METHOD:- STUDY OF ASSOCIATION BETWEEN TWO VARIABLES. WHTHER ONE VARIABLE IS ASSOCIATED WITH OTHER, IF YES WHAT IS THE DEGREE AND DIRECTION. VALUE LIES BETWEEN =+1 AND -1

•       ECONOMETRIC MODEL

•       ECONOMETRIC MODEL STUDY HISTORICAL RELATIONSHIP AMONG MACRO VARIABLE AFFECTING THE ECONOMY AND TRY TO FORECAST ITS IMPACT ON BUSINESS

•       MAIN METHODS:-

1.       ARIMA

2.       VECTOR AUTO REGRESSION

3.       BAYESIAN VECTOR REGRESSION MODEL

•       ARIMA METOD ( OR BOX-JENKIN TECHNIQUES)

•       BOX AND JENKIN DEVELOPED A METHOD OF FORECASTING USING INTEGRATIVE INTEGRATED MOVING AVERAGE. SUITABLE TO SITUATIONS WHERE THE INHERENT PATTERN IN UNDERLYING SERIES IS COMPLEX AND DIFFICULT TO UNDERSTAND. USED PRIMARY FOR SHORT TERM FORECASTING

•       FIVE STAGES OF ANALYSIS IN THIS METHOD

1.       Removal of the trend

2.       Model Identification

3.       Parameter Estimation

4.       Verification

5.       Forecasting

·         INPUT OUTPUT MODEL

1.       HERE FINAL OUTPUT OF ONE INDUSTRY BECOMES THE BASIS OF FORECASTING THE OUTPUT OR DEMAND OF EITHER INDUSTRY ON WHICH IT IS DEPENDENT FOR ITS INPUT

2.       USES IN BUSINESS TO BUSINESS DEMAND FORECASTING

3.       DEMAND OF CAR WILL DETERMINE THE OUTPUT OF TYRES AND AUTO ANCILLARY UNITS SUCH AS STEERING,CLUTCH AND WIND SCREEN.