A Study On Target Costing

A STUDY ON TARGET COSTING

A Project Submitted to University Of Mumbai For Partial Completion Of The Degree OF Master in Commerce Under The Faculty Of Commerce

BY CHETAN NARAYAN GHADI

Under The Guidance Of Prof. JYOTSNA PATEL

PRAKASH DEGREE COLLEGE KANDIVALI (WEST)

For The Year Of 2018-19 ~1~

A STUDY ON TARGET COSTING

A Project Submitted to University Of Mumbai For Partial Completion Of The Degree OF Master in Commerce Under The Faculty Of Commerce

BY CHETAN NARAYAN GHADI

Under The Guidance Of Prof. JYOTSNA PATEL

PRAKASH DEGREE COLLEGE KANDIVALI (WEST)

For The Year Of 2018-19

~2~

INDEX Chapt er No.

Particular

Page No. 7

1

INTRODUCTION

2

RESEARCH METHODOLOGY LITERATURE REVIEW

29

DATA ANALYSIS AND DATA PRESENTATION COCLUSION AND SUGGESTION BIBLIOGRAPHY

66

3 4 5 6

~3~

52

69 74

CERTIFICATE This is to Certify that MR.CHETAN NARAYAN GHADI has worked & duly Completed his Project Work for the degree of Master in Commerce under the Faculty of commerce in the ADVANCE COST ACCOUNTING & His Project is Entitled “ Deductions in Respect of A STUDY ON TARGET COSTING I further Certify that Entire Work has been done by the Learner Under My Guidance & that no part of it has been submitted Previously , for any Degree or Diploma of any University. It is his Own Works & facts Reported by his Personal Findings & Investigations.

Name & Signature Of Guiding Teacher

Date Of Submission :

External Examiner

~4~

Declaration by Learner I The Undersigned Miss/ Mr Chetan Narayan Ghadi Here by declare that the work embodied in this project work titled “A STUDY ON TARGET COSTING” forms my own contribution to the research work carried out under the guidance of Is a result of my own research work and has not been previously submitted to any other university for any other Degree/Diploma to this or any other university Wherever reference has been made to previous work of others, it has been clearly indicated as such and included in the bibliography. I, here by further declare that all information of this document has been obtained and presented in accordance with academic rules and ethical conduct.

Name and Signature of the learner

Certified by

Name and signature of the Guiding Teacher:

~5~

Acknowledgment To list who all have helped me is difficult because they are so numerous and the depth is so enormous I would like to acknowledge the following as being idealistic channels and fresh dimensions in the completion of this project. I take this opportunity to thank the University of Mumbai for giving me chance to do this project. I would like to thank my Principal, Dhanshree Mota for providing the necessary facilities required for completion of this project. I take this opportunity to thank our Coordinator Kinjal Patel for her moral support and guidance. I would also like to express my sincere gratitude towards my project guide Jyotsna Patel whose guidance and care made the project successful. I would like to thank my College Library, for having provided various reference books and magazines related to my project. Lastly, I would like to thank each and every person who directly or indirectly helped me in the completion of the project especially My Parents and Peers who supported me throughout my project.

~6~

Chapter 1: Introduction Target costing is an approach to determine a product’s life-cycle cost which should be sufficient to develop specified functionality and quality, while ensuring its desired profit. It involves setting a target cost by subtracting a desired profit margin from a competitive market price. A target cost is the maximum amount of cost that can be incurred on a Product, however, the firm can still earn the required profit margin from that product at a particular selling price. Target costing decomposes the target cost from product level to component level. Through this decomposition, target costing spreads the competitive pressure faced by the company to product’s designers and suppliers. Target costing consists of cost planning in the design phase of production as well as cost control throughout the resulting product life cycle. The cardinal rule of target costing is to never exceed the target cost. However, the focus of target costing is not to minimize costs, but to achieve a desired level of cost reduction determined by the target costing process. Successful organizations depend on the ability to continuously develop new products while meeting customer demand for improved cost, delivery, quality and flexibility. Many industries such as the aerospace Industry, are subjected to extremely competitive markets in which companies require accurate business cases and strategies for their products. Being in a competitive environment, cost has increasingly become one of the main parameters for clients. In response to improved cost, many

Manufacturers have begun to adopt tools and techniques: one such technique is target costing (TC). Target costing is used to understand the actual cost of a system, which will in turn find opportunities in reducing or improving the cost. The objective of this thesis is to accurately determine a fundamental question known to businesses: how much will the product cost?

~7~

Target costing is a system under which a company plans in advance for the price points, product costs, and margins that it wants to achieve for a new product. If it cannot manufacture a product at these planned levels, then it cancels the design project entirely. With target costing, a management team has a powerful tool for continually monitoring products from the moment they enter the design phase and onward throughout their product life cycles. It is considered one of the most important tools for achieving consistent profitability in a manufacturing environment.

~8~

1.1 Background

Target costing draws important links to concurrent engineering (CE), which aims at optimizing engineering design cycles. Bhuiyan et al. (2004)

State that CE reduces the overall lead time to design

components. As can be seen in the below picture, the CE process normally entails an iterative approach of four main activities in product development: product idea, evaluation, analysis and synthesis. Comprehending the target cost of the product idea phase will eliminate multiple time-consuming revisions of the process, hence optimizing engineering design cycles. CE refers to an integrated product development team consisting of engineering, finance, supply chain, marketing etc. in order to minimize the effort required downstream in the product development cycle. This team of experts will use a systematic approach described in the following figure to create a lean process in a products development in order to satisfy the customer’s needs.

Target cost models are utilized as critical decision tools to approximate the cost of products. According to Blanchard & Fabrycky (1998), ten to fifteen Percent of the total cost spent during the design phase commits eighty Percent of the total cost in the life cycle. Moreover, Davila (2000) argues that

~9~

70-80% of a product’s cost is set during product development and cannot be changed when the product reaches production. Hence, cost models used in early design phases are extremely important and the level of accuracy must be significantly high in order to reduce cost early on, when effective cost models will be most beneficial due to the lack of product knowledge. In other words, the technical specifications for the products are unknown or unclear most of the time. These technical specifications can be referred to as cost estimation relationships (CER). CERs correspond to a positively correlated relationship between a dependant variable and the corresponding independent variable. For example, in the aerospace industry, CERs are based on full data sets consisting of all available costs and technical data associated with a particular product (Book et al., 2011). Book et al. (2011), developed an extension to the traditional CER named the adaptive CER. The adaptive CERs goal is to have smaller estimating errors and narrower prediction bounds. CER’s are used as fundamental knowledge required in building target cost models Increasing competition in global markets drives companies to deliver high quality goods at lower price. Monden and Hamada (1991) discuss the necessity of target costing in new product development of an industrial assembly based manufacturers. Traditionally, businesses used to calculate the cost of a product, add a profit margin and then sell the product to the public where the firm can meet an acceptable rate of return (Sudhir, 2009). Moreover, traditional cost accounting was developed from mass production where profitability is maximized when labour and machine utilisation is maximized. Traditional costing practices are no longer effective since the competitive environment, where multiple suppliers can provide the same product, makes cost reduction initiatives, optimization and continuous improvement a necessity to the organization. In the end, a higher selling price for a similar product does not adhere to the end customer (Cooper and Chew, 1996). Traditional accounting structures do not support organizations and rarely enable effective decision making by managers (Maskell, 2006).

~ 10 ~

Traditional accounting systems are no longer effective due to the following inefficiencies in the reporting (Maskell and Baggaley, 2006):

1. Large, complex and wasteful processes 2. Measurements and reports that motivate large batch production 3. No concept to measure lean improvements 4. Use of standard product costs for decision criteria

Organizations are moving towards target costing which is a component of lean management. Target costing, a subset of lean accounting, seeks to replace traditional costing practices by providing more timely and relevant management information. This method creates value to the end customer by establishing the right market price and then works backwards to design and manufacture a product in a lean fashion. Ansari and Bell (1997) explained in their work a very simplistic way in which target costing can be derived. One must commence by assessing the selling price of a product determined by its market forces. Thereafter, the profit margin of a specific product must be established. From the above statements, the technique known as target costing can be formally stated as follows:

Target price - Target profit = Target Cost

~ 11 ~

1.2 Origin of Target Costing

In Japan, target costing is has gained importance and widely practiced in more than 80% of the companies in the assembly industries and more than 60% of the companies in processing industries. It emerged in Japan in 1960s as a consequence of difficult market conditions. A proliferation of consumer and industrial products of western firms were overcrowding the markets in Asia.

Japanese companies were also experiencing shortages of resources and skills needed for the development of new concepts, tools and techniques, which were required to achieve parity with the toughest western competitors in terms of quality, cost and productivity.

Many Japanese companies considered modified cross-functional activities, as used by western firms for manufacturing. They believed that good results can be achieved by combining employees from strategy, planning, marketing, engineering, finance and production into expert teams.

These teams were able to examine new methods and techniques for the design and development of new products and aimed at increasing the degree of integration between upstream and

~ 12 ~

downstream activities of a firm’s operations. Target costing thus emerged from this background.

A range of specialized tools, including functional analysis, value engineering, value analysis and concurrent engineering were introduced to support the target costing. This made Japanese companies particularly effective in the area of product design and development.

They were able to identify all relevant elements to formulate a holistic management approach in order to achieve performance levels to meet the firm’s objective.

~ 13 ~

1.3 History of Target Costing

Target costing was developed independently in both USA and Japan in different time periods. Target costing was adopted earlier by American companies to reduce cost and improve productivity, such as Ford Motor from 1900s, American Motors from 1950s-1960s. Although the ideas of target costing were also applied by a number of other American companies including Boeing, Caterpillar, Northern Telecom few of them apply target costing as comprehensively and intensively as top

Japanese, companies

such as Nissan, Toyota, Nippondenso Target costing emerged from Japan from 1960s to early 1970s with the particular effort of Japanese automobile industry, including Toyota and Nissan. It did not receive global attention until late 1980s to 1990s when some authors such as Monden (1992), Sakurai (1989),Tanaka (1993), and Cooper (1992) described the way that Japanese companies applied target costing to thrive in their business (IMA 1994). With superior implementation systems, Japanese manufacturers are more successful

than

the

American

costing. Traditional cost-plus

companies

pricing strategy

has

in

developing been

target

impeding

the

productivity and profitability for a long time, As a new strategy, target costing is replacing traditional cost-plus pricing strategy by maximizing customer satisfaction by accepted level of quality and functionality while minimizing costs.

~ 14 ~

1.4 Definition

Target costing can be defined as “a structured approach for determining the cost at which a proposed product with specified functionality and quality must be produced to generate a desired level of profitability at its anticipated selling price”. A critical aspect of this definition is that it lays emphasis on the fact that target costing is much more than a management accounting technique. Rather, it is an important part of a comprehensive management process aimed at helping a firm to survive in an increasingly competitive environment. Target costing is a management technique aimed at reducing a product’s life-cycle costs. A general concept of target costing is discussed here. Target Costing is a disciplined process for determining and realizing a total cost at which a proposed product with specified functionality must be produced to generate the desired profitability at its anticipated selling price in the future. CIMA defines target cost as “a product cost estimate derived from a competitive market price” Target Costing is a disciplined process that uses data and information in a logical series of steps to determine and achieve a target cost for the product. In addition, the price and cost are for specified product functionality, which is determined from understanding the needs of the customer and the willingness of the customer to pay for each function.

Target costing is a formal process that attempts to match a proposed product’s features (benefits) with a viable market price that achieves the company’s profitability goals by:

~ 15 ~

(a) Determining a price point (or range of prices) for an approximate combination of features and benefits. (b) Subtracting a desired profit from the market price to determine the maximum bearable level of costs. (c) Iterating the product design—eliminating or reducing unnecessary attributes with costs that can’t be recovered in higher prices—until the cost target is met. (d) Revising the market price for the redesigned product in view of changed market conditions.

~ 16 ~

1.5 Process of Target Costing

1. Conducting Market Research 2. Identify the Nature of the market 3. Translation of customers requirements into product features 4. Development of Product Design 5. Determine the price, Margin and Cost 6. Conducting value Engineering Process 7. Improve the Design to reach Target cost 8. Approval of Top Management

9. Maintenance of Top Accounts

~ 17 ~

1. Conducting

Market Research: The company should

determine the customer wants precisely through conducting marketing research. A new product can be designed or make changes in the existing product on the basis of the Customers expectations and perceptions.

2. Identify the Nature of Market: The market information can be collected in such a way that what type of products are available in the market, the level of competition prevailing, the number of competitors and the price at which the existing products are available. Besides, the company should find out the affordable price of the customers. If so, the target costing is followed. 3. Translation of Customers Requirements into Product

Features: The preference of one customer differs from another. These

preferences

are

collectively

called

as

customers

requirements. Now, the bundle of preferences are bringing into a tangible thing i.e. product. 4. Development of a Product Design: By considering the engineering analysis of market forces, customer needs, relevant technology,

competitors

models,

product

configuration

and

performance features, design alternatives, process capabilities, maintenance and service requirements etc., a suitable product designed is to be determined by the company. Such a product design assures a targeted profit and target cost for each component in total.

~ 18 ~

5. Determine the Price, Margin and Cost: Target selling price is determined on the basis of market survey, at which the product can be sold. The standard margin is also included in the target selling price. If so, it is possible to determine the target cost

6. Conducting Value Engineering Process: The company can conduct value engineering process to reach target cost. It is a well known fact that the difference between target selling price and the target profit is target cost. The target selling price cannot be changed at any cost; Hence, it is a duty on the part of company is that takes necessary steps to reach the target cost.

7. Improve the Design to Reach Target Cost: The company starts a minor trial production. Such a production ensures all product performances, target cost and target profit margin also. The trial production comes to an end whenever the product design matches the target cost.

8. Approval of Top Management: A detail report is presented before the top management for getting approval. The report contains the production process, elements of cost involved with the level of costs to be incurred and design of the specified product. A formal approval is given for starting commercial production.

9. Maintenance of Accounts: A separate accounting records are to be maintained for each product design. It is possible to verify whether the total expenses exceed the target cost. If the expenses

~ 19 ~

are not controllable at any time, the product design will be changed. Hence, the maintenance of separate set of books are highly required under target costing process.

10. Implement the Target Costing: The company can get the information regarding the expenses incurred for each design separately. A continuous watching is essential to bring the total cost within the target cost.

~ 20 ~

1.6 Advantages and Disadvantages

a. It reinforces top to bottom commitment to process and product innovation to achieve some competitive advantages. b. It helps to create a company’s market-driven management for designing and manufacturing products that meet the price required for the market success. c. It uses management control system to support and reinforce manufacturing strategies, and to identify market opportunities that can be converted into real saving to achieve the best value for money rather than simply achieving the lowest cost. d. Assures that products are better matched to their customers’ needs. e. Aligns the costs of features with customers’ willingness to pay for them. f. Reduces the development cycle of a product. g. Reduces the costs of products significantly. h. Increases the teamwork among all internal organizations associated with conceiving, marketing, planning, developing, manufacturing, selling, distributing and installing a product. i. Engages customers and suppliers to design the right product and to more effectively integrate the entire supply chain.

~ 21 ~

~ 22 ~

1.7 Features of Target Costing

1. Target Costing is viewed as an integral part of the design and introduction of new products. As such, it is part of an overall profit management process, rather than simply a tool for cost reduction and cost management. The first part of the process is driven by customers, market and profitability considerations. Given that profitability is critical for survival, a target profit margin is established for all new product offerings. The target profit margin is derived from the company’s long term business plan. Each product or product line is required to earn at least the target profit margin. 2. For any given product, a target selling price is determined using various sales forecasting techniques. Critical to setting the target selling price are the design specifications which reflect certain levels of functionality and quality of the new product. These specifications are based on customer requirements and expectations and are also affected by the products of competitors. Importantly, while setting the target selling price, competitive conditions and customers’ demands for increased functionality and higher quality, without significant increase in price, are clearly recognized, as charging a price premium may not be sustainable. Hence, the target selling price is market driven and should encompass a realistic reflection of the competitive environment. 3. Integral to setting the target selling price is the establishment of target production volumes, given the relationship between price and volume. The expected target volumes are also critical to computing unit costs, especially with respect to capacity-related costs (such as tooling costs), as product costs are dependent upon the production levels over the life cycle of the product.

~ 23 ~

Once the target selling price and required profit margin have been determined, the difference between these two figures indicates the allowable cost for the product. Ideally, the allowable cost becomes the target cost for the product. 4. The next stage of the target costing process is to determine cost reduction targets. Some firms will do this by estimating the “current cost” of the new product. The current cost is based on existing technologies and components, but encompasses the functionalities and quality requirements of the new product. The difference between the current cost and the target cost indicates the required cost reduction that is needed. This amount may the divided into target cost reduction objective and a strategic cost reduction challenge. The former is viewed as being achievable (yet still a very challenging target), while the latter acknowledges current inherent limitations. After analyzing the cost reduction objectives, a product level target cost is set which is the difference between the current cost and the target cost reduction objective.

5. A fair degree of judgment is needed where the allowable cost and the targeted cost differ. As the ideal is to produce at the allowable cost, it is important that the difference is not too great. Once the product level target cost is set, however, it generally cannot be changed, and the challenge for those involved is to meet this target. 6. Having achieved consensus about the product-level target cost, a series of intense activities commences to translate the cost challenge into reality. These activities continue throughout the design stage up until the point when the new product goes into production. Typically the total target is broken

~ 24 ~

down into its various components, each component is studied and opportunities for cost reduction are identified. These activities are often referred to as value engineering (VE) and value analysis (VA). VE involves searching for opportunities to modify the design of each component or part of a product to reduce cost, but without reducing functionality or quality of the product. VA entails studying the activities that are involved in producing the product to detect non-value adding activities that may be eliminated or minimized to save cost, but without reducing the functionality or quality of the product.

Where components are sourced from suppliers (which are often the case in automobile industry) target prices are established for each part and the company’s employees work with the suppliers to ensure that the targets are achieved. Overall the aim of the process is to ensure that when the production commences, the total cost will meet the target, and profit goal will be achieved. There is also a continuous improvement programme, known as kaizen costing, that focuses on the reduction of waste in the production process, thereby further lowering costs below the initial targets specified during the design phase. 7. To achieve the objectives of target costing, a team-based set up is required that integrates essential disciplines such as marketing, engineering, manufacturing, purchasing and finance. 8. To be successful at target costing, management must listen to the company’s customers. What products do they want? What features are

~ 25 ~

important? How much are they willing to pay for a certain level of product quality? Management needs to aggressively seek customer feedback, and then products must be designed to satisfy customer demand and be sold at a price they are willing to pay. In short, the target costing approach is market driven.

Design engineering is a key element in target costing. Engineers must design a product from the ground up so that it can be produced at its target cost. This design activity includes specifying the raw materials and components to be used as well as the labour, machinery, and other elements of the production process. In short, a product must be designed for manufacturability.

Every aspect of the production process must be examined to make sure that the product is produced as efficiently as possible. The use of touch labour, technology, global sourcing in procurement and every aspect of the production process must be designed with the product’s target cost in mind.

Manufacturing a product at or below its target cost requires the involvement of people from many different functions in an organisation: market research, sales, design engineering, procurement, production engineering,

production

scheduling,

material

handling

and

cost

management. Individuals from all these diverse areas of expertise can make key contributions to the target costing process. Moreover, a crossfunctional team is not a set of specialists who contribute their expertise and then leave; they are responsible for the entire product.

~ 26 ~

1.8 Problems of Target Costing

1. Team and cross-functional impediments: According to the specialists, the logical “target costing” approach is very easy, based on continuous communication and information between supply chain partners. In opposite, the “cost plus” approach is based on lack of communication between supply chain partners. The correct implementation of Target Costing employs all key functions in the entity and requires an extended time horizon and commitment. The implementation of Cost plus approach only employs part of the company’s key functions and requires reducing time horizon and commitment. The previous transfer of internal functions to partners or outsourcing may represent risks, due to inability to monitor or control the outputs of the desired function. When these functions are carried out at the manufacturer’s plant, expectations and standards are communicated and understood, but often these communications are lost when the function is transferred to one of the supply chain partners. One way to control this problem is to place one of the producer’s employees in the supplier’s plant, to monitor and assist with the supplier’s activities. Based on the above, the employees will learn faster and will understand costs better, and the company’s management will adopt Target Costing Method as a faster information flow, with more frequent reporting.

2. Insufficiency of accounting data: According to specialists, learning is due to the use of management accounting information systems based on advanced production technology. Thus, learning the Target Costing method and the rapid supply of information have lead to its rapid learning and expansion

~ 27 ~

among companies. Most specialists even suggest adoption of Target Costing method in entities that still use widely less advanced management accounting information systems, as it is a support for learning that includes accounting information on planning, control, production, budgets, forecasts, and performance standards. In this way, the information received helps to improve existing ones, thus achieving and developing a corporate learning system. Below we have described some differences between Target Costing method and the other existing traditional cost calculation methods used worldwide: 

The

Target

Costing

approach

may

be

used

in

combination/integration with Activity-Based Costing or Balanced Scorecard or other approaches; in comparison with Target Costing there are many Romanian economic entities which are using traditional management accounting and cost calculation methods (especially standard cost method or variable cost method).

3. Details of the production process: According to the specifics of the Target Costing, the designing process must be broken down into components to the lowest level. Any producer should analyze the diagram of the entire production process to its lowest level, even to the bill for materials, in order to find areas for improvement or connections that can be made by removing certain irrelevant components for the production process and for product assembly, by complying its technical specifications, as well as other ways to save a processing stage. These details require direct involvement in the production process, product design and marketing. implementation and execution becomes difficult. The aim of the target costs concept is to make minor changes and not major innovation

~ 28 ~

Chapter 2 : RESEARCH METHODOLOGY

2.1 STATEMENT OF THE PROBLEM For the comprehensive growth and specialization in all aspects of the processes and product component it is necessitated that all segments of the firm be induced to work in acquiring the required level of spirit and state. While reviewing the literature it was found that various studies have been conducted in the field of kaizen, target costing and cost management. In the implementation of kaizen and target costing techniques, it requires that the organization fabric be prepared to the desired level to grasp the philosophy. It was found that the studies on these aspects are much less conducted. It creates the scope for the research on this aspect. The philosophy of kaizen and target costing is a comprehensive approach which includes change in the thought process of members of the companies, zeal on the part of the members, real support of the managerial positions and so on. This is a research problem together with other aspects. After adopting the kaizen and target costing techniques how far and the extent the companies are able to achieve its competitive edge in the market.

~ 29 ~

2.2 SCOPE OF THE STUDY This study is an attempt to study various issues and aspect of kaizen costing and target costing. Though cost management includes activity based costing, standard costing ,uniform costing, direct costing, marginal costing, kaizen costing, target costing , out of which some are referred as traditional techniques and the remaining are modern techniques.

In order to determine a target cost model, one must first comprehend the different types of models under study. Regression models are classified into two broad categories namely linear and non-linear models (Rajarathinam and Parmar, 2011). These models are considered to be veritable tools for describing the functional form of the relationship between variables (Okereke, 2011). Many commercially available cost estimating packages use weight as the baseline cost driver and then generate measures of secondary cost drivers to refine the cost estimate (Curran et al, 2006). However, it is difficult to get an accurate weight estimate early in the early conceptual stages of design. To achieve the purpose of this study, two types of parametric equations will be presented and compared. They are the linear and non-linear regression models. Simple Linear Regression Model The simple linear regression model (SLRM) assumes that the relationship between the dependent variable, denoted y, and the independent variable, denoted x, can be approximated by a straight line. Linear regression is the most common method of studying the linear relation between two or more variables (Kahyani and Basiri, 2011). The simple linear regression (SLR) is

~ 30 ~

utilized when only one proven CER is capable of determining the cost of the product. The SLR model can be denoted as follows.

y  0  1X1 y, Target cost β0, Intercept β 1, Regression coefficient X1, Specified cost driver Ordinary least squares (OLS) is a method for estimating the unknown parameters in a linear regression model. OLS calculates the straight line through the data set which minimizes the error. When there is more than one CER present in the regression model and the data set continues to portray linear tendencies, the multiple linear regression model will be utilized. Multiple Linear Regression Model In the event that a cost model depends on several parameters, the multiple linear regression model (MLRM) will be employed. According to Kutner et al. (2004), this type of parametric model is considered complex. The relevant formula for the MLRM is as follows:

y ˆ y  0  1X1  2X2 ... k Xk where ŷ, Target cost β0, Intercept

~ 31 ~

βk, Regression coefficients Xk, Specified cost drivers In order to calculate the regression coefficients, the method of least squares was used.

Least Squares Estimation In least squares estimation (LSE), the unknown values of the parameters are called regression coefficients. The German scientist Karl Gauss (1777-1855) proposed estimating the regression coefficients to minimize the sum of the squared deviations between the observed data and the estimated data (Montgomery, 1984). This method can be observed in the below figure.

Deviations of the data from the estimated regression model The necessary derivation of calculating LSE can be demonstrated in the work of Faraway (2000), Montgomery (1984) and Sajadifar and Allameh (2008). The equation to calculate regression coefficients is listed below.

~ 32 ~

  (X X) 1 X  y

Sajadifar and Allameh (2008) have created different methods to compute multiple linear regression coefficients. They have modified the existing way to compute regression coefficients in order to make the computation more efficient. The least squared method of estimating the parameters under study is typically preferred to other methods as it yields unbiased estimators (El-Salam, 2011).

Multiple Non Linear Regression Model

In statistics, the NLM is a type of regression that utilizes data modeled in the form of non-linear combinations (Montgomery, 1984). As stated in the literature review, Salam (2009) introduced a parametric model based on estimating design effort of a compressor fan at Pratt & Whitney Canada.

Upon examining the above nonlinear model, the natural log (ln) of the entire equation must be taken in order to be in the proper format for regression analysis. The assumptions of linearity hold when manipulating the data from non linear to linear. It is much easier to use simple linear regression to estimate the parameters of a nonlinear model. Following the data linearization, evaluating the model parameters concludes direct solutions using the least squares method. The linear equation generated is described as follows and takes into account the error (ε) that postulates to the logarithmic.

In regression, it is known that the units of the dependant variables do not equate that of the independent variable. Therefore, the concept of dimensional analysis needs to be addressed. Dimensional analysis

~ 33 ~

(DA) is a method employed to restructure the variables of a regression model into a set of independent dimensionless products (Vignaux and Scott, 1999). DA explicitly uses the constraint that all terms of the model must have the same dimension. Parametric equations utilize DA as it takes multiple cost estimating relationship of any dimension, transforms them into dimensioned constants, called regression coefficients, to form a correct dimensional relationship that is homogeneous. Dimensionally homogeneous implies that all the dimensions of all terms are the same. DA divides each factor by the reference value in order to generate dimensionless units.

For example, an equation can be composed of multiple dimensions including mass with dimension [M], time [T] and volume [L]. Moreover, problems in economics often add a dimension of cost to the function. The regression analysis process takes the multiple variables and creates dimension constants, also known as regression coefficients, to form a dimensional relationship. These dimensionless products of the variables can be explained by the Buckingham Π theorem (Buckingham, 1914). Buckingham (1914) stated that all relationships can be reformulated as a function of a set of dimensionless products of the variables. The Buckingham Π theorem is a process that satisfies physical dimensional homogeneity which involves n multiples variables and reduces them to Π dimensionless variables (Bender, 1978). The following will be explained by the two subsequent formulas.

The original relationship of multiple dimensioned variables is written as the following equation, where n represents the number of dimensions.

F(X1,X2,……Xn)=0 If the above formula is dimensionally homogeneous, the Π theorem states that the expression can be expressed as a new function of a

~ 34 ~

set of dimensionless parameters written in terms of Π’s. Here m represents the number of fundamental dimensions in the relationship. The above formula represents a new function that is equivalent to the old one with fewer variables. Further research about DA and Buckingham Π theorem can be found in the work of most applied mathematics and physical modelling textbooks such as; Langhaar (1951) and Huntley (1967). The units are removed using the concept of DA furthermore they are still considered linear. Evaluating the suitability of the data is essential to postulate their linearity assumptions. Below are three techniques used to validate the linearity assumptions of a given data set.

Data Linearity Assumption: As stated by Okereke (2011), linear regression models are those that are linear in parameters. Data linearity is a critical parameter in regression. In order to determine if the function is linear, several key assumptions must be satisfied. These assumptions are listed below.

1. Linear Relationship 2. Multicollinearity 3. Homoscedasticity

Linear Relationship: Firstly, linear regression requires that the relationship between the dependant and independent variables to be linear. It is important to validate all the data points and determine if there are any outliers. Scatter plots or residual plots can test the linearity of a data set.

~ 35 ~

These plots allow for visual assessment of the relationship between the response and predictor variable (Weisberg, 2005).

In the case of the scattered plot, the standardized residual plotted against the non-standardized predicted value will determine if the data set is linear.

The scatter plot should not have any visual curvilinear patterns as demonstrated in the below figure.

Linear scatter plot

In the case where the data set renders a curvilinear pattern, the linearity test fails. A curvilinear representation can be seen in the following figure.

~ 36 ~

Non- linear scatter plot

The residual plot is a graph which demonstrates the residuals on the vertical axis and the independent variable on the horizontal axis. The difference between the observed value and the predicted value is called the residual. Below is a formula to calculate the residual error.

Residual = Observed value – Predicted value

If the points in a residual plot are randomly dispersed around the horizontal axis, a linear regression model is appropriate for the data, indicating a good fit for a linear model. Otherwise, a non- linear model is more appropriate. The residuals of a regression can be tabulated in the following example based on a sample size of ten units.

~ 37 ~

Example of Table of Residual Observed Value Predicted Value Residual(error)

Sample 1

98

103

-5

Sample 2

102

99

3

Sample 3

102

104

-2

Sample 4

102

97

5

Sample 5

105

105

0

Sample 6

103

101

2

Sample 7

98

97

1

Sample 8

103

101

2

The linear regression model can be employed if the linearity assumptions are satisfactory. To test the linearity of a data set, the scatter plot of the standardized residual against the predicted values is required. Figure 10 demonstrates a random dispersion of data which represents a linear relationship. The residual plot of the aforementioned example is displayed below.

~ 38 ~

Example of residual plot

Multicollinearity: Secondly, linear regression assumes that there is little or no multicollinearity in the data. In regression analysis, it is expected to have dependencies between the response variable and the regressor (Montgomery and Runger, 2007). Multicollinearity occurs when two or more predictors in the model are correlated and provide redundant information about the response. Multicollinearity can have serious effects on the estimate of the regression coefficients and on the general applicability of the estimate model. Multicollinearity can be tested several different ways, such as: the correlation matrix and the variance inflation factor (VIF).

The correlation matrix must yield correlation coefficients smaller than 1 in order to assume that multicollinearity is not present. Correlation can be interpreted as a statistical measurement of the relationship between two variables. Possible correlations range from +1 to –1. A correlation of zero indicates that there is no relationship between the variables. A correlation of +1 indicates a perfect positive correlation, meaning that both variables move in the same direction together. The correlation matrix will be demonstrated in the data analysis section.

~ 39 ~

The VIF of the linear regression must render a value smaller than 10 which determines if multicollinearity is present (Kutner, 2004). If a VIF greater than 10 is yielded, there is certainly multicollinearity in the data set. The equation for the variance inflation factor is shown below

VIF = 1/(1-R2)

where, VIF, Variance inflation factor R2, Coefficient of determination As can be seen in the work of Faraway (2000), who worked on a multicollinearity example on employment within a population, rendered a VIF value of 42. This value can be interpreted that the standard error is 42 times larger than it would have been without the presence of multicollinearity. To remove the presence of multicollinearity, examine the correlation matrix and remove the variables that do not have a large pairwise correlation with the other variables (Faraway, 2000). Thereafter, the process of calculating the R2 is repeated to determine the new VIF value with the omission of one less variable in the equation.

A stronger linear dependency of the independent variable(s) will enable a larger coefficient of determination. Hence, the VIF value will also yield a larger value. Montgomery and Runger (2007) discuss in detail about the presence of multicollinearity and different measures for solving this issue.

~ 40 ~

Homoscedasticity: The last data linearity assumption is homoscedasticity. Homoscedasticity can be observed when the data set exhibits similar amounts of variance across the range of values for an independent variable (Kim and Bentler, 2002). Equal variance is essential across the data for the linearity assumption to hold since the variance measures the dispersion of a set of data points around their mean value. The equation to calculate the variance is as follows;

σ2 = Σ (x - μ) 2 / N

~ 41 ~

where σ2, Variance Σ, Sum Observed data point x ,,

Mean of data set

, Number of data points N

The scatter plot is an excellent validation tool in order to determine whether the data demonstrates homoscedasticity. As can be seen in the following example, the variance across the data set is very similar, represented by the grey bars, which deems that the data set displays linear tendencies.

~ 42 ~

In this study, the three linearity tests will be conducted in the case study in Chapter IV. In order to comprehend which parametric model will generate

a

more

accurate

target

cost

between

both

models

aforementioned, the percent errors will be compared and two different methodologies will be employed to determine the significant factors.

The reliability of the models will be validated by two different methodologies to determine which parameter(s) become significant and will therefore be used to determine the cost. These methodologies are the analysis of variance and path analysis. It should be noted that these methodologies take into account the cost estimating relationships. The requirements to determine the most suitable regression model along with the ideal technical parameter(s) will be discussed in the following segment.

Cost Estimating Relationship: According to ISPA, cost estimating relationships can be defined as a mathematical expression of varying degrees of complexity expressing cost as a function of a cost driving variable. Cost drivers are any factors which cause a change in the cost of work performed in the lifecycle of a product. The identification and selection of cost drivers are fundamental to a cost estimating model. Without adequate data and CER, the cost models will have no added value in the early conception phase of estimation. The CER is an integral part of regression analysis and its validation is crucial.

~ 43 ~

CER Validation: Once the data collection and validation has occurred, it is imperative to pair the raw data with valid CERs. The ultimate test of the goodness will determine whether or not a particular CER can accurately predict the cost of a component. Several mathematical tests are available to comprehend the most significant CER. These tests are as follows:

1. Standard error of estimate 2. Coefficient of correlation 3. Coefficient of determination

The standard error of estimate measures the accuracy of the prediction. It can also be termed as the standard deviation of the data set. Therefore, if the data demonstrates large dispersions, a higher SEE will be calculated. This represents that the data set in the study tends to be far from the regression line. SEE uses the regression line that minimizes the sum of squared deviations from the prediction. The standard error of estimate’s equation is listed below.

𝑆𝑒𝑒 = √∑(𝑦 − 𝑦)/𝑁 Where SEE, Standard error of estimate ∑ Sum Y, actual result Ŷ, predicted result N, number of data points

~ 44 ~

The coefficient of correlation (r) and the related coefficient of determination (R2) are certainly the two most commonly used measure of goodness of fit. The value of r is calculated from the following equation.

𝑟 = ±√𝑟 These statistical results measure the amount of correlation between the prediction and the corresponding data set. Moreover, it is an excellent indicator of the accuracy and denotes the strength of the equation. For example, if the coefficient of correlation is r = 0.875, then the coefficient of determination is R2 = 0.766. This represents that 76.6% of the total variation can be explained by the linear relationship in the regression. The remaining 23.4% of the total variation in the equation remains unexplained. Hence, a high coefficient of determination yields a minimum amount of disparity in the equation. A metric for determining the proportion of the variation explained by the independent variables is the coefficient of determination (Montgomery et al., 2001)

There are multiple methods to determine the validity of a cost driver. The following section describes quantitative metrics to establish the significance of a CER.

CER Significance: The significance of the CER generates confidence in the regression equation and the assurance of its forecasting capability. It is known that there are numerous ways in statistics to evaluate the significance of a cost estimating relationship. The below table demonstrates several key criterion when determining the ideal CER.

~ 45 ~

Criteria P-value of the f-test P-value of the t-test Coefficient of Variation (CV) R-squared

Good ≤ 0.10 ≤ 0.10 ≤ 0.25

Marginal ≤ 0.15 ≤ 0.15 0.25 to 0.30

≥ 0.70

0.35 to 0.70

A caution is warranted when performing statistical analysis of a relationship. There is no one statistic that disqualifies a CER or model, nor is there any one statistic that validates a CER or model. The math modeling effort must be examined from a complete perspective, starting with the data and logic of the relationship.

CER Strengths: CERs can be excellent predictors when implemented correctly, and they can be relied upon to produce quality estimates when used appropriately. Several CER strengths pertaining to estimation are as follows:

1. Capability to reduce the amount of time to evaluate cost estimates 2. Ability to develop and produce prompt estimates 3. Minimal information is required concerning the product estimated 4. Practicability when estimating in early conceptual phase of a program.

~ 46 ~

CER Weaknesses: On the other hand, CERs must be used carefully and the process of selecting the ideal CER is essential. Below are certain examples of cost estimating relationship weaknesses.

1. Performing a detailed estimation can be more reliable than CER estimation 2. Employing incorrect cost or technical data may skew the chosen CER

Analysis of Variance: The analysis of variance (ANOVA) can be used to test the significance of regression analysis. According to Montgomery (1984), analysis of variance is a method of decomposing the total variability in a set of observations, as measured by the sum of the squares of these observations from their average, into component sums of squares that are associated with specific defied sources of variation. In order to judge the adequacy of a regression model, the coefficient of determination is utilized. The coefficient of determination represented by R2, measures the percentage of variation explained by the model between 0% and 100%. A marginal result can be found between 35% and 70%, however above 70% is considered good (ISPA, 2011). Additionally, the t-statistic and the related p-value are both important in this methodology since it estimates the probability level at which the statistical test would fail, suggesting the relationship is not valid. As per the ISPA, a p-value less than 0.10 is considered acceptable for inferring that the selected factor remains a significant cost driver. Moreover, a p-value less than 0.15 is deemed marginal. If the CERs are considered insignificant, the analysis is repeated by removing one CER at a time and the analysis is continuously repeated until only significant factors remain. Furthermore, the confidence interval

~ 47 ~

utilized for significance in this study is set at 90%. To do so, the probability level at which the statistical test would fail, suggesting the relationship is not valid should have a p-values less than 0.10. The analysis of variance will be employed to determine which CER(s) become significant. The ANOVA methodology will be compared to the path analysis which takes into account the effects of each parameter on the outcome.

Path Analysis: Wright (1934), known for his influential work on path analysis (PA), takes into account the approach used to study the direct and indirect effects of variables. This methodology analyses the effect that each of the parameters will have on the output of the equation. By examining the possible linkages between each potential cost driver, their respective path coefficients will determine a standardized method to conclude the significant factor(s). PA is designed to produce measures of relationship between variables (Smith and Murray, 1978). The path analysis process begins by determining the correlation between the potential cost drivers. In statistics, the correlation indicates the strength and direction of a linear relationship between two random variables, which is essential to comprehend their associations (Kutner et al, 2004). The higher the correlation, the better the effect of the CER will pertain to the regression. The regression coefficients, Bi, are generated in the summary output of the ANOVA. Thereafter, they are utilized in the below equation to determine their path coefficients. It should be noted that path analysis is not merely an ordinary regression analysis. Path analysis employs regression analysis to compute path coefficients. In path analysis, an equation represents a causal link whereas

~ 48 ~

in regression analysis, an equation represents the dependant variable as a function of the independent variable (Smith and Murray, 1978). Lastly, the value of U in relationship with the final equation is calculated, which represents the uncorrelated residual of the function (Li, 1975). A path diagram is a graphic display of the order in which variables are assumed to affect on another. The path analysis diagram is shown below.

Path diagram Land (1969) describes the convention for drawing path diagrams. It consists of arrows that can be drawn from variables acting as causes to variables acting as effects. The initial variables are linked to one another by curved lines with double arrows. Then the initial variables are linked to the output by a straight arrow and finally the output is linked to the uncorrelated value of the function by a straight arrow. The path analysis’ strengths and weaknesses are as follows.

~ 49 ~

Strengths of Path Analysis: Path analysis provides the study with unique advantages not available with other methods, such as;

1. Provides means for modeling of complex problems: Path analysis provides means of incorporating unobservable variable into evaluative studies. The resultant models are more representative of the dynamic reality (Smith and Murray, 1978). 2. Presents reflection of alternative models: Path analysis requires the explicit specification of presumed causal relationships and forces the researcher to consider several alternative models (Chamberlin, 1965). 3. Enables the study of both direct and indirect effects: Path analysis enables the study of both the direct and indirect effect on dependant variables by analyzing their correlations.

Weaknesses of Path Analysis: Some of the technical problems associated with the use of path analysis procedures are presented below: 1. Multicollinearity: No possibility to determine what proportion of the variance is accounted for by the variable when the data under study are interrelated. Multicollinearity often arises when multiple indicators are used in the regression for the cost drivers under study (Pedhazur, 1975).

2. Measurement errors: One of the assumptions in path analysis is the measurements are free of error. Otherwise a greater amount of ambiguity in the path coefficients will be inherent.

~ 50 ~

3. Complexity of interpretation: Path analysis models with four or more variables become increasingly complex to interpret. Moreover, the calculation of the coefficients and their interpretation can be tedious and inconclusive. In the end, path analysis provides a unique capability to study the direct and indirect effects of the variable on the output.

Limitation of CPM: 1. CPM operates on the assumption that there is a precise known time that each activity in the project will take. But, it may not be true in real practice. 2. CPM time estimates are not based on statistical analysis. 3. It cannot be used as a controlling device for the simple reason that any change introduced will change the entire structure of network. In other words, CPM cannot be used as a dynamic controlling device.

~ 51 ~

CHAPTER 3: LITERATURE REVIEW The Target Costing (TC) concept originated in Japan, emerging around the 60s (Feil et al., 2004). We can, however, state that it started before this time, since it stems from another American definition – value engineering. The value engineering technique, initially developed at General Electric, is a concept that tends to maximise product attributes while minimising their costs (Feil et al., 2004). During the 60s, value engineering was combined with the idea of influencing and reducing product costs as early as possible in the product planning and development process (Feil et al., 2004), because in this phase the majority of costs are fixed (Tani, 1995). This new concept was called genka kikaku and its first documented use in Japan took place in the well-known automotive company Toyota, in 1963, despite the inexistence of any mention in the literature to that fact until 1978 (Feil et al., 2004). Later, genka kikaku was translated into TC, the term currently known worldwide. Translating the phrase genka kikaku literally, gen means origin and ka price. Placing both together we obtain the phrase “price origin”. This directs us to the tight existing

link

between price and cost, and

we can therefore interpret this term in the sense that “the origin of price” is cost. Kikaku means plan. So, “cost plan” expresses the global approaching strategy to cost reduction which is a characteristic of the TC concept (Feil et al., 2004). Several authors define TC in different ways, but all of them share three common characteristics: market orientation, focus on product engineering, and focus on product functionality (Feil et al., 2004). Given the variety of existing definitions, the genka kikaku concept was determined and built from pieces of concepts and from its practical application in the competitive Japanese market. In its country of origin, TC is officially defined as a profit management process, by which factors such as quality, price, client trust,

~ 52 ~

delivery terms and times and others are set at the product planning and development stage. They are also set at levels that respect the needs of the consumer (Huh et al., 2008). In the 80s, the method was recognized as a significant factor in the development of Japanese companies, because it allowed the strengthening of their competitive position in the global market (Huh et al., 2008), giving rise to a curiosity regarding its use in other companies. In order to survive in the current market, companies are forced to become experts, and to develop products that offer quality and functionality to the consumer, while at the same time guaranteeing a desirable profit to the company (Cooper and Slagmulder, 1999). Cooper and Slagmulder (1999) consider that TC is a technique that allows the strategic management of future company profits. According to Cooper and Slagmulder (1999), the three components of the method – quality, functionality and perceived value – are essential for the success of the company. It has then to balance this tripod with the requirements imposed by the market and with the adopted strategy. The TC concept is characterized by starting with identifying the price consumers are willing to pay for the products, considering their quality and functionality (Albright and Lam, 2006). After establishing the target sales price, the company defines the desired profit margin. This margin – or target margin – can be determined according to the company’s strategy, profit expectations, historical results, competitive analysis and result simulations (Cooper, 1994). The difference between the target price and the desired margin is called target cost – a value used as a reference during the whole process of product conception, since it is used to control both the product’s design and production costs. Cost becomes therefore an input for the product development process, and not an outcome (Cooper and Slagmulder, 1999). According to Cooper (1994), after defining its target cost the company will assess its total production costs, which can never surpass the established target cost because it guarantees the product generates the intended profit, while applying the sales price accepted by the market.

~ 53 ~

When the product is in the conception stage and the previously defined target cost cannot be reached, the company use the value engineering (VE) tool. This is a multidisciplinary system that surveys the factors that may be influencing the cost of the product in order to allocate means to reduce and/or eliminate them without ever compromising the product’s quality and reliability (Cooper, 1994). Like TC, this tool is used during the product’s conception stage and has a crucial role in cost management by helping the company to manage the trade-off between functionality, quality and cost (Cooper, 1994). Another option to be assumed by the company is the cancelation of the project giving the production of the product (Helms et al., 2005). This can happen when the company considers the target cost cannot be reached and/or the profit margin does not ensure a satisfactorily return to the company. We estimate that 80% of product costs are tied up during the design stage, making it harder to obtain major cost reductions after this stage (Cooper and Chew, 1996; Davila and Wouters, 2004; Ansari et al., 2009). As main TC characteristics we can indicate the following six: price leads to cost, consumer focus, product design focus, multidisciplinary teams, focus on the costs incurred during the product’s lifecycle, and the involvement of the value chain as a whole (Huh et al., 2008). Regarding the first characteristic, the company starts to calculate the sales price in order to define the target cost. To establish the sales price the company has to consider several factors, such as the consumer traits, the product’s lifecycle, the estimated sales quantities, and the competitor’s strategy (Guilding et al., 2000). The price increases in the same proportion as the increment in the perceived value, i.e., the higher the perceived value seen by the consumer, the more the amount he is willing to spend (Modarress et al., 2005). The second characteristic – consumer focus – forces the company to adopt a market orientation without ever neglect quality, functionality, and product cost (Tani, 1995). The perceived value for the client regarding any incorporated characteristic and its functionality has to be greater than the cost the

~ 54 ~

company incurred into to provide such characteristics (Swenson et al., 2003; Ansari et al., 2009). The third characteristic – product design focus – demands it be sufficiently advanced so that quality and functionality can be adequately adjusted to consumer needs. The design changes need to happen before production starts. Then the company is able to reduce and/or avoid costs and decrease the market launch lead time (Swenson et al., 2003). Regarding the fourth characteristic – multidisciplinary teams – cooperation between the different departments is mandatory for TC success (Monden and Hamada, 1991), because the teams involved in the process are responsible for the complete integrity of the product, from conception to final production (Swenson et al., 2003). This multidisciplinary cooperation is essential because in order for the cost reduction to be effective the company has to balance all the new product targets – cost, quality, functionality, volume of production and invested capital (Everaert et al., 2006; Cooper and Chew, 1996). As a rule, the engineering team has the most influential role in this type of teams, unlike accounting whose opinion is seen as unimportant (Dekker and Smidt, 2003). TC’s fifth characteristic orders the control of the total costs of the product’s lifecycle during its duration. These costs include the purchasing price for the components, and operational, maintenance, distribution and product placement costs (Ansari et al., 2009). The TC requires a continuous estimation of the production costs, while the product moves along the conception stage, in order to analyze the impact caused by design decisions in the costs, and also as a way to monitor the progress towards meeting the goals related with expense reduction (Everaert et al., 2006). To make that estimation we need a detailed cost information, that has to be reliable and accurate (Everaert et al., 2006). During the product development process we also need to continuously compare the new product expense levels with the target cost to be met (Everaert et al., 2006).

~ 55 ~

Regarding the sixth characteristic, TC demands that all members of the value chain, such as suppliers, service providers, distributors and consumers are included in the process (Ansari et al., 2009), thus focusing the entire chain in the global objective of waste, excess and irregularity elimination (Helms et al., 2005). One of the requirements of this method is that the company must involve its suppliers as partners in the product conception stage, when the target costs are defined (Cooper and Chew, 1996). When selecting its suppliers, since all suppliers will be involved in the whole development process of a new product, the company prefers relevant factors such as trust, cooperation, ability to produce quality parts, the amount of engineers and design experts employed, the reputation held within the industry by the services provided, and dependability (Helms et al., 2005). This relationship between the company and its suppliers generates the sharing of information on cost reduction (Helms et al., 2005), and on production methods and techniques essential to TC’s success, besides ensuring that both the supplier and the buyer will meet their target cost (Ellram, 2000). Workers are the ones that present the best ideas for the company to reach a continuous improvement. This happens because they are the ones closer physically to the jobs and see, primarily, the type of faults that happen and their main causes (Cooper, 1996). The company must therefore motivate all its employees to meet the target cost during the product conception stage, using their creativity to create alternative plans that allow for a bigger cost reduction, and making TC an activity that generates profits for the whole company (Ibusuki and Kaminski, 2007). For this, employees need access relevant information in order to better reach the proposed goals (Modarress et al., 2005). When there is such information, co-workers internalize know-how faster and there is a greater understanding of costs and of the organization in itself (Helms et al., 2005). Co-worker training and instruction is of the highest importance in any TC initiative, because a lack of knowledge over the process can pose a barrier to its implementation (Ansari et al., 2009; Helms et al., 2005).

~ 56 ~

REVIEW OF EMPIRICAL STUDIES The origin of TC has intrinsically to do with the Japanese automotive market characteristics and environment. When Toyota decided to include this method in its product conception process (Feil et al., 2004), the automotive market was going through profound changes in its surroundings. This industry was witnessing a shortening of the product’s lifecycle, demand diversification, and the existence of a sharp competition (Monden and Hamada, 1991). Besides, Japan was suffering the consequences of an appreciation of its national currency - yen (Monden and Hamada, 1991). As a consequence the sector searched for new cost management methods that would be useful for manufacturing new products. They intended thus to satisfy demand and client imposed specifications at the lowest cost, as well as reducing costs in existing products by eliminating waste (Monden and Hamada, 1991). TC was therefore a very useful tool. According to Monden and Lee (1993), TC conquered the industrial market by being an effective cost management system in the product’s design and development stage. The recognized success in its country of origin, Japan, where it has a usage rate of 100% in the automotive sector (Böer and Ettlie, 1999), restates its suitability to the industrial area. With the goal to analyze the most important target costing characteristics, according to the manufacturer, we analyzed the articles presented in Table 1, published during the period between 1995 and 2002, which report the use of this method in six automotive companies (Toyota, Nissan, Ford, Chrysler, Daihatsu, Fiat), from different geographic origins. The literature review allows the identification of six characteristics associated with TC: price leads to cost; consumer focus; product design focus; multidisciplinary teams; focus on the costs incurred; and involvement of the value chain as a whole.

~ 57 ~

Target Costing Characteristics in Automotive companies The Toyota Motor Corporation, the largest and most successful Japanese automotive company is known for its corporate innovation, its function integration ability and its internal philosophy (Hopper and Joseph, 1995). Its constant search for innovation led the company to adopt TC. This was done with the intention to improve its conception and production process, while facing the Japanese economic recession that was happening at the time (Monden and Hamada, 1991). In the Bhimani and Okano article (1995), the authors study Toyota in the United Kingdom – Toyota Manufacturing UK Ltd. – and have identified two TC characteristics present in the company: price leads to cost, and the involvement of the entire value chain. The same authors mention that the company calculates its target cost through the difference between sales price and the required profit margin. The first step starts with the definition of the sales price, for which the company takes into consideration the European sales estimates and the prices of cars previously marketed or currently on the market (Bhimani and Okano, 1995). Later the desired profit margin is removed and the target cost to be met is obtained. The importance of the role performed by the suppliers in the product conception stage is also mentioned. The supplier selection process must be approved by the mother-company, based in Japan, and by Toyota UK (Bhimani and Okano, 1995). Suppliers are selected according to the quality and functionality of the prototypes supplied, but they are also expected to get involved in the product conception and to give ideas that allow cost reduction and operational improvement (Bhimani and Okano, 1995). There is consequently a high know-how exchange between the suppliers and the company, because Toyota makes a point of promoting training sessions for the suppliers that explain the whole company production process, from the materials used to the implemented operational system (Sako, 2004). The role undertaken by the co-workers is also important for the organization, since it promotes

~ 58 ~

employee involvement in the conception and production process through suggestions that allow product improvement (Hopper and Joseph, 1995). The Carr and Ng (1995) article has as main theme the study of the company Nissan Motor United Kingdom, a subsidiary of the mother company Nissan Motor Company, Ltd., the second largest Japanese automotive company. In what regards TC, the authors managed identify four characteristics: price leads to cost, multidisciplinary teams, control of costs incurred, and involvement of the value chain. Regarding the definition of the product unit sales price, the company defines it according to information picked up in the market. The margin is defined considering the company’s activity plans, and the target cost is calculated as being the difference between the sales price and the margin (Carr and Ng, 1995). The authors also mention the existence of multidisciplinary teams formed by co-workers originally from departments such as Quality, Design, Engineering, Buying and Finances (Carr and Ng, 1995). Authors say that about 80% of the production costs concern materials and components used in the product, thus giving greater importance to cost controlling. Regarding the value chain, Nissan UK makes sure suppliers and co-workers are involved in the product conception and design process. Its co-workers are regarded by the company as a source of innovation, since they know the product specifications as well as their jobs, and consequently their cost reduction suggestions are more effective (Carr and Ng, 1995). Dealings with suppliers are tightly controlled since their supplies represent 80% of total costs. However, the trade relation existing between both parties is based on trust, rather than in pressuring them to respect prices and margins (Carr and Ng, 1995). Nissan bets on the know-how exchange with its suppliers, and in the promotion of workshops as a way to increase its performance regarding themes such as quality, productivity and efficiency (Carr and Ng, 1995). The suppliers are selected according with the criterion called QCDDM – Quality, Cost, Delivery, Development and Management – and their performance on these five components is evaluated (Carr and Ng, 1995:357).

~ 59 ~

In the article presented by Mintz (1995), the company being analyzed is Ford Motor Company, the second largest American automotive company. Mintz (1995) identifies two method characteristics present in the company: multidisciplinary teams and the involvement of the value chain. The company develops its entire car and truck design process having always under consideration the price given by its suppliers (Mintz, 1995:29). This happens because the company considers that a lack of careful attention to the main costs leads to a complete method failure. Ford also uses the concept of multidisciplinary teams making them up with people from different internal areas of the company, and giving them the responsibility to create the product (Mintz, 1995). According to the author, Ford sees its co-workers as fundamental resources for the company. It takes a chance on them and finances their training (Mintz, 1995). The company encourages its coworkers to rotate between departments and also to benefit from an international professional experience (Mintz, 1995). We should also mention that the company has a huge reputation for financial training, an area where the company makes sure its co-workers receive years of intense training (Mintz, 1995). The Chrysler Group LLC is known worldwide as an American automotive company, with headquarters in Detroit, Michigan (Dyer, 1996). The article of Dyer (1996) is centered in the relationship established between the company and its suppliers, and in the description of the SCORE program. This program was developed by the company and its goal is to improve the relations between the company and its suppliers. The same author is able to identify two TC characteristics in the company: multidisciplinary teams and the involvement of the value chain. The SCORE program – Supplier Cost Reduction Effort – has as main goal a cost reduction that benefits both business partners without injuring the suppliers’ profits. It defines that the manufacturer is responsible for encouraging, revising and applying the suppliers’ ideas in the quickest way possible, while sharing the benefits generated by those ideas (Dyer, 1996). The same program predicts that the suppliers will offer suggestions that will translate in a cost reduction equal to

~ 60 ~

5% of that supplier’s sales to the company, without penalties in the case the goal is not reached. To make all this suggestion process easier, the company has decided to give their suppliers the possibility to submit their ideas and consulting their performance on an online platform (Dyer, 1996). As an example, during the first two years of application the SCORE program generated 875 ideas worth $170.8 million in annual savings for Chrysler (Dyer, 1996). Through this program, the company is able to deeply involve its suppliers in the conception process, and they both try as a partnership to find ways to decrease the production costs while sharing the resulting savings (Dyer, 1996). The responsibility for supplier selection is given to a multidisciplinary team that evaluates the suppliers’ ability to design and produced a specific component or system (Dyer, 1996). Considering their performance, Chrysler offers term contracts to the selected suppliers for a period usually corresponding to the lifecycle of a model of car, or more – an average of 4.4 years (Dyer, 1996). The company also makes a point in maintaining a high level of communication with the supplier, having for that purpose created a common e-mail account that makes information sharing easier (Dyer, 1996). Chrysler considers therefore that TC has completely changed the relation of the company with its suppliers, by focusing it on cost instead of on price. This change in focus allows the company to work together with its suppliers in order to respect common costs and functional goals, making it also possible to share a relation of trust (Dyer, 1996). The article also refers that the company has introduced the concept of multidisciplinary teams, which allows all work stations to work together in order to develop the best product possible at the best price. Chrysler sustains that these teams fuel continuity, coordination and trust, both at an internal level and with the suppliers (Dyer, 1996), as well as innovation and problem solving, creating a faster and cheaper product development process. Consequently, all these improvements lead the company to produce components of the highest quality at a lower cost, while reducing the market launch lead time. From the development of a relationship with the supplier we can highlight four advantages (Dyer, 1996): reduction of the

~ 61 ~

product’s production cycle; reduction of the product’s total costs; reduction of supplier’s costs; increase in market share and profit. Daihatsu Motor Co. Ltd. was established in 1907 and is currently defined as a mini-cars manufacturer, partly owned by Toyota (Lee and Monden, 1996). This organization has headquarters in Osaka, Japan, and is already present in more than 120 countries in its market segment (Lee and Monden, 1996). TC is seen by the company as a specific manifestation of an improvement and operational control system (Lee and Monden, 1996). The characteristics – price leads to cost, focus on the costs incurred and involvement of the value chain – are identified as the most essential by the company (Lee and Monden, 1996). At Daihatsu, the method is strongly related to the long term planning process of products and returns. This connection allows the company to adopt a strategy that joins the two mentioned components (Lee and Monden, 1996). As such, the target cost value results from the difference between the sales price and the desired margin, and represents the top management desired goal. The target profit margin is defined for each period and each car model. Because it’s easy to calculate and given the already mentioned focus on profit, the return on sales ratio for similar products is used to define the margin (Lee and Monden, 1996). Considering the control of costs during the product’s lifecycle the estimated cost is controlled and reduced using the value engineering technique (Lee and Monden, 1996). Daihatsu also has a co-worker encouragement policy that promotes their participation in the design process and in the development of new products, through suggestions that may add value to the product and lead to waste elimination (Lee and Monden, 1996). The article of Zirpoli and Caputo (2002) is centered in the study of the relation between the automotive company Fiat and its suppliers. Fiat, an acronym for Fabbrica Italiana Automobili Torino, is the largest Italian car manufacturer, with headquarters in the city of Turin. Regarding TC, the authors have managed to identify two characteristics of the method: the existence of multidisciplinary teams, and the involvement of the value chain.

~ 62 ~

The involvement of its main suppliers in the product design process has proven to be an essential strategy for the brand and has triggered a radical change in the entire value chain (Zirpoli and Caputo, 2002). In its relation with the suppliers, Fiat adopted a philosophy based in companionship, where it involves the supplier in a process of co-design and in the sharing of information (Zirpoli and Caputo, 2002). The supplier selection process is seen as crucial, because a less than good choice causes the company to enter into high costs. The company policy states that the choice of suppliers cannot be solely based on quality or on the price of the supplied component (Zirpoli and Caputo, 2002). Therefore the choice decision is based on two criteria: technical and political. First, the company assesses the presented components against about fourteen parameters that check the technical characteristics of the part in question and its consistence in terms of technology, cost and performance (Zirpoli and Caputo, 2002). The second step may be considered a political move, since the company analyses and assesses the supplier’s portfolio, hopping to balance the quantities bought from the suppliers while creating a controlled competition among them (Zirpoli and Caputo, 2002). The authors describe also that, as a rule, Fiat shares the suggestions and benefits resulting from the suppliers’ suggestions equally among them. In the case the suggestion emerges from Fiat, the benefit is given solely to the manufacturer (Zirpoli and Caputo, 2002). The suggestions proposed by the internal divisions or by the suppliers are immediately applied to the existing products. In the case there is a decrease in costs, the suppliers are awarded with an amount equal to 50% of the cost reduction during the first year the suggestion is applied (Zirpoli and Caputo, 2002). Therefore, the suppliers chosen by the company have an interest in showing efficiency and in constantly improving their performance in exchange for long term contracts that usually coincide with the vehicle’s lifecycle (Zirpoli and Caputo, 2002). The company also values geographical proximity with its supplier and uses money as an incentive to have them move to a closer location (Zirpoli and Caputo, 2002).

~ 63 ~

In March of 2000, Fiat established an agreement with General Motors (GM) in which both would acquire a set of raw materials and other goods and services from the same suppliers together. As a consequence the supplying market became more competitive, with suppliers from both companies noting a doubling in their competition and business possibilities (Zirpoli and Caputo, 2002). The authors also report the existence of multidisciplinary teams with people from different divisions and external suppliers being involved in the same team (Zirpoli and Caputo, 2002). Considering now the theme in hand – the most important TC characteristics – we note that in the automotive sector the type of existing relationship with the suppliers is crucial for developing the business. With a constant reduction in the product’s lifecycles, and with the commitment of about 80% of the costs in the conception stage, the focus on design, research and development and in production planning is crucial for the survival of the companies (Carr and Ng, 1995; Lee and Monden, 1996). According to the opinion of Toyota, Nissan, Chrysler and Fiat, supplier participation in the product conception process explains the reinvention of their own design and development process as well as the business (Bhimani and Okano, 1995; Carr and Ng, 1995; Dyer, 1996; Zirpoli and Caputo, 2002). Instead of a competitive and exhausting relationship, with pressure being exchanged between the parties, the nature of the cooperating relationship between the supplier and the company is characterized by concepts such as trust or dependency (Zirpoli and Caputo, 2002). Suggestions that add value to the product are requested from the key business suppliers. This allows the reduction of waste and the improvement of the production line. The benefits stemming from these improvements are usually shared between both parties (Zirpoli and Caputo, 2002; Dyer, 1996). Besides, the company ensures the establishment of long term contracts with its suppliers, with a more extended duration normally coinciding with the product’s lifecycle, as shown by Chrysler and Fiat (Dyer, 1996; Zirpoli and Caputo, 2002). The suppliers are thus encouraged to learn and absorb the organizational abilities specific to that company, and are then considered an integrating part of it

~ 64 ~

(Sako, 2004). The existence of multidisciplinary teams can also be seen in Nissan, Ford, Chrysler and Fiat. The mixing of people from various areas in the same team gives birth to more structured ideas, more suited to the main goal: the creation of value for the product.

~ 65 ~

CHAPTER 4: DATA ANALYSIS INTERPRETATION AND PRESENTATION

1. Data Collection: The data collection phase is one of the most critical and time consuming steps in producing an estimate. All parametric estimating techniques require credible data before they can be used effectively. Without credible data and data sources, the outcome of a cost model, which is the target cost, is irrelevant. The parametric techniques utilized in this thesis require two types of data collection. Firstly, historical cost data is gathered and assembled specifically for the current research project. The data is collected with the use of the contractual prices. Secondly, technical data is gathered seeing as it describes the physical, performance, and engineering characteristics of a product. According to Bengtsson and Sjöblow (2006), expert interviews are necessary to create understanding and avoid misunderstanding of the system under study. In essence, cost interviews were conducted with numerous supply chain agents and managers in order to normalize the cost in a comparable fashion. The cost data can be subject to a lot of variability due to the fact that the negotiations were performed on different programs, commercial versus business, and in different fiscal years dating back in the early 80’s. Therefore, negotiation techniques and business cases have drastically changes over the past 30 years, cost transparency increases as well as increases in the aircraft planning base which drives the cost of the system. Similarly, technical interviews with the engineering and advanced design teams were conducted in order to get multiples point of

~ 66 ~

for views of the different potential cost drivers to be studied. Interviewing multiples sources and triangulating the results is a key factor in order to build credibility in the technical and commercial data gathering phase. The interview grids can be seen in APPENDIX C. In order to make a successful study, once the technical parameters are assessed, the technical data must be collected and validated by numerous sources. It is known that the aerospace industry is subject to data scarcity. Bombardier Aerospace was launched in 1986 with the acquisition of Canadair Ltd. Hence, the data points collected are limited to the programs on hand. Knowing that a new aircraft program is launched every few years, data scarcity can be a predominant factor in this study. The cost data for this thesis was collected with the contractual agreement between the supplier and BA in a given year. The technical data are precise constraints that can be found in the technical requirements

document.

This

document

provides

technical

requirements of the commodity under study. All data points have been cross validated by many sources in the company. The assessment of data linearity is crucial in order to suggest a good fit a linear regression model. The validity and reliability of the data are an important factor due to the fact that it is used as a basis of the research.

2. Data Validity and Reliability: To assure the credibility of a cost model, special attention must be placed on the validity and reliability of the data. Data validity can be defined as the ability of accurately measuring information, whereas data reliability is best described as the degree to which the data are free

from

error

and

yield

consistent

results

(Eriksson

and

Wiedersheim-Paul, 2001)

The collected data must be adjusted for items such as production rate, improvement curve, and inflation. This is also referred to as the data

~ 67 ~

normalization process. According to Bengtsson and Sjöblow (2006), it is important to reflect upon which errors could have occurred while gathering information. In the aerospace industry, the main variability within the commercial data is due to the commodities negotiation strategy. Some commercial strategies involve paying the suppliers development costs up front in order to obtain a reduction on ship set price. Whereas, in some cases where funding is scarce, commodities tend to make the supplier absorb all development costs which yields a higher recurring price. The thesis takes into account as a baseline that all commercial agreements are treated equally and that the collected data is assumed to be fair. Due to confidentiality purpose, the data collected at Bombardier aerospace cannot be divulged to the public therefore data masking techniques were performed.

~ 68 ~

~ 69 ~

CHAPTER 5: CONCLUSION AND SUGGESTION When purchasing a new product, understanding its target cost is essential, as it will help the company in planning and during negotiations. Target costing focuses on managing cost and profitability in product development (Martin Carlsson-Wall, 2011) during the conceptual design phases, designers and decision makers often need to know accurate cost information to assess and compare multiple alternatives to determine a preferred design. Upper management needs to evaluate cost reduction possibilities and alternatives affecting system performance. Therefore, appropriate cost estimating models must portray accurate and robust cost estimates to support design to cost studies in the early conceptual phase of new programs. This study focused on target cost models and adapting this knowledge to a case study in the aerospace industry. A case study involving the MLG at Bombardier Aerospace served as proof in demonstrating which type of regression analysis improves the cost estimating accuracy. Even though this empirical study is for a specific application, the methodology utilized can be applied to various industries, as the model has the flexibility of being generalized. The study is concluded in the following sections.

5.1 Major Findings: The main purpose of this research is to explore target costing methodologies and how the concept can be utilized in the early stages of product development. This will guide designers at the early phases of complex products. Moreover, parametric equations, a subset of target costing was analyzed in order to obtain higher predictive accuracy of cost estimation and guide designers at the early design phases of complex products. It is understood that accurately estimating cost is not an easy task at the early

~ 70 ~

stage of complex product development when only a few conceptual attributes of the product are known. This study shows a comparative parametric modeling technique, between linear and non-linear regression, to estimate the target cost of a major commodity at Bombardier Aerospace. Three parameters were initially considered as potential CER’s for the MLG: Height, MTOW and Weight. As per the results shown in section 4, the MLG has a relatively high degree of correlation with the MTOW. The confidentiality of the data was appropriately masked to ensure that the data remains confidential. In conclusion, the non-linear regression model will generate a better accuracy to predict the target cost. This study shows that the overall performance of the NLRM is superior to that of the MLRM based on the following key criterion:

1. Lower percent error for the trial data (10 data points) 2. Lower percent error for the validation data (3 data points) 3. Acceptable coefficient of determination as per the ISPA

It has been proven by both methodologies; ANOVA and path analysis, that the maximum takeoff weight is the most predominant CER factor. This demonstrates that both methodologies will converge on the same findings. It is important to point out that a cost model cannot reflect the reality to one hundred percent, but the goal is to build a model that is as close a possible. The regression analysis is limited by the scarcity of data points and that the contractual prices were assumed to be fair.

~ 71 ~

5.2 Limitations: One of the limitations of this thesis is the number of data points collected for the case study. Many other data points could have been considered from other aircraft manufacturers into the regression process, however due to sensitive contractual costs the data points were impossible to collect. Moreover, data obsolescence may be present due to the fact that the ten data points equalling ten programs at BA can take a decade to populate. In essence, a good database of technical and commercial data would lead to a more credible statistical significance and parametric relationship in order to extract the necessary cost drivers. However, it is extremely difficult to obtain a large number of data points in aerospace. The difficulty in accurate cost estimation should not be underestimated. The current study found that in each case for the 10 programs under study excluding the validation points, the NLM outperforms the MLRM and can be seen in the below table. This is consistent with the work of Karadag et al. (2006), who describe that the non-linear regression model has a better performance for analyzing experimental data in the field of ammonium exchange compared to the linear regression. This is also demonstrated in the work of Herman and Scherer (2003), where the non-linear regression outperforms the linear regression in the first order degradation of pest control substances in soil.

5.3 Enterprise applicability: The development of target costing methods, more precisely parametric equations, can be extremely practical in any type of industry. Parametric equations give the user the practicality of understanding a high level estimate based on historical data. For example, in the supply chain when negotiating a new aircraft part or system, parametric equations give you the ability to determine if the initial commercial bid of a supplier is under or

~ 72 ~

overpriced. This will enable the supply chain to follow with an appropriate commercial strategy. In the advanced design team, parametric equations give the ability to perform trade off studies in the early concept phase in order to determine whether the aircraft concept is profitable. Internal studies at Bombardier on the distribution of the aircraft recurring cost describe that 65% of the cost of the aircraft is associated with 6 major systems or structures.

5.4 Future Research: The application of target costing methodologies in the early conception phase is essential because a large portion of a company’s costs are decided interactively with customers and suppliers during product development (Wall, 2011). There can be several future applications of this thesis. This thesis is limited to comparing linear and non linear regression analysis. However, other cost models can be developed, not limiting itself to a specific commodity, rather the model can be generalized. Moreover, this research can be utilized towards any type of industry, such as the automotive industry, the aerospace industry and many other manufacturers. In order to seek the possibility of establishing a superior cost estimate, different types of cost models can be utilized to enhance the credibility of a cost estimate, such as: hierarchical regression models described in the literature review and the artificial neural networks (ANN). According to Suh (2005) and shown in the work of Anderson (1995), ANN is a mathematical model that adapts its structure based on the inflow of information through the network during the learning phase.

~ 73 ~

CHAPTER 6: BIBLIOGRAPHY   

www.google.com www.youtube.com Wikipedia



~ 74 ~