The volatility of decoupling points in a changing strategic climate A case study at Tata Steel

(1)

The volatility of decoupling points in a changing strategic climate

A case study at Tata Steel

Master Thesis

Date: 26-02-2015

Reinout Troelstra

MSc Technology and Operations Management Faculty of Economics & Business

University of Groningen

First supervisor: dr. J.Veldman Co-assessor: dr.ir. W.H.M. Alsem

Company supervisor: M. van Randwijk

Student ID: s1913247 Hendrikstraat 1 9724 NA Groningen rtroelstra@gmail.com

(2)

2

Abstract

As a consequence of global overcapacity and high customer demands, the European steelmaking business is under pressure to keep prices low, to manage their large product variety and to improve customer relationships. An important element in the production and supply chain organization of steel is the customer order decoupling point (CODP). Optimal or robust CODP placement can contribute to efficient management of product variety. Therefore in this study it is investigated what constitutes the optimal placement of a CODP in the steel industry, and how this decision is affected by changes in strategic characteristics. In order to gain insight in the steel business and the CODP decision therein, a case study and quantitative modelling have been performed at a large integrated steel mill (ISM) in The Netherlands.

Currently this ISM operates almost fully in make-to-order (MTO) mode and the resulting delivery performance is poor. Customers demand better delivery performance and find that competing steel firms are better able to serve their needs. Action is thus required and it is not likely that the supply chain with the current CODP location can meet the customers’ needs. A change in production mode is therefore required. There are many characteristics influencing the CODP location however, and there are indications that existing frameworks do not take all relevant characteristics into account.

(3)

3

Preface

With this master thesis I will finish my master programme Technology and Operations Management at the University of Groningen. I am very grateful for everyone who has contributed to the success of this thesis. First of all I would like to thank my two supervisors, J. Veldman and W.H.M. Alsem for the guidance, pragmatism and insights they provided during all stages of writing my thesis. Furthermore I would like to express my gratitude towards M. van Randwijk, who has provided the opportunity for me to perform my graduation research at Tata Steel. This has been a very interesting experience and it helped me gather in-depth information from practice. I would also like to thank my interviewees for their time, insights and willingness to help me out with my research. Furthermore I am grateful towards all those who I have met on the work floor during my internship for their openness and collegiality.

This master thesis is also the final chapter of my career as a student in Groningen. It has not always been easy for me to keep motivated, during my studies. These acknowledgements would therefore not be complete without mentioning the unconditional and seemingly unwavering support that I have received from my parents. My close family and friends have always been there for me, and they should know how important they have been for the success in my studies. Thank you!

Reinout Troelstra

(4)

4

List of abbreviations

CODP: Customer order decoupling point ISM: Integrated steel mill

KPI: Key performance indicator

MLE: Mainland Europe

MTA: Make to availability

MTO: Make to order

MTS: Make to stock

OPP: Order penetration point, synonym of CODP ROTIF: Ready on time in full

SKU: Stock keeping unit TOC: Theory of constraints

WACC: Weighted average cost of capital

(6)

6

1 Introduction

The European steel industry is currently facing large challenges caused by global economic effects. While the costs for raw materials and energy are rising, demand is in decline and competition is fiercer than ever (Anonymous, Steel Industry Profile: Europe, 2013). In addition, the global steelmaking capacity surpasses demand, causing low prices (Robertson, Gibson and Flanagan, 2002; Woertler, et al., 2002). With these challenges comes the recognition of opportunities for improvement throughout the steel-making industry. Suppliers of steel traditionally focus on either one of the well-known value propositions by Treacy and Wiersema (1993): customer intimacy; product leadership or; operational excellence. However, the need has been recognized to set very high standards on all these values (Sinha and Ghoshal, 1999). This requires rethinking the way supply chain operations, production and customer communication are performed. More importantly, these aspects must not be viewed in isolation: synergy is usually the way to success (Sinha and Ghoshal, 1999).

This research will focus on streamlining a large level of product variety as is offered by many integrated steel mills (hereinafter: ISMs) in Europe and North America. Since the number of design choices is “virtually infinite” (Kumar, et al., 2006, p. 5), offering a large product variety while maintaining service performance and keeping operational costs to a minimum, is highly challenging. Management of this high product variety is therefore a key driver of profits for ISMs (Denton, Gupta and Jawahir, 2003). The large variety of end products requires enormous capital investments in raw materials, semi-finished products and semi-finished goods. A small percentage reduction in stock can therefore lead to substantial cost savings. However, customers demand for more reliable deliveries and the easiest way to increase delivery performance is to keep more end-products stock. Therefore in managing all these end products, a balance is sought between inventory costs and delivery performance.

(7)

7 As steelmaking is a so called ‘few-to-many’ process, in which many end products originate from only a few raw materials, decisions on the CODP have large implications for costs and lead times (Denton, Gupta and Jawahir, 2003). Olhager (2003) made an excellent overview of market, production and product characteristics that influence the CODP decision, but there is another category of characteristics that impact the way that the CODP location is responsible for a competitive advantage. Strategic characteristics, such as variations in raw materials and energy costs, capital costs, the exchange rate and the price level of steel can have an impact on the optimal CODP strategy. This study will examine the volatility of this MTO/MTS decision when these strategic characteristics change. With this research, an answer on the following research question is sought:

To what extent and how is the optimal placement of the CODP at an ISM influenced by strategic characteristics?

In order to come to an answer to this research question, a case study will be performed at Tata Steel Europe in IJmuiden, a large ISM in the Netherlands. This ISM currently operates in MTO mode for more than 90 per cent of the total production, and is currently not able to cope with customers’ demand for better delivery reliability. The supply chain and product portfolio of one customer of Tata Steel, which is representative for the type of problems faced, is selected to be analysed in greater detail.

After thorough analysis of the customer, the strategic characteristics that are of relevance for the CODP decision are determined. Subsequently, the research will show the way that competitive advantages are addressed by shifting the CODP and the role of strategic characteristics therein. A quantitative model will show what the optimal placement of the CODP is given certain circumstances, and how volatile this decision is. Finally, it will be investigated how customers can play a role in CODP shifting processes. The latter is important because a CODP shift is not without consequences. It has already been noted that keeping a large variety of products on stock is very expensive. Therefore it is important to research whether the variation in end-products can be reduced in collaboration with the customer.

(8)

8

2 Theoretical background

In this theoretical section a description of the relevant facts and theories on the steel industry and the CODP is provided. The literature review is divided into several sections: first the CODP and its placement considerations are investigated, subsequently a brief explanation of the steelmaking process at an ISM is provided and finally the strategic developments in the steel industry are outlined.

2.1 The CODP

More and more companies today realize that the placement of customer order decoupling points (CODPs) is an important strategic decision in supply chain and manufacturing operations (Olhager, 2010). The CODP (sometimes also referred to as the order penetration point, OPP), is defined as the point in the supply chain where products are no longer produced to forecast but rather produced to order. As such, it represents the stage where every individual product is specified to meet the demands of a single customer (Olhager, 2003). Therefore it specifies the mode of production that is applicable, either MTS/MTO or some hybrid system. Next to MTS and MTO modes of production, assemble-to-order (ATO) and engineer-to-assemble-to-order (ETO) systems are used. These systems are outside of the scope of this research, as they are linked to the design- and assembly phases of production (Olhager, 2003) and therefore relevant to an ISM. In a hybrid MTS/MTO system, inventory points in the production system are served in MTS mode and the final processing step(s) are performed in MTO mode when an order arrives. Figure 1 illustrates the location of the CODP in the production modes under consideration.

Figure 1: Different production modes, dashed lines represent production based on forecast, whereas solid lines represent order-based production (Adapted from: Rafiei and Rabani, 2011)

(9)

9 2.1.1 CODP placement

This section provides a concise literature review of factors influencing the CODP placement. Olhager (2003) has provided an excellent overview of CODP positioning considerations based on the work by Sharman (1984), Olhager and Östlund (1990), Giesberts and Van Der Tang (1992), Hoekstra and Romme (1992), Olhager (1994), Van der Vlist et al. (1997), Fisher (1997), Olhager and Wikner (1998, 2000), Pagh and Cooper (1998), Lehtonen (1999), D’Alessandro and Baveja (2000), Mason-Jones, Naylor and Towill (2000) and Van Donk (2001). He chose to divide the most important influencing variables into three categories: market, product, and production characteristics. To narrow the focus, only the variables that are possibly relevant in this study are outlined below.

Market-related factors Explanation

Delivery lead-time requirements Demands from the market, limiting the backward placement of the CODP.

Product demand volatility Indicates the opportunities for order or make-to-stock production

Product volume Related to demand volatility, relative volatility is lower for high-volume items

Product range and product customization requirements

Broad range of customization requirements renders MTS impossible

Customer order size and frequency Indicates volume and repetitiveness of demand

Table 1: Market related factors influencing the CODP placement

Product-related factors Explanation

Customization opportunities Customization opportunities offered in anticipation of customer demands influences delivery strategy Material profile Convergent/divergent product structure, e.g. few raw

materials used to make many end-products Product structure Is an indicator of product complexity, which may

subsequently lead to long lead times

Table 2: Product related factors influencing the CODP placement

Production-related factors Explanation

Production lead time Important factor related to delivery lead-time requirements

Number of planning points Potential CODP positions, e.g. stock points or production steps

Flexibility of the production process Flexibility originating from for instance short set-up times

Bottleneck positioning Positioning of the bottleneck relative to the CODP, decoupled bottleneck does not face volatile demand, however downstream bottleneck only works on actual orders

(10)

10 Olhager (2003) recognizes the interrelatedness of the factors influencing CODP positioning, and the hierarchical influence of market characteristics on product and production characteristics. Figure 2 is a graphical representation of the factors impacting the CODP. It shows that the aforementioned factors do not only impact the CODP directly, but also via the delivery and production lead time, which are both highly important in the CODP placement.

Figure 2: Conceptual impact model of factors influencing CODP placement (adapted from Olhager, 2003)

The overview of market, product and production characteristics that Olhager (2003) identified to influence the CODP placement seems to be rather complete. However it does not take into account the characteristics from the external environment that directly impact the cost price and profitability of products at a firm, hereinafter called the strategic characteristics.

Examples of possibly relevant strategic characteristics in the steel industry are raw materials and energy costs, capital costs, the exchange rate and the market price of steel. These characteristics can have a considerable impact on cost price and profitability of products, which could mean that the optimal location of the CODP is affected as well. This is because changes in the CODP location should be strategically motivated and therefore address competitive advantages; e.g. costs, delivery reliability (Olhager, 2003). This will be elaborated upon in sections 2.3.2 and 2.4.

The direct impact of strategic characteristics in general has been recognized by Melnyk, Stewart and Swink (2004), Perona, Saccani and Zanoni (2009), Rafiei and Rabbani (2011), MacCarthy (2013) and more specific for the steel industry by Potter et al. (2004), Kerkkänen (2007), Kumar et al. (2006), Singer and Donoso (2006). Research linking strategic characteristics to the CODP placement however, seems to be lacking. The volatile nature of the steel industry and its strategic characteristics is reason to believe this topic requires considerable attention.

(11)

11

2.2 The steel industry and the ISM

In order to better understand the strategic characteristics and their relation to the CODP problem, literature on the global steel developments has been reviewed. To keep the main text of this thesis concise, the majority of the review is provided in an appendix, whilst the most relevant insights are outlined here. The second part of this section elaborates upon the production process at an ISM, to enable the reader to understand how an ISM works. It is also an attempt to provide some insight in the unique nature of steel production.

2.2.1 The steel industry characteristics

The global overcapacity is the biggest challenge for the steel industry in Europe. In the past decade, a lot of extra capacity has been added, mainly through the construction of additional steel plants in China and India. Governments do not allow ISM’s to discontinue their business because of the impact on the unemployment rates. As an example, one of the largest ISM’s in Europe, Riva in Italy, is supported financially by the government to make sure that the 15.000 employees do not lose their jobs. As a result of these government interventions, market-driven capacity reduction is not occurring. At the same time, the steel consumption in Europe is in a steady decline. A key reason for this is the reduction of steel intensity per product: less steel is required because steel is produced in higher strengths. As a consequence, steel manufacturers are steering on higher efficiencies, competition becomes fiercer, imports increase and the steel prices drop. Appendix A describes the macro-economic developments with regard to the supply and demand of steel, indicating the gravity of the overcapacity.

(12)

12 Key disturbing forces for European ISMs are considered in Appendix B. The most important takeaways from this literature review are that the competitive pressure on an ISM is indeed very strong, and that the context in which steel firms operate is changing, building the need for more efficient operations. Moreover, the literature review provides initial evidence that the strategic characteristics mentioned at the end of section 2.1.1 are indeed relevant concerning the CODP placement problem, and that these characteristics are currently highly volatile in nature.

2.2.2 The ISM production processes

Apart from the magnitude of operations and the huge variety of end-products that is delivered, steelmaking is a relatively simple process and the basic process steps have not changed significantly throughout history. In Figure 3 the core operations of an ISMs are depicted. The first step is the iron making, which is conducted in a blast furnace. In the blast furnace, iron ore and coal (which has been earlier transformed into coke) are heated so that molten iron or hot metal is created (Sinha, et al., 1995). The molten iron, mixed with scrap steel and secondary materials, is reduced in an oxygen furnace to form steel that is put in a ladle; a kind of container (Denton, Gupta and Jawahir, 2003). The molten steel is further refined and cast into slabs of steel by a so called continuous casting process. From an inventory point, the slabs are pulled to be reheated and hot-rolled into coils. These are stored in inventory as well, to be further processed by for instance cold rolling and/or finishing operations, such as annealing and galvanizing.

Figure 3: The elementary steelmaking steps performed by an ISM. Source: Tang and Wang (2008)

2.3 Applying the CODP placement framework

(13)

13 sections provide insight in the steel industry-specific CODP placement complexities, as well as in the attempts that have been made in the past to apply a CODP placement framework.

2.3.1 CODP placement at an ISM

According to a study by Denton and Gupta (2004), balanced CODP placement by ISMs is essential for maintaining high efficiency and competitiveness. ISM operations generally encompass high capital expenditures and long cycle times, stressing the importance of time and inventory trade-offs. They also touch upon the subject of mini-mills; smaller steel mills processing recycled steel, which have a cost benefit as well as shorter cycle times in plain carbon-steel markets. These mini-mills pose a threat to the much larger and capital-intensive ISMs, thereby increasing the urgency for the large steel manufacturers to decrease lead times, while maintaining product variety and low costs. The conclusion can be drawn that the steel industry is a prime candidate for CODP optimization.

A typical integrated steel plant produces thousands of unique end-products which leads to great complexity of CODP placement. Denton, Gupta and Jawahir (2003) found however that a large share of the portfolio of steel manufacturers is comprised of recently introduced products, and that especially products with more finishing operations or special grades (‘high grades’) have high contribution margins. The management of all this product variety is thus both worthwhile and demanding. To illustrate the amount of end products that is delivered by a steel producer, Figure 4 depicts exemplary order specifications applied to products at various stages.

Figure 4: Potential specifications of items downstream of the basic steelmaking process (Source: Morikawa, Takahashi and Hirotani, 2014)

(14)

14 and downstream CODP placement does not support a high degree of customization (Kumar, et al., 2006; MacCarthy, 2013). The optimal CODP placement can therefore not be generalized for all products, one product family or even a product type. Moreover, the complexity of the CODP decision is greater when the amount of product variation grows and strategic characteristics become increasingly dominant, as is the case in the steel industry (Kerkkänen, 2007).

2.3.2 CODP placement in practice The previous sections show that:

1. The CODP location has a large effect on important KPI’s of any supply chain

2. Optimal CODP placement at an ISM is determined by a wide range of market, product, and production characteristics, but probably also by strategic characteristics.

3. The steel industry and its strategic characteristics are volatile, so optimal CODP placement might also be.

4. CODP placement is for an ISM both highly relevant and challenging

Thus, sufficient evidence has been provided to justify a critical analysis of the CODP placement decision at an ISM. The question however, is how to proceed with this analysis. Perona, Saccani and Zanoni (2009) stress that there has been very limited research done on the applicability of CODP placement considerations in practice. They find that theoretical approaches focus on multi-item and multi-stage cost optimization methods that are often too complex to be used in practice, whilst practitioners commonly resort to their common sense and knowledge without a theory-based rationale. Subsequently they have tried to bridge the gap between the theory and practice themselves by proposing a four-step framework for the CODP placement, which is then used in a case study. Through selection and elimination a subset of products is selected for which the CODP is determined, based on a simple analytic model.

The four-step framework, proposed by Perona, Saccani and Zanoni (2009): 1. Segmentation of products into homogenous product groups 2. Decoupling point determination

3. Replenishment policy assignment 4. Parameter setting

(15)

15

 The selection of MTO/MTS or a combining strategy is done based on a typology of product groups. Subsequently each group is assigned a suitable strategy according to the work of Huiskonen, Niemi and Pirttilä (2003) and Soman, Van Donk and Gaalman (2007). This product group and CODP assignment is however based solely on the analysis of physical volumes and the number of customer orders. Additional strategic characteristics are not taken into account.

 The results indicate a 200% increase in holding costs, while in other companies the new method might not lead to inventory changes, because holding costs are more expensive.

 Planning activities such as capacity management and purchasing, which can be very relevant, are not included in the model.

 The model is somewhat limited in its quantitative and rational decision-making support.

 It does not provide guidelines for reviewing the CODP placement when strategic characteristics change.

 There is no sensitivity analysis conducted. It is unknown how the outcomes (costs, lead times) change when the CODP placement is changed.

Perona, Saccani and Zanoni (2009) have thus made an interesting attempt to apply CODP placement in practice, but their approach lacks support of strategic characteristics and they do not solve the complexity of assessing all relevant cost drivers when shifting the CODP. This cost estimation problem will be addressed in section 2.4.1 of this thesis.

2.3.2 Including the strategic characteristics

This thesis will attempt to build upon the CODP placement framework by Olhager (2003), so that the relevant strategic characteristics in the steel industry are taken into account. Market, product and production characteristics will not be taken into account here. The strategic characteristics that are possibly of relevance here, are depicted in Table 4.

Strategic

characteristic Definition

Raw materials costs Cost of all the most important raw materials in steel production, iron ore and coking coal, required for making a tonne of steel

Energy costs Cost of energy for making a tonne of steel

Capital costs Interest rate charged for internal and external funds

Exchange rate Exchange rate between the euro and the dollar on the currency exchange Market price of steel Market price for a tonne of steel

Table 4: Definition of strategic characteristics

(16)

16 (Olhager, 2003; Van Donk, 2001). Raw materials costs are an important factor in the cost price of steel (Singer and Donoso, 2006) and therefore influence inventory costs. Similarly, energy costs, capital costs and the exchange rate impact the cost of inventory and thus possibly the CODP position. The market price of steel is highly volatile; the price of a tonne of hot rolled steel for example varied between $250 and $1200 between 2000 and 2009 in the US market (Ozario, Bastian-Pinto, Baidya and Brandão, 2013). These large fluctuations could mean that the profit margin varies greatly as well, which can be a strategic driver to change the location of the CODP.

Detailed analysis, using various sources of information in the case company, should reveal whether Table 4 should be complemented with additional strategic characteristics or that perhaps certain characteristics should be eliminated.

2.4 Competitive advantages of CODP shifting

The goal of the entire process of evaluating and possibly relocating the CODP is of course to improve the firm so that it is able to remain competitive and possibly even outperform the competition on certain KPI’s. Relocating the CODP either more up- or downstream is called backward and forward shifting respectively. According to Olhager, any shift in the CODP should “strengthen a competitive priority” and it is thus a strategic decision (Olhager, 2003; p.325). Figure 5 provides an overview of the competitive advantages that are addressed by forward- or backward shifting the CODP, as identified by Olhager (2003). Backward shifting means production will be (more) MTO, forward shifting means (more) MTS. There is obviously a trade-off in the competitive advantages that are addressed when a shift is initiated. As an example, a forward shift could lead to enhanced delivery reliability and/or speed, but might negatively influence the product range or product mix flexibility.

Costs

MTO Product range Product mix MTS

flexibility Quality Delivery reliability Delivery speed CODP Forward shifting Backward shifting * More opportunities for process optimization Less WIP inventory

Scope

Cost component 1: Cost component 2:

(17)

17 Figure 5 illustrates how the CODP should be shifted in order to strengthen one or more competitive advantages. In different contexts, these competitive advantages will have different weights or importance. The focus here is on the impact of shifting the CODP on the competitive advantages of costs, delivery speed and delivery reliability. Delivery speed- and reliability are fairly straightforward concepts, and from the previous theoretical sections the forward shifting effect on the CODP is clear. Delivery speed and delivery reliability are hereinafter together also referred to as delivery performance. Pertaining to costs, there can be both forward- and backward shifting components. A detailed explanation why these costs are assumed to be a backward shifting competitive advantage is provided in the following paragraph.

2.4.1 Cost reduction or cost increase?

There is one key difference between Figure 5 and the work of Olhager (2003); Olhager (2003) argues that forward shifting leads to cost reduction because of increased process optimization opportunities. In a similar fashion he associated backward shifting with lower costs because of less work in process inventory. The work on CODP placement by several authors is inconclusive on the effect of CODP shifting on total system cost; defined here as the monetary costs involved in the production and delivery. Moreover it shows that the total system cost is highly dependent on the characteristics of the firm and industry under investigation.

As an example, the CODP placement framework by Perona, Saccani and Zanoni (2009) is based on the assumption that holding costs are fairly low and are thus penalized less in the holding costs vs. setup costs trade-off. This assumption is certainly not valid across different contexts; think of business with high value-added production processes. Van Donk (2001) finds a similar trade off as Olhager (2003) in the sense that the upstream effect caused by cost of steps in the primary process are balanced against the downstream effect of stockholding costs, but in his work that trade-off is severely influenced by the perishable nature of food, making stockholding more expensive for perishable products. In addition, Van Donk (2001) recognizes that two important determinants on the value of stock are the value of the raw materials purchased and the value added in production. If procurement of raw materials in an MTO situation can be postponed until an order arrives, the aggregate inventory will be lower. If the production is highly value-added, the CODP is expected to be more upstream, as higher value inventories are more expensive to store (Van Donk, 2001).

(18)

18 the CODP upstream. There is some considerable value added in the process though, especially when dealing with better steel grades. The aggregate inventory will therefore probably cost more in MTS production than in MTO production. Even more so in comparison with many industries in which MTO production requires considerable inventory of mid-level parts, such as in assembly lines.

(19)

19

3 Conceptual model

The theoretical background study of chapter 2 of this thesis, and the gaps identified therein, have led to the conceptual model of Figure 6. The model must be interpreted in line with the scope that has been defined in Table 4 and Figure 5. Hence, only the moderating effect of the strategic characteristics on the relationship between the CODP placement and the competitive advantages is under consideration here. The competitive advantages that are addressed here are: advantages of delivery speed, delivery reliability and costs.

Strategic characteristics

CODP location Competitive advantages

Figure 6: Conceptual model of the moderating role of strategic characteristics on the relationship between the location of the CODP and the competitive advantages that are addressed

In this model, the CODP location is the independent variable, the competitive advantages of costs, delivery speed and delivery reliability are the dependent variables. The strategic characteristics of Table 4 have a moderating role, in the sense that they affect the way that the CODP location influences either one or all of the three competitive advantages. This is a key difference in comparison to the influence of market, product and production characteristics that Olhager (2003) identified. The latter characteristics influence the CODP location, and ultimately the way that the competitive advantages are addressed. The strategic characteristics, in contrast, only affect the relationship between the CODP location and the competitive advantages. This means that when a certain CODP location is chosen, the competitive advantages are addressed to a lesser or greater extent dependent on the strategic characteristics. Paradoxically, this might result in the optimal CODP location to shift when the strategic characteristics change.

(20)

20

3.1 Research questions

In order to validate the conceptual model, research questions have been formulated. The main research question is:

To what extent and how is the optimal placement of the CODP at an ISM influenced by

strategic characteristics?

This main research question has been split up and divided in more narrow-focused sub-questions (SQ), which allow for individual aspects of the problem to be under investigation. The first sub-question (SQ 1) is answered based on empirical research. The latter sub-questions (SQ 2-4) are answered with a model-based approach. The methods chapter will elaborate upon the details of the research design.

SQ 1: How do strategic characteristics influence the relation between the CODP location and the competitive advantages that are addressed?

With this research question an answer is sought on how the strategic characteristics change the way that costs, delivery speed and reliability are addressed by the CODP positioning. The basic mechanism of Figure 6 is under investigation here. To get a grip on this relation, the strategic characteristics of the steel industry and their impact need to be understood.

SQ 2: What are the performance effects of relocating the CODP?

Changing the location of the CODP can have a significant effect on costs and delivery performance. It is however unclear to what extent these variables will change when the CODP is shifted, especially for the steel industry. An indication of the impact of CODP relocation will provide decision-makers with the information they need.

SQ 3: How volatile is the optimal placement of a CODP?

(21)

21

SQ 4: To what extent and how can the CODP be relocated by coordination efforts with businesses downstream of the ISM?

(22)

22

4 Methods

This section outlines the methodology that has been used to answer the research questions as introduced in section 3.1. First, a description of the case company and the case selection are provided. In the subsequent section the research design, including the data collection and data analysis, is treated.

4.1 Case company

This thesis research has been performed at an ISM owned by Tata Steel Europe, located in IJmuiden. This facility is one of three ISMs run by Tata Steel Europe. Two others are located in the United Kingdom. Tata Steel in IJmuiden is responsible for 9000 employments, and with 750-acres it is the largest industrial site in The Netherlands. The mill produces slab, hot and cold rolled coil and coated strip products (galvanised coil and packaging steels). Its predominantly high-grade steels are used for e.g. automobiles, construction and packaging.

Tata Steel Europe is the result of two mergers; first in 1999, Dutch-owned Hoogovens and British Steel together formed Corus. Later, in 2006, Corus became part of the Indian concern Tata Steel Group to form the fifth largest firm globally. In 2013, caused by the rise of major Japanese and Chinese steel firms, Tata Steel Group has dropped in ranking to 11th_{place (World Steel, 2014). Their subsidiary Tata} Steel Europe is the second largest steel company in Europe.

4.2 Case selection

(23)

23

4.3 Research design

The research questions guide the analysis towards identification and exploration of the link between the location of the CODP and strategic characteristics. The research should also create insight in the performance effects and volatility of this CODP decision. These phases of the research have been conducted with a combination of qualitative and quantitative methods. The following sections provide a validation of the methods that have been used. The first section justifies the methods that have been used for each sub-question. The data collection and data analysis sections then show how the data required for the various research methods has been acquired and analysed.

4.3.1 Sub-question 1

SQ 1: How do strategic characteristics influence the relation between the CODP location and the competitive advantages that are addressed?

This sub-question is highly exploratory and therefore lends itself for case study research. Meredith (1998) provides an excellent plea for the use of case studies in operations management research. The author argues that in comparison with rationalist approaches, such as optimization/simulation models and surveys, case study research is more concerned with understanding phenomena, whereas rationalist approaches are used to explain phenomena. Advantages of case study research, introduced by Benbasat, Goldstein and Mead (1987) are:

1. Information can be gathered in a natural setting, so relevant theories can be developed from observing actual practice. Case research is thus very suitable for capturing the knowledge of practitioners.

2. “How” and “why” questions can be answered with a case study, to grasp the nature and complexity of the phenomena that are studied.

3. Case research allows for early, exploratory research.

(24)

24 4.3.2 Sub-questions 2-4

SQ 2: What are the performance effects of relocating the CODP? SQ 3: How volatile is the optimal placement of a CODP?

SQ 4: To what extent and how can the CODP be relocated by coordination efforts with businesses downstream of the ISM?

For researching these sub-questions a more rationalist approach is adopted. By means of quantitative modelling, the findings and data of the case study have been used to create a better understanding of CODP placement for this case. It was a way to explore the effects of the mechanisms observed in the case on a variety of values (Van Kampen and Van Donk, 2014), and to relate it to more general problems.

The focus has been be on quantitative analysis; modelling performed in Excel provides the desired numerical feedback on the CODP decision during various strategic characteristics. The result is a simple linear model in which strategic parameters can be changed to see the effect on the CODP decision. However, a combination of methods (both qualitative and quantitative) has been applied here. The input data for a quantitative model can be complex and sometimes interpreted in various ways. Only by combining it with findings from other methods (interviews, archival research), a strong argumentation was possible.

4.3.3 Data collection

Interview protocol

(25)

25

Informal conversation

As this research project was conducted in the form of an internship at the supply chain department, a large knowledge base was accessible as well in the form of informal conversation. In addition to the interviewees, many more employees have contributed to the information in this thesis. Various factories at the ISM have been visited, and meetings and workshops on the improvement of delivery performance were attended.

Archival research

A lot of information on delivery performance, improvement programs, costs and inventory could be found on the case company’s intranet. In addition, employees from Tata Steel who are working in close collaboration with the selected case customer have supplied relevant information on orders, order volumes, delivery performance and sales/marketing information.

4.3.4 Data analysis

For a good understanding of the results section, the definition of ROTIF and a description of how it is calculated are provided here. Furthermore the quantitative model that has been used is introduced.

ROTIF

Readiness on time in full (ROTIF) is defined as the fraction of orders without quality defects that are ready in full for delivery towards the customer at the time that was promised. If for instance 200 tonnes were ordered, and only 180 tonnes are ready at the promised date, the ROTIF is 0%. This is the measurement for delivery performance that is used in the case company. It is deemed adequate because it excludes possible delivery issues, so the internal supply chain performance can be viewed in isolation. Furthermore the ROTIF is easy to calculate because the buyer and supplier always agree upon a due date and quantity. Because every order is tracked, computer systems can show exactly how much of the ordered quantity is ready for delivery at the promised due date.

(26)

26

Quantitative model

(27)

27

5 Results

The subsequent sections contain the results of the interviews, archival research and quantitative modelling. First, the selected case customer is introduced. This vital step allows the reader to understand the context-specific dilemmas and complexities. Moreover it is done so the subsequent information can be related to this specific case. In the remainder of this chapter, the sub-questions are answered consecutively.

5.1 Customer analysis

As described in the methods section, one customer has been selected to be under investigation. This customer is Company X; a manufacturer of xxxxxxxxxx, mostly for specialized applications. They are active globally, with production facilities in xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx..

Company X has been a customer of Tata for about six years. It is served directly by the production hub in IJmuiden as well as by steel service centres in Maastricht. Company X is very dissatisfied with the delivery performance that is realized by Tata Steel. Because they order steel types with a high profit margin, Company X is seen as an important customer of Tata Steel. This customer is assigned the highest priority by the sales and marketing department. Contributing to this is the belief that if the delivery performance is improved, this buyer will grow as a customer. This is a key assumption that will be retained throughout the rest this thesis.

In order to understand this customer and the type of challenges that Tata faces in serving this customer, a detailed analysis is provided in the following sections.

5.1.1 Demand side

Although Company X is active in global markets, Tata Steel Europe is only concerned with Western Europe. Company X is active in several xxxxxx markets in this part of the world, the most important being: xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx. The market share of Company X differs across these markets. The following figures show the market share that is realized by Company X, in comparison with other businesses.

--Confidential--

(28)

28 5.1.2 Supply side

The total demand for steel by Company X is about 000.000 tonnes (t). Its total demand is expected to be stable over the coming years. Last year, Tata Steel supplied about 00.00 t (0%), whereas in the current year this is expected to grow to about 00 kt (00%). The most important competitors in this market are ArcelorMittal and Thyssenkrupp, which respectively supplied about 000 and 000 kt in the previous years. The rest of Company X’s demand (approx. 00 kt) is fulfilled by smaller suppliers (<00kt). There is a desire and presumed opportunity to grow to a volume of about 50kt in 2018, and even 100kt in the more remote future. Company X is buying its products from more than one supplier to reduce risks of lead time uncertainty and quality issues, as is common for most firms. Satisfaction with the price and quality of products, as well as excellent technical support are reasons for Company X to buy products from Tata.

--Confidential--

Figure 8: Steel supply by Tata and its competitors

5.1.3 Supply chain

The diagram below depicts the most relevant supply routes to the different production locations of Company X. Multisteel in Maastricht is a Tata-owned distribution and slitting centre. Vogten is a third-party service centre, also located in Maastricht, where steel can be pickled, decoiled, cut and slit. In order to limit the scope of this research, in the subsequent sections no distinction is made between the two service centres.

--Confidential--

Figure 9: Supply routes from Tata towards Company X in xxxxx and xxxxxxx

Multisteel is holding about four weeks of stock designated for Company X to buffer for uncertainty. Additionally, Company X is keeping about two weeks of stock because the deliveries are so unreliable. One of the reasons for this is that Company X orders specific products that cannot be bought elsewhere if the product is too late.

5.1.4 Current delivery reliability

(29)

29 meeting its own goals either; the ROTIF in the past two years has been between 65% and 85%, whereas they strive to have an average ROTIF of 90%.

The most important demand by Company X is the improvement of the delivery reliability. There have been periods recently in which a ROTIF as low as 25% has been experienced. Many products that Company X orders are so specialized that they cannot easily be procured elsewhere on a short time basis when the delivery is delayed. A steel warehouse does not have these products on stock. Moreover, the procurement of (more commoditized) products when deliveries fail introduces new supply chain problems. The products that are delivered too late from Tata are redundant and have to be kept at stock in Maastricht, incurring additional costs. However, Multisteel is able to forecast with reasonable reliability the demand for the next three months. Perhaps better forecastability of the products is playing a role in the performance difference between direct and indirect deliveries. If Tata wants to increase either the profit margin or the total order volume for a customer like Company X, significant improvements are necessary. The improvements should lead to a ROTIF in excess of 90% for Company X to be satisfied. Delivery performance is in that sense a precondition for growth. The direct competitors of Tata are believed to have a delivery performance that is much closer to Company X’s demands.

5.1.5 Lead time demands

A lead time of 6-8 weeks or even 10-12 weeks would be acceptable, but only when the delivery performance is significantly increased in excess of 90% ROTIF. In comparison, competitive steel firms deliver between six to eight weeks. However, Company X would of course always like to have the lead time as short as possible if there is no extra charge involved.

5.1.6 Volatility of customer quality demands and order portfolio

(30)

30 5.1.7 Internal supply chain

Figure 10 shows the routing and yearly volume (in tonnes) of the different steel grades that Company X has ordered in 2014. No distinction has been made in the two xxxxxxx locations of Company X, as this would be too extensive for this research. From Figure 10 it can be understood that all direct product flows are towards Company X’s location in xxxxxxx. There are two processes that can be used to make a steel coil: using slabs that are cast beforehand (hot rolling) or by directly rolling the steel after it is cast in the direct sheet plant. xxxxx-grade steels require an additional processing step (cold rolling) before they are ready for the customer. The diagram shows about half of the total volume is delivered to Company X via service centres in Maastricht, and the other half originates directly from the ISM in IJmuiden

5.1.8 Delivery performance and inventory points

(31)

31 --Confidential--

(32)

32 --Confidential--

(33)

5.2 How do strategic characteristics influence the relation between the CODP location and

the competitive advantages that are addressed?

The findings that relate to this sub-question will be presented in twofold. First, the strategic characteristics under consideration (Table 4) have been checked for their relevance in the steel firm. The second part describes how these characteristics relate to the CODP decision for Company X.

5.2.1 Relevance of strategic characteristics

In order to see what strategic characteristics impact the CODP and how, correct understanding of the current strategic characteristics in the steel industry is vital. The interviews have provided much insight, confirming most findings set out in the theoretical framework of this discourse.

Raw material costs

Raw materials costs make up about three quarters of the cost price of steel (between $20 and $200 per tonne). There is a strong correlation between raw materials costs and the cost price of steel. There is a constant struggle to translate higher raw materials costs into higher sales prices, and to prevent higher raw materials cost to incur a price decrease. This is quite hard however because of the transparency of the markets. The raw materials market is quite volatile: the iron ore market is dominated by three large mining firms. The largest steel manufacturer (ArcelorMittal) only serves a fraction of the world market, whereas the mining companies operate under oligopolistic conditions. The steel manufacturers therefore do not have any buyer power. In the coking coal market, similar mechanisms are occurring.

Energy costs

Energy costs are not a large contributing factor in the cost price of steel. Also, steel production incurs gas and thermal energy production, which can be sold or used to generate electricity.

Exchange rate

(34)

34

Market price

Steel prices are different for different customers, different steel grades and different markets. Prices are the result of a negotiation process between customer and supplier. In general, prices used to fluctuate 30-40% over time, currently the price range is a bit narrower and the variance is dictated by raw materials prices. Currently the raw materials costs are low so steel firms attempt to maintain the sales price.

Capital costs

It is hard to say what exactly the cost of capital is at Tata Steel. Sources at Tata say that whenever required, a weighted average cost of capital (WACC) of 10% is assumed. The acquisition of Corus by Tata Steel has led to reasonably high interest rates set by the parent company.

Concluding remarks

The preceding section confirms the importance of most strategic characteristics outlined in Table 4. The energy costs however, shall not be taken into account because of its small role in the cost price of steel.

5.2.2 Connecting strategic characteristics to the CODP for Company X

Raw materials cost, capital cost, market price and the exchange rate are the strategic characteristics that have been identified to be of importance here. But how do they relate to the competitive advantages that should be addressed by shifting the CODP? The aforementioned factors clearly all relate to cost/price or value. The link with delivery speed and/or delivery reliability might be a little less clear. The following paragraph will provide some insight.

Size of the prize

(35)

35 Hence, the goal of forward shifting is to increase the throughput at the expense of increased inventory costs. More specifically, by enhancing the delivery performance customer should order more or be willing to pay more for increased service. Forward shifting would be pointless without the chance of either of these two last named scenarios occurring. The tremendous competitive pressure in the steel market indicates that customers are not likely to pay a premium, which is confirmed by experts. Even more so because reasonable delivery performance is experienced as a market qualifier. This is especially true for Company X, which has faced bad delivery performance ever since it became a customer. Roth and Miller (1990) found that the financial performance of a firm is enhanced with the improvement of the delivery performance, but only when the plant progresses from performing below standard to an average performer (cited in Corbett and Claridge, 2002, p.113). Taking into account that the current delivery performance towards Company X is far below standard (elaborated in section 5.1.4), it is highly likely that there are financial benefits to gain for Tata.

The ‘size of the prize’ to be acquired is thus the additional order volume that is realized by better delivery performance. An estimate of the value of the extra order volume is the total extra volume in tonnes, multiplied by the difference between sales price and cost price. Hence, all strategic factors have an impact on this extra value; either because the cost price varies or because the market price is changing.

Cost of the prize

(36)

36 inventory at all times. One per cent difference in capital costs has a large impact on the inventory costs, for instance for Company X.

Conclusion

There is no doubt that strategic characteristics have an impact on the decision to shift the CODP either up- or downstream for a customer like Company X. If delivery performance is key, MTS seems very suitable, especially when stockholding costs are relatively low. It is however not unthinkable that market prices are low while stockholding costs are at a peak; Tata could then make the strategic choice not to shift the CODP downstream and try to optimize the delivery performance in another way. The specific characteristics of Company X do however opt for a strategy leaning towards MTS. This is less relevant regarding customers that order lower-grade steels and/or experience better delivery performance.

5.3 What are the performance effects of relocating the CODP?

As currently steel production is almost fully MTO-based, relocating the CODP can only mean one thing; a forward shift. As has been laid out in the previous section, a forward shift is expected to lead to delivery performance increase and subsequently order volume growth. On the other hand, inventory costs will rise because of higher value end-stock. In this section, an attempt is made to quantify the value of these components. In order to do so, reference is made to the case company, Company X. A detailed explanation of the calculations in this section is provided in Appendix C.

5.3.1 Extra order volume

(37)

37 characteristics in the steel industry. Note that the additional order volume is expressed in extra value here and not in additional profit, because at this stage, increased inventory costs are not yet taken into account. Section 5.3.4 will indicate how this extra value can also lead to additional profit.

5.3.2 Delivery performance

The most important precondition for the order volume growth is an improvement in delivery performance. More specifically, Company X desires a better delivery reliability. Delivery speed is less of an issue, and Company X is even willing to accept longer lead times if that improves reliability. A forward shift towards MTS should unequivocally lead to better delivery reliability, or growth in sales will not occur. It is therefore important to assess how much the delivery performance is expected to increase when a forward shift is initiated.

The current situation, depicted in 5.1.8, is insightful for what lies ahead. Multisteel and Vogten are currently keeping stock designated for Company X, based on their own forecast. Even though Company X is placing orders in regular MTO mode at Tata, the service centre operates in a way that closely resembles MTS because of four weeks stock kept at their location. The effect is clearly visible: a ROTIF of about 90-95% is achieved, whereas direct delivery performance from Tata IJmuiden towards Company X is much worse with a ROTIF between 45-60%. However, only about 50% of the volume is currently delivered via the Multisteel/Vogten route, as shown in Figure 10. If all products for Company X are kept on stock, 90-95% ROTIF should be viable, depending on stock levels. Assuming full MTS production, a 30-50% increase in ROTIF is therefore possible.

5.3.3 Inventory costs

(38)

38 flattening of the curve: there are certain products that are ordered so irregularly, that incredibly high stock levels would be required in order to satisfy demand for these products at all times. Figure 12 shows that for an increase in ROTIF of about 30%, 30 days of inventory must be kept.

Figure 12: Delivery reliability increase expected when inventory of all products is kept

The inventory costs are quantified using of Figure 20 in Appendix C. Thirty days of end-stock inventory (1600 tonnes) costs about €80.000 annually, assuming a cost price of €420 per tonne, and 13% stock-keeping costs. Taking into account that MTS production requires less stock of raw materials, inventory costs are only €20.000-€40.000 higher in comparison with MTO production

5.3.4 Combining the measures

Inventory increase relates to delivery performance increase and the latter will most likely lead to extra order volume, as established in section 5.2.2. Once more a relationship following an s-curve is assumed between delivery reliability increase and extra order volume. Taking into account that the current reliability is very unsatisfactory, the customer will not order much more volume if the delivery performance is only marginally increased. If significant steps are taken however, the

(39)

39

Figure 13: Order volume growth expected when delivery performance is improved

Combining the two S-curves, inventory increase can be related to gaining order volume (via better delivery performance), depicted in Figure 14. The parameters of the s-curves have been set such that a credible distribution is created. Figure 14 shows that small amounts of inventory will not lead to much higher delivery reliability, and not a lot of extra order volume is gained. However when a reasonable amount of inventory is kept, significant steps in order growth are possible. Currently, 14 days of inventory is kept, but only for the products that are supplied by Multisteel/Vogten. Tata therefore currently operates on the beginning of the curve. Taking the aspiration into account to have a ROTIF of more than 90%, between 20 and 30 days of inventory must be kept and between 20 and 30 kt of order volume growth is possible.

Figure 14: Relationship between inventory and order growth

(40)

40

Figure 15: Cost of inventory versus value of increased order volume Conclusion

There is a trade-off between the costs of inventory that is kept and the extra order volume that will be gained. Stock-holding can be worthwhile, albeit under specific circumstances. A forward shift of the CODP always leads to higher inventory costs, but this will more than made up for if the order volume growth is significant. Changing the strategic characteristics has an effect on the value of extra orders and the inventory costs, and the break-even point will shift accordingly. The next section will show in what way and how much the values change when the strategic characteristics change.

5.4 How volatile is the optimal placement of a CODP?

(41)

41

Table 6: Sensitivity analysis of the strategic characteristics. The left column shows the base case and its outcomes.

The sensitivity analysis has two empty entries; this is because here the cost price surpasses the market value, causing a negative profit margin. In reality, the market price is adjusted here to compensate for the increased cost price. An important finding from the sensitivity analysis is that when the difference between the market price and cost price is at least €30, - and increasing, stock-holding is profitable. The larger this gap becomes, less inventory is required to reach the break-even point. Raw materials such as coking coal and iron ore have a considerable influence as well. Even more so, when you take into account that these raw materials prices can vary up to 200%. The exchange rate has an impact as well, but is less likely to vary much. The capital costs finally, have less impact than might be expected beforehand. The explanation for this is that the extra value of inventory kept compared to the MTO situation is not very high.

Conclusion

Because the strategic characteristics have been identified as highly volatile and have effects on both the cost price and the market price, it is difficult to point out how exactly the CODP should be shifted when one of these characteristics will change. Once more, the analysis shows that in most situations, stock-holding will deliver more value compared to a MTO situation. Only when the profit margins are very small MTO will be the best solution. The analysis does point out that decisions on the CODP should be taken in accordance with the strategic characteristics that hold at that time.

5.5 To what extent and how can the CODP be relocated by coordination efforts with

businesses downstream of the ISM?

(42)

42 communication and agreements with the customer. The following sections will show why this should be done and how this can be initiated.

5.5.1 Reducing the number of SKU’s

Looking at Figure 11, we can see that Company X orders 70 different SKU’s, from eight types of steel grades. Every SKU is unique and thus non-interchangeable with another SKU. The consequence of this is that total MTS production means that you have to keep 70 different items on stock. This also implies that you will stock products that face very irregular or minimal demand: this is not very efficient from an inventory costs point of view. Therefore, a selection of products will be kept on stock. The subsequent section outlines how this is currently performed at Tata Steel.

In addition to a MTO mode of production, which is still the most applied strategy for Tata Steel as well as at similar ISM’s, there are developments to shift the CODP much more downstream. The philosophy that is applied at Tata is based on the work of Eliyahu Goldratt and his theory of constraints (TOC). In TOC, a new order fulfilment strategy and commercialization of this strategy as an alternative to MTS is outlined, called Make-to-Availability (MTA). MTA is defined as “the manufacturer’s general declaration to provide immediate supply whenever needed” (Schragenheim, Dettmer and Patterson, 2009, p.92).

In terms of the CODP, MTA is a pure MTS-system. Nonetheless, there are some key differences between traditional MTS modes of production and the MTA as it is being implemented at Tata: the first and foremost difference is the way that stock is being replenished. Whereas in most MTS systems, stock is being replenished based on a forecast of customer orders, MTA is based on actual demand. Through a system called dynamic buffer management (DBM), which is Kanban-style system, the end-stock buffer is maintained through a pipeline of end-stock. When the order volume and variance are reasonably constant, this system can serve all orders within a very short lead time. When the order volume and variance get out of bounds, the sizes of the buffers are updated to reflect the new parameters. The main advantage of this system compared to MTS is the absence of forecasts. Only the initial buffer setting is based on a rough-estimate or forecast, but subsequently no forecast information is used. Forecasting can be time-consuming and inaccurate, furthermore forecasts are not always interpreted and used correctly (Xu, Fang and Zhu, 2013). Bypassing the need for an accurate forecast is the main reason that Tata is adopting the MTA strategy.

(43)

43 frequency of the SKU’s. Although the classification of products in runners, repeaters and strangers can be found in literature (Jina, Bhattacharya, and Walton, 1997; Armistead, 1996), the scale of the parameters are industry-specific and thus the classification that originates from Tata is assumed to be suitable.

Segment Variance in demand: Frequency:

Runner Relative standard deviation <100% And Ordered in >80% of weeks

Repeater Relative standard deviation <200% And Ordered in >30% of weeks Stranger Relative standard deviation >200% Or Ordered in <30% of weeks

Table 7: Classification of SKU’s into runner, repeater and stranger segments

5.5.2 Driving processes for SKU reduction

The company-wide implementation of MTA is more an overhaul of the way orders are being accepted, produced and delivered, than solely the placement of the CODP. Just like MTS, MTA requires relatively stable demand and medium to large size order volumes. Consequently, the entire organization is affected; in order to fulfil more orders on a MTA-basis, the number of end products is decreased (through product rationalization and overspecification), the number of individual SKU’s from an administrative perspective is decreased (order book rationalization), orders can be fulfilled by either UK or IJmuiden production facilities (homologation) and hybrid MTO/MTS systems emerge (delayed differentiation). Discussing all of these measures is too extensive for this research. Therefore only product rationalization and overspecification will be outlined here.

Product rationalization

(44)

44 processes. It is important however not to rationalize in markets in which Tata wants to differentiate, because customers will expect that you can deliver products.

Overspecification

(45)

45

6 Discussion

The results that have been presented are very interesting from both a theoretical and practitioner’s point of view. However, it must once again be acknowledged that this has been a very exploratory research. The limited time and resources available did not allow for a more precise and extensive quantitative model to support the conclusions. Nonetheless, the findings have some important implications for both theory and practice.

6.1 Assumptions

There are a few assumptions that have been made that can be challenged. One of these assumptions has been outlined in 2.4.1; the assumption that costs will increase when a forward shift is initiated. When a sufficient fraction of orders is fulfilled in MTS mode the company will experience a cost decrease because of increased opportunities for process improvements. However, the results already indicate that a move towards more MTS production is a good idea. If the possible process improvement opportunities would be added to the list of benefits, this forward shifting strategy would be strengthened even more.

Another assumption is that the order volume will grow when the delivery performance is increased. Although this assumption may seem reasonable, it is also only driver for the forward shifting process. When the order volume growth will not be realized, even when delivery performance is significantly improved, inventory costs will rise but no benefits are gained. It is hard to quantify the effect of increasing delivery performance on order volume growth because the latter is hard to measure: If you get more orders next year than this year, is this due to increased delivery performance? This is very hard to tell because there are many factors that can contribute to order volume growth.

The s-curve shaped relationships used to perform the necessary calculations can be disputed. Even when the basic assumption of a non-linear relationship between the variables is established, it is debatable what the exact relationship will be. The s-curve are however believed to give a somewhat prudent estimation of reality, because they take slow initial growth into account. This cautiousness improves the credibility of the results.

(46)

46 to stock items, assuming that the order growth is substantial. The same holds for decreasing cost and market prices. Only when the market price is dropping faster or increasing considerably slower than the cost price, the results might become different.

6.1.1 Single case limitations

The volatility of decoupling points in a changing strategic climate A case study at Tata Steel