Demand Forecasting Of Nokia 3lx48
This document was ed by and they confirmed that they have the permission to share it. If you are author or own the copyright of this book, please report to us by using this report form. Report 3i3n4
Overview 26281t
& View Demand Forecasting Of Nokia as PDF for free.
More details 6y5l6z
- Words: 3,032
- Pages: 16
Demand Forecasting
1
2
“Demand forecasting is an estimate of sales in dollars or physical units for a specified future period under a proposed marketing plan”. This definition was given by the American Marketing Association.
INDEX
1) Introduction.......................................................................................03
2) Methods of demand forecasting............................................................04
3) Company Profile.................................................................................08
4) Method used for Nokia by me..............................................................10
3
Demand forecasting is the tool to predict the likely demand of a product in the future. The concept of matching supply with demand is straightforward. It is to get the balance between the inventory control and the customers’ need. Buying too much wastes time, money and space, also underestimating demand leads to back orders, cancellations etc. So demand forecasting provides a sophisticated yet easy method of to the challenges of many companies. Demand forecasting overcomes the problems by giving a simple source for statistical demand forecasting throughout the enterprise. This frees the planners and the buyers to concentrate their time on the fine points of merchandise planning and inventory optimization. The better a company can assess the demand the better it can use its resources. Factors are mainly three: company, competitive and macro economic factors. Company factors mainly include market share trends, changes in strategies and implementation. Competitive factors include competitor advertising, competitor product offerings, market share etc. Macroeconomic changes include mainly income, economic growth and shocks. There are many methods to assed or forecast demand, but none of the methods are to be a 100% success as the various factors involved can be independent or dependent of the other factors. But it still allows the company to have a view to how much does it need to supply the products in the market. Most of the companies use a single or a variety of methods. They may use the Simple sales analysis, market size and market share research. One of the most used nowadays method is that of Market size research and mind share research.
4
Forecasting is the method of estimation in totally unknown situations. Forecasting in the recent past has evolved into the practice of demand planning in day to day planning for manufacturing companies also for supply chain management. Forecasting benefits mainly include: 1) Saves time, money and effort 2) Reduced inventory 3) Increased sales and capacity.
Methods of Demand Forecasting There are a lot of methods to forecast the demand but the apt and correct has to be used to get the so called intended result from the forecast that is to be done. The choice of the type of the method to be used can be chosen on the various given factors: 1) Imminent objectives of forecast whether it’s a new product or to get impact from an
ment. 2) Cost involved, it should not be more than the benefits received or to be received
3) Time perspective, whether it’s for a short run or for a long run 4) Complexity of the technique 5) Nature and the quality of available data. The various methods involved for forecasting are: Methods based on Judgement Unaided judgement It is used when the following conditions do not hold: experts aren’t biased, or when large changes are unlikely to take place, when relationships are well understood by experts. It was found that experts were no better than chance when they used unaided judgement. Prediction markets 5
Some commercial organizations provide internet markets to allow participants to bet by trading s. Markets can be used to predict as the percentage of US households with HDTV by the end of a given time period. Confidential betting markets are also set up to bet on new entrants into the market. This method is used in the US and is mainly based on betting. Delphi method It was developed at RAND Corporation in 1950 to help capture the knowledge of diverse experts while avoiding the disadvantages of traditional group meetings. To forecast, the should recruit about 10-15 experts and poll them for their expert views. Then the gives anonymous statistics on the forecasts, and the experts’ reason for their forecast. This is repeated till there is a change in forecast between 2-3 rounds. The Delphi method or the result is the median or mode of the experts’ final forecasts. It was found out that the forecasts from the Delphi groups were more accurate than from unaided judgement and the traditional groups. Structured analogies The outcomes of similar situations from the past may help to forecast the outcome of a new target situation. This method is not carried out in a structured manner. For example, they might search for an analogy that suits them and then stop, so it’s not structured. To use this method, an prepares a description of the target situation and selects experts who have knowledge about the analogues; the experts identify and describe analogues situations, preferably direct situations. They rate their similarity to the target situation and match the outcomes of their analogies with potential outcomes in the target situation. Then the derives the forecast. Judgemental decomposition It is to divide the forecasting problem into parts that are easier to forecast than the whole. One then forecasts the parts individually using methods to each part. Then they’re combined to get the final forecast. Judgemental bootstrapping This method converts subjective judgements into structured procedures. Experts are asked for their methods and asked to make decisions for diverse cases which can be real or 6
hypothetical. The resulting data is then made to a model by estimating a regression equation relating the judgemental forecasts to the information used by the forecasters. This method allows experts to see how they’re weighing various factors. This knowledge can help to improve judgemental forecasting. Simulated interaction It’s a form of role playing for predicting decisions by people who are interacting with othes. Useful when conflicts arise. To simulate it, an prepares a description and gives all possibilities. Roles are essayed and from that a decision is reached and this is used to make the forecast.
Intention and Expectation surveys Here, people are asked how they would intend to behave to a particular situation and in an expectation survey the people are asked about their expectations. Then the probabilities scales such as the eleven scale method are used. By aggregating their intentions and expectations the demand is forecasted. Cont analysis Surveying of consumers is done about their preferences for alternative product design in a structured way; it is a possible to infer how different features will influence demand. In general, this method gives accuracy which increases with increasing realism of the choices presented to respondents.
Methods requiring quantitative data Extrapolation This method uses historical data to forecast the demand. Exponential smoothening is the most popular and cost effective of the methods. It implements the principle that recent data should be weighted and ‘smoothes’ out fluctuations to forecast the trend. Here the first
7
cleans and de seasonalises the data and selects reasonable smoothing factors. He then calculates an average and trend from the data and forecsts the demand. Quantitative analogues Experts identify situations that are analogues to those situations. These are then used to extrapolate the target situation. Experts collect those data’s and construct a model target situation and another using analogues data, combines the parameters of the model and forecasts it. Rule based forecasting Here a manager’s knowledge is integrated with the domain of time series data in a structured and in expensive way. One must find identify features of the series using statistical analysis, inspection and domain knowledge. The rules are then used to adjust data and to estimate short and long range models. Neutral nets These are computer intensive methods that use decision processes analogous to the human brain. They have the capabilities of learning as patterns changes and updating their parameters estimates. Also its found these methods are more accurate in real. Data mining It uses sophisticated analyses to identify relationships. It ignores theory and prior knowledge in search of patterns. Casual models These are based on prior knowledge and theory. Time series and cross sectional regression are the commonly used ones. They allow one to examine the effects of marketing activity thus providing information for contingency planning. For this method to work, suitable variables are to be chosen that can help this method. Then based on the prior knowledge and theory the forecasting can be carried out and the demand can be found out. Segmentation It involves breaking a problem into independent parts, using each data for each part to make a forecast and then the whole data is taken to forecast. It can be done for particular products. 8
For example, wool carpets are separately forecasted according to the climate and the region, so this method is very helpful and useful.
Company Profile The company that has been taken into here is NOKIA. Nokia’s history spans more than a hundred years and contains many stories, events and milestones brought about by the many twists and turns of world history and industrialization. It started in 1865 when engineer Fredrik Idestam established a wood pulp mill in Finland. Soon cardboard nokia became successful and it attracted a large workforce and a small community grew as such. It then became Finnish rubber works and then after the IInd world war it bought Finnish cable works and was collectively called Nokia group. It manufactured cables for telegraph and telephone networks and in 1970’s forayed into telecommunications and then on to become what it is today. Today Nokia has become one of the largest producers of mobile phones ranging from cheap ones to high end communicating devices. As it suggests, Nokia is really good at ‘connecting people’. This Finnish icon is the world’s largest manufacturer of mobile devices and has around 40% of the global device market. Nokia is very, very successful and, in 2006, generated revenue that for the first time was in excess of Finland’s state budget. Nokia has always used innovation as a key driver for growth: first, by pioneering GSM and then by reinventing the concept of product personalisation. These days Nokia’s challenge is to maintain its position in a world increasingly converged and dominated by the likes of Google 9
and Microsoft. However, while these companies have strong brands and interesting plans for the future, they don’t have control over the handset. Nokia is bundling great services with tailored, -friendly hardware. With a billion customers and relationships with hundreds of operators around the world, Nokia may well manage to hold its place. Core innovation strengths within Nokia include speed of action, interconnection between products and services and the strategic use of design. Over the past few years, as margins have been three times those of its nearest competitors, the company has re-emphasised the importance of design to place it literally at the heart of the organisation’s operations. This has moved Nokia from being a feature-driven to a design-driven manufacturer where, amongst others, ethnography is now a core capability linking consumer behaviour around new uses of digital media directly into the development process. Nokia’s leadership in mobile devices has given it a solid platform upon which to build a services business that expands outside the core. And here’s the rub, while Nokia has consolidated its traditional strengths of R&D and product design, the big story in 2007 was its strategic shift into the multimedia services space. Nokia created the N Series to deliver high-end multimedia phones. In addition, its nearubiquitous camera-phones have provided revenue opportunities from -generated content. In of the social networking trend, Nokia has formed alliances with partners such as Yahoo to allow s to share photos using its Flickr service. 2006 saw the first tablet device with internet, not cellular connectivity, and the success of this product has led to a nextgeneration of internet-only devices including webcams and higher levels of VoIP. By eliminating the SIM card and breaking the connection to the networks, Nokia is using alliances with the likes of Google and Skype to offer more freedom of choice direct to consumers. At the lower end of the market, Nokia is the leading brand in China and India and is well positioned for further growth. It continues to lead on process innovation to drive down cost and is opening its tenth factory in India with the capability to turn out 20m phones a year. Nokia’s innovation roap weaves its software and services into a seamless package. Smart acquisitions have increasingly played a considerable role in this. Seeing that that location and content services provide major opportunities Nokia bought Navteq, the leading provider of digital map information for over $8bn in cash. Nokia expects that the truly mobile internet with multiple connectivity options that enable faster access to music, video, TV and mobile 10
navigation and massive multiplayer gaming services will be a major factor in driving further growth. With overall industry handset volumes growing, Nokia is in the strongest position of any manufacturer. Add into this mix Nokia’s acquisitions and partnerships that are building new service portfolios and it is clear that this company continues to be the leading source of innovation in the telecommunications sector.
Method used to forecast sales of Nokia
Trend Projection Trend is a general pattern of change in the long run. This method is referred to the ‘classical method’ and is also a- very powerful tool used to predict future values of a variable on the basis of time-series data. Time series is composed of Secular trend- refers to the change occurring over a long time. Seasonal trend- refers to seasonal variations of the data within a year Cyclical trend- refers to cyclical movement in the demand for a product that may have a tendency to recur in a few years Random events- can be calamities, social unrest, foreign aggression etc. Here, the equation for linear trend is
Y=a+bX
11
Smoothing technique Most of the series do not show a continuous trend, some increase and decrease in values can be seen in any time series. To take care of these seasonal or random variations are made to ‘smooth’ the series. Smoothing techniques are used when the time series data exhibit little trend or seasonal variation. Here, the moving average method is used to forecast on the basis of demand values during the recent past. The formula used is:
Least square method For quarterly’s Year
2004 Q3 2005 Q3 2006 Q3 2007 Q3 2008 Q3 n=5
Sales
in Deviations from
millions(Euro)
2006 (X)
Y 6939 8403 10100 12898 12237 ∑Y= 50577
-2 -1 0 1 2 ∑X=0
XY
4 1 0 1 4 ∑
=10
-13878 -8403 0 12898 24474 ∑XY= 15091
Solving the equations: ∑Y = na + b∑X ∑XY = a∑X + b∑ We get a = ∑Y/n = 10115.4
b = ∑XY/∑
= 1509.1
So the eqn is Y = 10115.4 + 3956.5X
12
So for 2009 Q3 we assume: Y = 10115.4 + 1509.1(3) = 14642.7 million
Year
2004 Q4 2005 Q4 2006 Q4 2007 Q4 2008 Q4 n=5
Sales
in Deviations from
millions(Euro)
2006 (X)
Y 9063 10333 11701 15717 12662 ∑Y= 59476
-2 -1 0 1 2 ∑X = 0
XY
4 1 0 1 4 ∑
= 10
-18126 -10333 0 15717 25324 ∑XY= 12583
Solving the equations: ∑Y = na + b∑X ∑XY = a∑X + b∑ We get a = ∑Y/n = 11895.2
b = ∑XY/∑
= 1258.3
So the eqn is Y = 11895.2 + 1258.3X 13
So for 2009 Q4 we assume: Y = 11895.2 + 1258.3(3) = 15670.1 million
Yearly Year
2004 2005 2006 2007 2008 n=5
Sales
in Deviations from
millions(Euro)
2006 (X)
Y 29267 34191 41121 51058 50710 ∑Y= 206347
-2 -1 0 1 2 ∑X = 0
XY
4 1 0 1 4 ∑
= 10
-58534 -34191 0 51058 101420 ∑XY= 59753
14
Solving the equations: ∑Y = na + b∑X ∑XY = a∑X + b∑ We get a = ∑Y/n = 41269.4
b = ∑XY/∑
= 5975.3
So the eqn is Y = 41269.4 + 5975.3X So for 2009 Q4 we assume: Y = 41269.4 + 5975.3(3) = 59195.3 million
Smoothing technique In this method the average of the previous three months are taken and its average is used to get the demand for the next month. For 2005 Q1 is 7396, so for the next month it is 7396/3 = 2465.3 For 2005 Q2 is 8059, so for the next month it is 8059/3 = 2686.3 15
For 2005 Q3 is 8403, so for the next month it is 8403/3 = 2801 For 2005 Q4 is 10333, so for the next month it is 10333/3 = 3444.3 For 2006 Q1 is 9507, so for the next month it is 9507/3 = 3169 For 2006 Q2 is 9813, so for the next month it is 9813/3 = 3271 For 2006 Q3 is 10100, so for the next month it is 10100/3 = 3366.66 For 2006 Q4 is 11701, so for the next month it is 11701/3 = 3900.3 For 2007 Q1 is 9856, so for the next month it is 9856/3 = 3285.3 For 2007 Q2 is 12587, so for the next month it is 12587/3 = 4195.6 For 2007 Q3 is 12898, so for the next month it is 12898/3 = 4299.3 For 2007 Q4 is 15717, so for the next month it is 15717/3 = 5239 For 2008 Q1 is 12660, so for the next month it is 12660/3 = 4220 For 2008 Q2 is 13151, so for the next month it is 13151/3 = 4383.66 For 2008 Q3 is 12237, so for the next month it is 12237/3 = 2801 For 2008 Q4 is 12662, so for the next month it is 12662/3 = 4220.6 For 2009 Q1 is 9274, so for the next month it is 9274/3 = 3091.3 For 2009 Q2 is 12898, so for the next month it is 12898/3 = 4299.3 For 2009 Q3, the estimated is 14642.7, so for the next month it is 14642.7/3 = 4880.9 For 2009 Q4, the estimated is 15670.1, so for the next month it is 15670.1/3 = 5223.3
Bibliography 1) Managerial Economics by Geetika, Piyali ghosh and Choudhary 2) www.buseco.edu
16
1
2
“Demand forecasting is an estimate of sales in dollars or physical units for a specified future period under a proposed marketing plan”. This definition was given by the American Marketing Association.
INDEX
1) Introduction.......................................................................................03
2) Methods of demand forecasting............................................................04
3) Company Profile.................................................................................08
4) Method used for Nokia by me..............................................................10
3
Demand forecasting is the tool to predict the likely demand of a product in the future. The concept of matching supply with demand is straightforward. It is to get the balance between the inventory control and the customers’ need. Buying too much wastes time, money and space, also underestimating demand leads to back orders, cancellations etc. So demand forecasting provides a sophisticated yet easy method of to the challenges of many companies. Demand forecasting overcomes the problems by giving a simple source for statistical demand forecasting throughout the enterprise. This frees the planners and the buyers to concentrate their time on the fine points of merchandise planning and inventory optimization. The better a company can assess the demand the better it can use its resources. Factors are mainly three: company, competitive and macro economic factors. Company factors mainly include market share trends, changes in strategies and implementation. Competitive factors include competitor advertising, competitor product offerings, market share etc. Macroeconomic changes include mainly income, economic growth and shocks. There are many methods to assed or forecast demand, but none of the methods are to be a 100% success as the various factors involved can be independent or dependent of the other factors. But it still allows the company to have a view to how much does it need to supply the products in the market. Most of the companies use a single or a variety of methods. They may use the Simple sales analysis, market size and market share research. One of the most used nowadays method is that of Market size research and mind share research.
4
Forecasting is the method of estimation in totally unknown situations. Forecasting in the recent past has evolved into the practice of demand planning in day to day planning for manufacturing companies also for supply chain management. Forecasting benefits mainly include: 1) Saves time, money and effort 2) Reduced inventory 3) Increased sales and capacity.
Methods of Demand Forecasting There are a lot of methods to forecast the demand but the apt and correct has to be used to get the so called intended result from the forecast that is to be done. The choice of the type of the method to be used can be chosen on the various given factors: 1) Imminent objectives of forecast whether it’s a new product or to get impact from an
ment. 2) Cost involved, it should not be more than the benefits received or to be received
3) Time perspective, whether it’s for a short run or for a long run 4) Complexity of the technique 5) Nature and the quality of available data. The various methods involved for forecasting are: Methods based on Judgement Unaided judgement It is used when the following conditions do not hold: experts aren’t biased, or when large changes are unlikely to take place, when relationships are well understood by experts. It was found that experts were no better than chance when they used unaided judgement. Prediction markets 5
Some commercial organizations provide internet markets to allow participants to bet by trading s. Markets can be used to predict as the percentage of US households with HDTV by the end of a given time period. Confidential betting markets are also set up to bet on new entrants into the market. This method is used in the US and is mainly based on betting. Delphi method It was developed at RAND Corporation in 1950 to help capture the knowledge of diverse experts while avoiding the disadvantages of traditional group meetings. To forecast, the should recruit about 10-15 experts and poll them for their expert views. Then the gives anonymous statistics on the forecasts, and the experts’ reason for their forecast. This is repeated till there is a change in forecast between 2-3 rounds. The Delphi method or the result is the median or mode of the experts’ final forecasts. It was found out that the forecasts from the Delphi groups were more accurate than from unaided judgement and the traditional groups. Structured analogies The outcomes of similar situations from the past may help to forecast the outcome of a new target situation. This method is not carried out in a structured manner. For example, they might search for an analogy that suits them and then stop, so it’s not structured. To use this method, an prepares a description of the target situation and selects experts who have knowledge about the analogues; the experts identify and describe analogues situations, preferably direct situations. They rate their similarity to the target situation and match the outcomes of their analogies with potential outcomes in the target situation. Then the derives the forecast. Judgemental decomposition It is to divide the forecasting problem into parts that are easier to forecast than the whole. One then forecasts the parts individually using methods to each part. Then they’re combined to get the final forecast. Judgemental bootstrapping This method converts subjective judgements into structured procedures. Experts are asked for their methods and asked to make decisions for diverse cases which can be real or 6
hypothetical. The resulting data is then made to a model by estimating a regression equation relating the judgemental forecasts to the information used by the forecasters. This method allows experts to see how they’re weighing various factors. This knowledge can help to improve judgemental forecasting. Simulated interaction It’s a form of role playing for predicting decisions by people who are interacting with othes. Useful when conflicts arise. To simulate it, an prepares a description and gives all possibilities. Roles are essayed and from that a decision is reached and this is used to make the forecast.
Intention and Expectation surveys Here, people are asked how they would intend to behave to a particular situation and in an expectation survey the people are asked about their expectations. Then the probabilities scales such as the eleven scale method are used. By aggregating their intentions and expectations the demand is forecasted. Cont analysis Surveying of consumers is done about their preferences for alternative product design in a structured way; it is a possible to infer how different features will influence demand. In general, this method gives accuracy which increases with increasing realism of the choices presented to respondents.
Methods requiring quantitative data Extrapolation This method uses historical data to forecast the demand. Exponential smoothening is the most popular and cost effective of the methods. It implements the principle that recent data should be weighted and ‘smoothes’ out fluctuations to forecast the trend. Here the first
7
cleans and de seasonalises the data and selects reasonable smoothing factors. He then calculates an average and trend from the data and forecsts the demand. Quantitative analogues Experts identify situations that are analogues to those situations. These are then used to extrapolate the target situation. Experts collect those data’s and construct a model target situation and another using analogues data, combines the parameters of the model and forecasts it. Rule based forecasting Here a manager’s knowledge is integrated with the domain of time series data in a structured and in expensive way. One must find identify features of the series using statistical analysis, inspection and domain knowledge. The rules are then used to adjust data and to estimate short and long range models. Neutral nets These are computer intensive methods that use decision processes analogous to the human brain. They have the capabilities of learning as patterns changes and updating their parameters estimates. Also its found these methods are more accurate in real. Data mining It uses sophisticated analyses to identify relationships. It ignores theory and prior knowledge in search of patterns. Casual models These are based on prior knowledge and theory. Time series and cross sectional regression are the commonly used ones. They allow one to examine the effects of marketing activity thus providing information for contingency planning. For this method to work, suitable variables are to be chosen that can help this method. Then based on the prior knowledge and theory the forecasting can be carried out and the demand can be found out. Segmentation It involves breaking a problem into independent parts, using each data for each part to make a forecast and then the whole data is taken to forecast. It can be done for particular products. 8
For example, wool carpets are separately forecasted according to the climate and the region, so this method is very helpful and useful.
Company Profile The company that has been taken into here is NOKIA. Nokia’s history spans more than a hundred years and contains many stories, events and milestones brought about by the many twists and turns of world history and industrialization. It started in 1865 when engineer Fredrik Idestam established a wood pulp mill in Finland. Soon cardboard nokia became successful and it attracted a large workforce and a small community grew as such. It then became Finnish rubber works and then after the IInd world war it bought Finnish cable works and was collectively called Nokia group. It manufactured cables for telegraph and telephone networks and in 1970’s forayed into telecommunications and then on to become what it is today. Today Nokia has become one of the largest producers of mobile phones ranging from cheap ones to high end communicating devices. As it suggests, Nokia is really good at ‘connecting people’. This Finnish icon is the world’s largest manufacturer of mobile devices and has around 40% of the global device market. Nokia is very, very successful and, in 2006, generated revenue that for the first time was in excess of Finland’s state budget. Nokia has always used innovation as a key driver for growth: first, by pioneering GSM and then by reinventing the concept of product personalisation. These days Nokia’s challenge is to maintain its position in a world increasingly converged and dominated by the likes of Google 9
and Microsoft. However, while these companies have strong brands and interesting plans for the future, they don’t have control over the handset. Nokia is bundling great services with tailored, -friendly hardware. With a billion customers and relationships with hundreds of operators around the world, Nokia may well manage to hold its place. Core innovation strengths within Nokia include speed of action, interconnection between products and services and the strategic use of design. Over the past few years, as margins have been three times those of its nearest competitors, the company has re-emphasised the importance of design to place it literally at the heart of the organisation’s operations. This has moved Nokia from being a feature-driven to a design-driven manufacturer where, amongst others, ethnography is now a core capability linking consumer behaviour around new uses of digital media directly into the development process. Nokia’s leadership in mobile devices has given it a solid platform upon which to build a services business that expands outside the core. And here’s the rub, while Nokia has consolidated its traditional strengths of R&D and product design, the big story in 2007 was its strategic shift into the multimedia services space. Nokia created the N Series to deliver high-end multimedia phones. In addition, its nearubiquitous camera-phones have provided revenue opportunities from -generated content. In of the social networking trend, Nokia has formed alliances with partners such as Yahoo to allow s to share photos using its Flickr service. 2006 saw the first tablet device with internet, not cellular connectivity, and the success of this product has led to a nextgeneration of internet-only devices including webcams and higher levels of VoIP. By eliminating the SIM card and breaking the connection to the networks, Nokia is using alliances with the likes of Google and Skype to offer more freedom of choice direct to consumers. At the lower end of the market, Nokia is the leading brand in China and India and is well positioned for further growth. It continues to lead on process innovation to drive down cost and is opening its tenth factory in India with the capability to turn out 20m phones a year. Nokia’s innovation roap weaves its software and services into a seamless package. Smart acquisitions have increasingly played a considerable role in this. Seeing that that location and content services provide major opportunities Nokia bought Navteq, the leading provider of digital map information for over $8bn in cash. Nokia expects that the truly mobile internet with multiple connectivity options that enable faster access to music, video, TV and mobile 10
navigation and massive multiplayer gaming services will be a major factor in driving further growth. With overall industry handset volumes growing, Nokia is in the strongest position of any manufacturer. Add into this mix Nokia’s acquisitions and partnerships that are building new service portfolios and it is clear that this company continues to be the leading source of innovation in the telecommunications sector.
Method used to forecast sales of Nokia
Trend Projection Trend is a general pattern of change in the long run. This method is referred to the ‘classical method’ and is also a- very powerful tool used to predict future values of a variable on the basis of time-series data. Time series is composed of Secular trend- refers to the change occurring over a long time. Seasonal trend- refers to seasonal variations of the data within a year Cyclical trend- refers to cyclical movement in the demand for a product that may have a tendency to recur in a few years Random events- can be calamities, social unrest, foreign aggression etc. Here, the equation for linear trend is
Y=a+bX
11
Smoothing technique Most of the series do not show a continuous trend, some increase and decrease in values can be seen in any time series. To take care of these seasonal or random variations are made to ‘smooth’ the series. Smoothing techniques are used when the time series data exhibit little trend or seasonal variation. Here, the moving average method is used to forecast on the basis of demand values during the recent past. The formula used is:
Least square method For quarterly’s Year
2004 Q3 2005 Q3 2006 Q3 2007 Q3 2008 Q3 n=5
Sales
in Deviations from
millions(Euro)
2006 (X)
Y 6939 8403 10100 12898 12237 ∑Y= 50577
-2 -1 0 1 2 ∑X=0
XY
4 1 0 1 4 ∑
=10
-13878 -8403 0 12898 24474 ∑XY= 15091
Solving the equations: ∑Y = na + b∑X ∑XY = a∑X + b∑ We get a = ∑Y/n = 10115.4
b = ∑XY/∑
= 1509.1
So the eqn is Y = 10115.4 + 3956.5X
12
So for 2009 Q3 we assume: Y = 10115.4 + 1509.1(3) = 14642.7 million
Year
2004 Q4 2005 Q4 2006 Q4 2007 Q4 2008 Q4 n=5
Sales
in Deviations from
millions(Euro)
2006 (X)
Y 9063 10333 11701 15717 12662 ∑Y= 59476
-2 -1 0 1 2 ∑X = 0
XY
4 1 0 1 4 ∑
= 10
-18126 -10333 0 15717 25324 ∑XY= 12583
Solving the equations: ∑Y = na + b∑X ∑XY = a∑X + b∑ We get a = ∑Y/n = 11895.2
b = ∑XY/∑
= 1258.3
So the eqn is Y = 11895.2 + 1258.3X 13
So for 2009 Q4 we assume: Y = 11895.2 + 1258.3(3) = 15670.1 million
Yearly Year
2004 2005 2006 2007 2008 n=5
Sales
in Deviations from
millions(Euro)
2006 (X)
Y 29267 34191 41121 51058 50710 ∑Y= 206347
-2 -1 0 1 2 ∑X = 0
XY
4 1 0 1 4 ∑
= 10
-58534 -34191 0 51058 101420 ∑XY= 59753
14
Solving the equations: ∑Y = na + b∑X ∑XY = a∑X + b∑ We get a = ∑Y/n = 41269.4
b = ∑XY/∑
= 5975.3
So the eqn is Y = 41269.4 + 5975.3X So for 2009 Q4 we assume: Y = 41269.4 + 5975.3(3) = 59195.3 million
Smoothing technique In this method the average of the previous three months are taken and its average is used to get the demand for the next month. For 2005 Q1 is 7396, so for the next month it is 7396/3 = 2465.3 For 2005 Q2 is 8059, so for the next month it is 8059/3 = 2686.3 15
For 2005 Q3 is 8403, so for the next month it is 8403/3 = 2801 For 2005 Q4 is 10333, so for the next month it is 10333/3 = 3444.3 For 2006 Q1 is 9507, so for the next month it is 9507/3 = 3169 For 2006 Q2 is 9813, so for the next month it is 9813/3 = 3271 For 2006 Q3 is 10100, so for the next month it is 10100/3 = 3366.66 For 2006 Q4 is 11701, so for the next month it is 11701/3 = 3900.3 For 2007 Q1 is 9856, so for the next month it is 9856/3 = 3285.3 For 2007 Q2 is 12587, so for the next month it is 12587/3 = 4195.6 For 2007 Q3 is 12898, so for the next month it is 12898/3 = 4299.3 For 2007 Q4 is 15717, so for the next month it is 15717/3 = 5239 For 2008 Q1 is 12660, so for the next month it is 12660/3 = 4220 For 2008 Q2 is 13151, so for the next month it is 13151/3 = 4383.66 For 2008 Q3 is 12237, so for the next month it is 12237/3 = 2801 For 2008 Q4 is 12662, so for the next month it is 12662/3 = 4220.6 For 2009 Q1 is 9274, so for the next month it is 9274/3 = 3091.3 For 2009 Q2 is 12898, so for the next month it is 12898/3 = 4299.3 For 2009 Q3, the estimated is 14642.7, so for the next month it is 14642.7/3 = 4880.9 For 2009 Q4, the estimated is 15670.1, so for the next month it is 15670.1/3 = 5223.3
Bibliography 1) Managerial Economics by Geetika, Piyali ghosh and Choudhary 2) www.buseco.edu
16
Related Documents 3h463d
Demand Forecasting Of Nokia 3lx48
December 2021 0
Demand Forecasting 1i3s4q
December 2019 58
Demand Forecasting f663f
June 2020 6
Demand Forecasting 1i3s4q
November 2019 48
Demand Forecasting 1i3s4q
December 2019 53