Content uploaded by Namchul Shin
Author content
All content in this area was uploaded by Namchul Shin on Nov 21, 2014
Content may be subject to copyright.
Research Paper
Value Capture in the Global
Electronics Industry: Empirical
Evidence for the “Smiling Curve”
Concept
NAMCHUL SHIN*, KENNETH L. KRAEMER** & JASON DEDRICK†
*Seidenberg School of Computer Science and Information Systems, Pace University, New York, USA
**University of California, Irvine, USA, †Syracuse University, Syracuse, New York, USA
ABSTRACT This research asks who captures the greatest value in the global electronics industry by
testing the concept of the “smiling curve”, which predicts that the greatest value is captured by upstream and
downstream firms, and the lowest value is captured in the middle of the value chain. We test the concept using
the Electronic Business 300 data-set for 2000 –2005. We find that lead firms and component suppliers earn
higher gross margins and net margins compared to contract manufacturers. However, the differences are
minimal for return on assets (ROA) and return on equity (ROE). We also find that active component suppliers
gain higher profits than passive component suppliers. These findings suggest that the smiling curve is right if
value is defined in terms of gross margins, but the cost of sustaining a position on either end of the curve is so
high that returns on investment are similar across the curve.
KEY WORDS: Electronics industry, value chain, smiling curve, lead firm, component supplier
1. Introduction
In today’s global electronics industry, companies outsource production and even product
development to global networks of contract manufacturers (CMs), original design
manufacturers (ODMs) and component suppliers. In such global production networks, value
created from a successful product is distributed not only to a lead firm, usually the company
whose brand appears on the product, but also to partners in the firm’s value chain, such as
componentsuppliers as well as CMs/ODMs. While the lead firm captures a significant portion of
1366-2716 Print/1469-8390 Online/12/020089 –19 q2012 Taylor & Francis
http://dx.doi.org/10.1080/13662716.2012.650883
Correspondence Address: Namchul Shin, Seidenberg School of Computer Science and Information Systems, Pace
University, 163 William Street, New York, NY 10038, USA. Tel.: þ1 212 346 1067; Fax: þ1 212 346 1863; Email:
nshin@pace.edu
Industry and Innovation,
Vol. 19, No. 2, 89–107, February 2012
Downloaded by [Namchul Shin] at 05:10 20 February 2012
the value by focusing on the creation, penetration and defense of markets for the product, other
firms also benefit by pursuing core technological innovations and offering complementary
products or services. Since no single company is the source of all innovations, a leadfirm works
closely with global partners to bring new products to market (Linden et al.,2009).
This paper addresses the question of who captures the greatest value in such global value
chains by empirically testing the “smiling curve” concept in the electronics industry. The smiling
curve (Shih, 1996; Everatt et al., 1999) or the smile of value creation (Mudambi, 2008) repre-
sents a pattern of value-added along the value chain. It predictsthat higher value is added both
upstream (at the input end) and downstream (at the output end), with the lowest value-added in
the middle of the value chain. From the firm’sperspective, however, the goal is not to add value
but to capture value in the form of profits. Thus, this research examines whether the “smiling
curve” concept can be applied to value captured by firms ineach part of the global value chain.
1
This research can inform us about the importance of position in the value chain for the
profitability of firms, and also for the financial benefits to countries participating in global
value chains. Thus, beyond the question of which firms capture more value, we also raise a
question about the value captured by countries, particularly between advanced and
emerging economies. Given that there tend to be more lead firms and component suppliers
in advanced economies, the smiling curve would predict that the value captured by firms in
these economies is higher than that by firms in newly emerging economies, which tend to
specialize more in labor-intensive assembly.
We hypothesize these relationships and test the hypotheses by using data from the
Electronic Business 300 data-set. We find that lead firms and component suppliers capture
more value as measured by gross margin and net margin, compared to various contract
manufacturers (e.g. CMs/ODMs). We also find that active component suppliers gain higher
profits than passive component suppliers
2
and that firms based in advanced economies earn
higher value in terms of gross margins, compared to firms based in emerging economies. Our
findings suggest that high levels of innovation, sales and marketing, and branding can build
barriers to entry and help firms capture higher profits in global production networks.
In the next section, we describe the concept of the “smiling curve”, analyze the concept
based on resource-based theory, dynamic capabilities and industrial economics, and propose
hypotheses. Section 3 describes our research methods and data sources. We present our
results in Section 4. Implications of the results and conclusions are provided in Section 5.
2. Theoretical Background
2.1 The Concept of the Smiling Curve
A firm’s value chain activities can be broadly grouped into three categories: the upstream
(input), the downstream (output or market) and the middle (Mudambi, 2007, 2008). While
1
Value capture, which can be indicated by gross profit, does not equal value-added because it excludes the amount
of wages for direct labor (workers who are involved in production, that is, converting inputs to a salable product). Gross
profit estimates the value a company captures from its role in the value chain, which it can use to reward shareholders
(dividends), invest in future growth (R&D), cover the cost of capital depreciation and pay its overhead expenses
(marketing and administration) (Linden et al., 2009).
2
The classification of active and passive components is described in Section 3.1.
90 N. Shin et al.
Downloaded by [Namchul Shin] at 05:10 20 February 2012
upstream activities comprise design, basic and applied R&D, downstream activities typically
comprise marketing, distribution, brand management and after-sales services. Activities in
the middle comprise manufacturing, assembly and other repetitious processes in which
commercialized prototypes are implemented on a mass scale.
Based on his analysis of the computer industry’s value chain, Acer founder Stan Shih
(1996) argued that the value-added curve of the industry takes a smiling shape. The smiling
curve shows that while higher value is created by both upstream and downstream firms
(located at both the left and right side of the curve), such as component suppliers and lead
firms, system assembly firms (located in the middle) add the lowest value (Figure 1).
According to Shih, the major factors determining the level of value-added are entry barriers
and accumulation of capability: the higher the entry barriers and the greater the
accumulation of capabilities, the higher the value-added.
3
For example, the establishment of
a brand name business in microprocessor manufacturing comes with high entry barriers
such as intellectual property and brand equity, and requires many years of investment in
R&D and marketing (branding), respectively. On the other hand, entry barriers and switching
costs are lower for computer assembly because it is relatively easy to build the needed
capabilities and therefore subject to rapid imitation and intense competition. In fact, Acer
itself spun off its ODM business as a separate company, called Wistron, and concentrated
its own efforts on developing its brand name business in order to avoid the commodity
assembly trap.
4
This research employs the resource-based theory and dynamic capabilities approach
as well as industrial organization to analyze the concept. These theoretical approaches look
at entry barriers but with different foci. Industrial organization focuses on industry forces,
whereas the resource-based theory and dynamic capabilities approach focuses on
resources and capabilities (difficult to replicate), respectively. However, they are closely
related, and our research emphasizes the resource-based theory and dynamic capabilities
approach.
2.2 Barriers to Entry: Resources and Dynamic Capabilities
Industrial organization economics considers entry barriers as the fundamental prerequisite
for market power that confers large profits (monopoly rents) (Baumol et al., 1982). It focuses
on the external environment, emphasizing industry attractiveness as the primary basis for
superior profitability. Observing that competition for profits goes beyond direct competitors,
Porter (1980) extends the concept of industry rivalry based on five competitive forces that
include customers, suppliers, potential entrants, substitute products and direct competitors.
He argues that this extended rivalry defines an industry’s structure and shapes the nature of
3
Although Shih uses value-added for the “smiling curve” concept, he implicitly seems concerned with sustainable
incomes (value capture) that are delivered to firms positioning themselves in different ways in global value chains.
4
Acer, a leading manufacturer of notebook and desktop computers, spun off its contract manufacturing service unit
as Wistron in 2001 when its sales were slumping in a weakening computer hardware market. By separating its branded
and contract manufacturing operations, Acer could focus on its branded computer product operations. More recently,
Asustek announced in December 2009 its plans to spin off its contract manufacturing unit as Pegatron Technology.
The company was looking to focus more on creating its own branded line of business. In each case, the spinoff
company was free to pursue other customers and gain economies of scale.
Value Capture in the Global Electronics Industry 91
Downloaded by [Namchul Shin] at 05:10 20 February 2012
competition within an industry; thus, firms should analyze their external environment,
choose strategies that give them competitive advantage in that environment and then
acquire the resources needed to implement their strategies. However, he places little
emphasis on the importance of idiosyncratic firm attributes (e.g. resources) for a firm’s
competitive advantage.
By contrast, the resource-based theory suggests that in a rapidly changing environment
in which customer demand is volatile, and technology is continually evolving, an externally
focused orientation does not provide a secure foundation for formulating long-term strategy
(Barney, 1991).
5
Although Porter tends to emphasize strategic positioning in terms of cost
leadership, differentiation and focus as the primary source for superior profitability,
6
fundamental to strategic choices is the resource position of the firm: a firm gains and
sustains large profits from resources that are rare, valuable, hard to imitate and immobile
(Grant, 1991). Grant argues that barriers to entry are built up by resources that incumbent
firms possess such as scale economies, patents, brand value and customer relationships,
which new entrants can acquire only slowly or at disproportionate expense. Barney (1991)
also argues that barriers to entry exist when competing firms are heterogeneous in terms of
the strategic resources they control.
While resources include firms’ tangible and intangible assets, capabilities refer to a
firm’s ability to appropriately deploy, coordinate and integrate its resources for production
(Grant, 1991; Teece et al., 1997; Coombs and Bierly, 2006). The dynamic capabilities
Value
added
Component-level
R&D Product-level R&D,
marketing and branding
Assembly
Passive components:
capacitors and resistors
motherboards
Component suppliers CMs/ODMs Lead firms
Active components:
key integrated circuits
hard drives
visual displays
Figure 1. Smiling curve: adapted from Shih (1996)
5
The resource-based theory is related to the work of David Ricardo (1891), Joseph Schumpeter (1934) and Edith
Penrose (1959). The returns to the resources that confer competitive advantage are referred to as Ricardian rents,
compared to monopoly rents, that is, the returns to market power (Grant, 1991).
6
Cost leadership, differentiation and focus are proposed by Porter (1980) as a set of generic strategies that can
help firms gain competitive advantage in an industry.
92 N. Shin et al.
Downloaded by [Namchul Shin] at 05:10 20 February 2012
approach explains the sources of competitive advantage over time in competitive markets
(Teece and Pisano, 1994; Teece et al., 1997). It emphasizes firm capabilities (difficult to
replicate) for superior firm performance, which enable firms to sense changing customer
demands and technological opportunities, seize the opportunities by developing new
products through investments in tangible and intangible resources, and maintain
competitiveness through enhancing, integrating, protecting and reconfiguring those
resources (Teece, 2007). According to Teece (2007), in the fast changing global economy
characterized by open innovation, outsourcing and offshoring, particularly in high-technology
sectors, sustainable advantage requires more than the ownership of difficult-to-replicate
knowledge assets. It also requires unique and difficult-to-replicate dynamic capabilities.
7
Chesbrough (2003) argues that as sources of innovation are geographically dispersed, firms
reach out beyond their boundaries to access and integrate technology developed by others
(i.e. open up technological opportunities through engaging in R&D and tapping into the
research output of others).
Knowledge integration capability is critical for superior firm performance in such an
open innovation environment, illustrated by the computer industry (Iansiti and Clark, 1994).
Brusoni et al. (2001) argue that multi-technology firms, such as computer firms, need to
have knowledge in excess of what they need for what they make. They outsource
manufacturing while focusing on in-house concept design and system integration
capabilities to coordinate the work of suppliers, who do new technology development and
manufacturing. Knowledge assets embodied in people and organizational routines are not
tradable and are hard to replicate in a market; thus, the creation, protection, integration and
leverage of such intangible assets is critical for firms to achieve superior firm performance
and avoid the zero-profit trap (Teece, 2007).
Morrison et al. (2008) argue that complex and tacit knowledge may affect the balance of
power and the pattern of governance in global value chains. According to them, buyers (or in
our case, lead firms and branded firms) are undisputed leaders since they coordinate and
govern global value chains, based on knowledge of the whole product system as well as
concept design, branding, marketing and system integration capabilities. Gereffi (1994)
argues that global buyers can and do exert a high degree of control (or power) over spatially
dispersed value chains by building global scale production and distribution systems without
direct ownership. They manage such globally fragmented production networks and bring
together all the pieces of the business into an integrated whole, for example, understand
customer needs and integrate upstream (or component) innovations into new product
developments.
On the upstream end, component suppliers also can generate sustainable high profits
by possessing valuable resources such as intellectual property, superior design skills and
the ability to commercialize new technologies (Gereffi, 2001; Gereffi et al., 2005). Some
suppliers of key components and technologies, such as Intel, Qualcomm, TI and Nvidia, are
7
Teece (2007) argues that supra competitive returns are earned through dynamic capabilities that enable
entrepreneurship, innovation, semi-continuous asset orchestration, resource combinations and reconfiguration. The
returns to dynamic capabilities are referred to as Schumpeterian rents, compared to Ricardian rents, that is, the
returns to resources.
Value Capture in the Global Electronics Industry 93
Downloaded by [Namchul Shin] at 05:10 20 February 2012
able to earn higher profits by controlling key standards, thereby holding dominant positions
in some segments of the industry (Shin et al., 2009).
8
By contrast, firms in the middle of the value chain are not in a position to develop unique
intellectual property, control key product standards, or develop strong brand names or
customer relationships. They must compete largely on cost and operational excellence, and
find it difficult to build any barriers to entry or create switching costs for their customers. Thus,
we argue that in today’s highly competitive global electronics industry, lead firms and
component suppliers are in a better position to develop valuable resources, build barriers to
entry and capture greater value. These theoretical arguments lead to the following hypothesis:
Hypothesis 1: Firms both at the upstream and downstream ends, such as component
suppliers and brand name vendors, capture higher value than firms in the
middle, such as CMs/ODMs.
2.3 Value Capture by Type of Component
Shih (1996) argues that the level of value-added from component manufacturing (activities at
the upstream end of the value chain) differs by the types of components.
9
Active components,
such as integrated circuits, visual displays and hard drives, generally require large capital
investments and high-level manufacturing capabilities. These components are highly
specialized, compared to passive components, such as capacitors and resistors, or printed
circuit boards (e.g. motherboards), which are more standardized.
10
Active components are
capable of a greater degree of differentiation and perhaps even branding, such as the “Intel
inside” branding campaign. Performance aspects of active components are likely to be more
visible to the final customers than other components. For example, most customers would
recognize the difference between 50GB and 500GB hard drive while few would recognize
the implications of improvements in the performance of capacitors or resistors. Therefore,
active components are at the higher left side of the curve while passive components are lower
on the curve. Shih (1996) ranked the level of value-added from component manufacturing in
the following order (from high to low): software, microprocessor, DRAM, LCD, ASIC, monitor,
HDD and motherboard. Therefore, we propose the following hypothesis:
Hypothesis 2: Firms manufacturing active components capture higher value than firms
manufacturing passive components.
2.4 Value Capture by Country
According to Mudambi (2008), firms from advanced economies, those from emerging
economies and those from recently developed countries are all conforming to the smiling
8
According to Sturgeon (2003), since standards and protocols are dynamic, major advantages accrue to companies
that actively participate in the rule-setting process, which favors established firms and locations. Most other value
chain participants, such as CMs/ODMs, must adjust to the rules (or parameters) developed by those firms.
9
We are grateful to anonymous reviewers for helping to clarify this discussion.
10
Active components are those that require electrical power to operate. This could include the power supply, fans,
storage device, transistors, diodes and other integrated circuits. Passive components such as the chassis, capacitors
or enclosures do not require electrical power to operate.
94 N. Shin et al.
Downloaded by [Namchul Shin] at 05:10 20 February 2012
curve: high-value activities at the downstream and upstream ends are largely concentrated in
advanced economies, while low-value activities in the middle of the value chain are moving
(or have moved) to emerging economies (Gereffi, 1999; Smakman, 2003; Pyndt and
Pedersen, 2006). Firms based in advanced economies (so-called insiders), such as the Triad
of North America, Europe and Japan, are more likely to capture higher value, compared to
firms based in emerging economies such as Taiwan, China and Korea (Spencer, 2003;
Mudambi, 2008; Shin et al., 2009).
11
These established players were earlier entrants and
established barriers to entry in many high-value segments of the industry. They continually
innovate in order to maintain their competitive advantage, while startups such as Qualcomm,
Broadcom and Nvidia have become highly profitable as fabless chip designers. A few
emerging country firms have been able to develop successful brand names (e.g. Samsung,
LG, Acer, HTC, Huawei) or compete in certain component markets (e.g. Samsung in memory
chips and displays), but these are the exception. The innovations in developed countries are
increasingly design driven, recognizing the highly diverse needs of individual markets. All
these firms’ design strategies are aimed at buttressing and enhancing the value of their
brands (Mudambi, 2008). Hence, we propose the following hypothesis:
Hypothesis 3: Firms based in advanced economies capture higher value than firms in
emerging economies.
3. Research Methods
In order to test the hypotheses proposed in the previous section, we employ the one-way
analysis of variance (ANOVA—F-test) procedure, the non-parametric
x
2
(Kruskal and
Wallis) and median tests.
Although the one-way ANOVA is a method of our choice for testing for differences
between multiple groups, it assumes that the variances of the groups are equal and that the
distribution of the test variable is reasonably normal. ANOVA is robust to unequal variances
when the groups are of equal or near equal size. However, when both the variances and the
sample sizes differ, we may need to transform the data (for example, the log transformation)
or perform a non-parametric test (Norusis, 2004). Non-parametric procedures are designed
to test for the significance of the difference between multiple groups when the assumptions
of ANOVA are invalid or suspect. They make no assumptions about the mean and variance
of a distribution, nor do they assume that any particular distribution is being used (Conover,
1980; Siegel and Castellan, 1988; Norusis, 2004). We employ the non-parametric
x
2
(Kruskal and Wallis) and median tests for the robustness of our analysis.
3.1 Data Sources and Coding
This study employs two data sources: the Electronic Business (EB) 300 data-set and the
Hoovers database for the six years from 2000 to 2005. The EB 300 data-set includes
11
Korea, and to an extent Taiwan, might be somewhere in between, as they have some major brand name companies
like Samsung and Acer. They are also major suppliers of high-value components like LCDs and DRAM although they
do not compete in software, microprocessors and specialized chips such as graphics. Taiwan and Korea also have far
higher GDP per capita than China and most of South East Asian countries.
Value Capture in the Global Electronics Industry 95
Downloaded by [Namchul Shin] at 05:10 20 February 2012
the top 300 electronics firms ranked by electronics revenue. The electronics revenue is
derived from segmentation information and Reed Research estimates (Electronic Business,
2006). It includes revenue from the sale, service, license or rental of electronics/computer
equipment, software or components. Data items such as sales, cost of goods sold (COGS),
return on equity (ROE), return on assets (ROA), R&D expense and the number of
employees are obtained from the Hoovers database for the same firms included in the EB
300 data-set.
For measures of value capture, we use gross profit (net sales minus cost of goods sold,
which combines the wage bill with the cost of purchased inputs) and gross margin (the ratio
of gross profit to net sales). We also employ net margin, ROA and ROE to measure a
company’s bottom-line financial performance.
Since we focus on three types of firms in the global electronics industry as lead firms,
CMs/ODMs and component suppliers, we select only the firms operating in the following four
industries: computer and peripheral equipment manufacturing, communications equipment
manufacturing, audio and video equipment manufacturing, and semiconductor and other
electronic component manufacturing.
12
The selection is based on the four-digit North
American Industry Classification System (NAICS) code. The NAICS codes for the above
four industries are 3341, 3342, 3343 and 3344, respectively.
We code these firms as lead firms, CMs/ODMs or component suppliers. Lead firms are
branded firms at the head of a value chain and closest to distribution and retail. We only
include firms if they can be classified as “pure” lead firm, CM/ODM or component supplier.
13
Then, we classify component suppliers further into two categories: active and passive
component suppliers.
14
A passive component refers to a component that consumes energy,
but does not produce power. An active component is a component that produces power by
consuming energy. We use the Yearbook of World Electronics Data (2003) for the
classification. Active components include most key components, such as visual displays,
hard drives and key integrated circuits. On the other hand, passive components include
such components as capacitors, resistors, connectors and motherboards.
We also code these firms as firms based in advanced economies (insiders) and firms
based in emerging economies (outsiders). Our samples include firms in 14 different
countries, such as the USA, Canada, Germany, Switzerland, Netherlands, Finland,
Sweden, France, Japan, Taiwan, South Korea, Singapore, Hong Kong and China. We
classify North American, European and Japanese firms into firms in advanced economies,
and other Asian firms into firms in emerging economies. The sample includes 622
observations for the six years from 2000 to 2005. The sample statistics are shown in Table 1.
12
Other industry segments are left out because most firms in those segments cannot be classified as pure lead firms,
CMs/ODMs or component suppliers, and electronic revenue of those firms does not equal total revenue. Highly
integrated firms or large conglomerates are not included since they have mixed sales figures such as sales from brand
products, from contract manufacturing and from components.
13
We compared the sample firms (622 observations for 2000– 2005) and omitted firms (1,178 observations for 2000–
2005) in terms of electronic revenue, gross margin, net margin, ROA and ROE. By conducting the ANOVA, the non-
parametric
x
2
and median tests, we found that the omitted firms were not systematically different from the sample
firms for all of the measures.
14
We only include firms that can be classified as pure active and passive component suppliers. Diversified firms, as
well as other types of firms, such as contract component manufacturers and storage firms are excluded.
96 N. Shin et al.
Downloaded by [Namchul Shin] at 05:10 20 February 2012
Table 1. Sample statistics (2000 – 2005)
Lead firms CMs/ODMs Component suppliers
Variables Mean Std. dev. Obs. Mean Std. dev. Obs. Mean Std. dev. Obs.
Sales (millions) 11,069.5 15,010.6 204 5,398.6 4,925.5 112 3,424.1 4,573.0 306
Gross profit (millions) 4,275.6 5,053.2 160 719.2 1,783.5 74 1,321.1 2,720.8 275
Gross margin (%) 31.90% 14.12% 160 10.67% 14.22% 74 32.29% 16.65% 275
Net profit (millions) 74.0 2,820.5 204 227.8 589.2 106 31.7 3,523.3 304
Net margin (%) 0.61% 17.95% 204 20.44% 8.83% 106 1.25% 24.63% 304
ROA (%) 0.97% 18.57% 163 0.04% 12.01% 94 20.10% 31.47% 293
ROE (%) 27.51% 154.2% 162 22.24% 35.51% 90 24.11% 101.5% 284
R&D expense (millions) 1,163.2 1,435.8 159 45.4 51.3 48 445.0 746.2 232
R&D ratio (% of sales) 8.36% 5.72% 159 0.90% 1.17% 48 11.38% 7.71% 232
S&GA costs (% of sales) 19.17% 7.64% 156 5.08% 5.92% 71 12.56% 6.48% 273
Employees (thousands) 35.0 36.2 160 28.7 35.4 92 16.6 16.4 274
Value Capture in the Global Electronics Industry 97
Downloaded by [Namchul Shin] at 05:10 20 February 2012
4. Results
4.1 Comparison of Lead Firms, Component Suppliers and CMs/ODMs
Our results show that value captured by the three types of firms (lead firms, component
suppliers and CMs/ODMs) are significantly different for gross margin, gross profit and net
margin (Table 2): lead firms and component suppliers earn profits higher than CMs/ODMs. All
three test statistics of ANOVA (except for net margin), non-parametric
x
2
and median tests
are significant at a level of 0.001. However, the differences are minimal for ROA. On the other
hand, CMs/ODMs perform better than lead firms and component suppliers in terms of ROE
although it is negative for all three types of firms. The non-parametric
x
2
and median tests are
significant at levels of 0.10 and 0.05, respectively. Our results also show that lead firms and
component suppliers spend more on selling and general administration (S&GA) expense and
invest more in R&D, compared to CMs/ODMs.
Positioned close to the consumer markets in the global value chain, lead firms specialize in
high value-added activities such as R&D, product design, marketing and branding. They may
have a well-known brand, better marketing and sales capabilities, and a keen understanding of
Table 2. ANOVA, non-parametric
x
2
(Kruskal–Wallis) and median test results (2000 –2005)
NMean F
x
2
Median test (
x
2
)
Gross margin Lead firms 160 32.3% 60.320*** 131.887*** 77.370***
CMs/ODMs 74 10.7%
Comp. suppliers 275 31.9%
Ln(gross profit) Lead firms 160 7.57 65.776*** 88.832*** 40.567***
CMs/ODMs 74 5.91
Comp. suppliers 272 6.54
Net margin Lead firms 204 0.61% 0.271 18.340*** 19.816***
CMs/ODMs 106 20.44%
Comp. suppliers 304 1.25%
ROA Lead firms 163 0.97% 0.096 4.236 2.792
CMs/ODMs 94 0.04%
Comp. suppliers 293 20.10%
ROE Lead firms 162 27.51% 0.074 4.617
þ
6.055*
CMs/ODMs 90 22.24%
Comp. suppliers 284 24.11%
R&D/sales Lead firms 159 8.36% 51.909*** 116.291*** 59.810***
CMs/ODMs 48 0.90%
Comp. suppliers 232 11.38%
Ln(R&D) Lead firms 157 5.97 61.860*** 89.110*** 50.051***
CMs/ODMs 47 3.30
Comp. suppliers 232 5.35
S&GA/sales Lead firms 156 19.17% 111.625*** 185.980*** 113.038***
CMs/ODMs 71 5.08%
Comp. suppliers 273 12.56%
Notes: ROA ¼return on assets; ROE ¼return on equity; S&GA ¼selling and general administration expense.
The log transformation of net profit is not used because the number of observations with a negative value is high.
***p,0.001; *p,0.05;
þ
p,0.10.
98 N. Shin et al.
Downloaded by [Namchul Shin] at 05:10 20 February 2012
customers, compared to CMs/ODMs (Shin et al., 2009). Component suppliers, particularly
suppliers of visual displays, hard drives or key integrated circuits, invest heavily in R&D and
pursue high levels of innovation by embodying proprietary knowledge, compared to
CMs/ODMs. Such capabilities as branding (for lead firms) and R&D (for component suppliers)
create entrybarriers and help lead firms and component suppliers gain higherprofits. However,
the costs of conducting R&D, sales and marketing can negatively affect both lead firms’ and
component suppliers’ bottom-line financial performance. Net margins are still higher for lead
firms and component makers than for CMs/ODMs, but while the difference is statistically
significant, it is very small in practical terms. Lead firms and component makers both earn
average grossmargins of about 32 per cent,compared to 10 per cent for CMs/ODMs. However,
the net margins are 0.61, 1.25 and 20.44 per cent, respectively (Table2). We do not detect any
significant differences in ROA among the three types of firms, and ROE is higher for CMs/
ODMs, compared to lead firms and component suppliers. The reason might be that in contrast
to ROA and ROE, net margin is not dependent on differences in asset intensity or equity (or
debt) financing.
15
More significant from an investment perspective is the fact that ROE is
negative for all three groups, illustrating how brutally competitive the electronics industry is.
Figure 2 depicts the means of gross margins of lead firms, component suppliers and
CMs/ODMs. Interestingly, the shape of the mean plot is similar to the “smiling curve” shown
in Figure 1.
16
Figure 3 shows the mean plot for ROA, along with the mean plot for gross margin, for
the three types of firms. The shape of the mean plot for ROA is somewhat distorted (tilted
“smiling curve”), compared to the mean plot for gross margin.
Figure 4 shows the mean plot for ROE, along with the mean plot for gross margin, for
the three types of firms. The shape of the mean plot for ROE looks the reverse of the “smiling
curve”.
17
As mentioned earlier in Section 2.1, some lead firms, such as Acer and Asustek,
separated their branded and contract manufacturing operations in order to focus more on
their branded product operations. Motorola also spun off its upstream component business
(chip fabrication) as Freescale Semiconductor, and outsourced more of its production. Such
a strategy is consistent with our findings.
In order to directly examine if the spinoffs of these lead firms capture more value, we
analyze the post-spinoff performance of Acer and Motorola.
18
Our results show that the
15
Our results for ROE might be influenced by the high leverage of CMs/ODMs. Since the leverage of those firms can
be higher than lead firms and component suppliers, we analyze the differences of the debt-to-equity ratio for the three
types of firms. The non-parametric median test shows that the debt-to-equity ratio is significantly higher for
CMs/ODMs, compared to lead firms and component suppliers. However, the results are insignificant for the one-way
ANOVA test and opposite for the Kruskal –Wallis
x
2
-test.
16
The lines are straight because the figure plots the means of gross margins of the three types of firms categorized into
discrete variables.
17
The differences in results between ROA and ROE look strange since the two measures are closely related. ROE can
be decomposed into: ROE ¼net income/equity ¼net income/assets (ROA) £assets/equity (leverage). Therefore,
when ROA increases, ROE may increase. However, ROE is also affected by leverage. Figure 4 shows the impact of
the leverage of CMs/ODMs (please refer to footnote 15).
18
We do not include Asustek in the analysis since its spinoff was announced fairly recently (December 2009). It would
be interesting if future studies conduct the analysis with more firms, including not only lead firms, but also component
suppliers, which have spun off their contract manufacturing or fabrication operations.
Value Capture in the Global Electronics Industry 99
Downloaded by [Namchul Shin] at 05:10 20 February 2012
companies’ bottom-line financial performance has significantly improved after the spinoffs
(Table 3). Acer’s performance improved in 2002 after spinoff: ROA (from 0.9 to 7.9 per cent),
ROE (from 1.7 to 13.0 per cent) and net margin (from 0.9 to 8.1 per cent). The performance
of Motorola has also improved in 2005 after spinoff: ROA (from 5.0 to 12.8 per cent), ROE
(from 11.5 to 27.5 per cent) and net margin (from 4.9 to 12.4 per cent). We also compare the
post-spinoff performance of the two lead firms to the firms spun off (i.e. Acer vs. Wistron and
Motorola vs. Freescale Semiconductor). Using the performance of three-year averages, we
found that these lead firms (Acer and Motorola) have outperformed the spun-off firms
(Wistron and Freescale Semiconductor), respectively, in terms of ROA, ROE and net
margin. The comparison of the post-spinoff performance is shown in Table 4.
Mean of
gross margin
Component suppliers Lead firmsCMs/ODMs
35%
30%
25%
20%
15%
10%
Figure 2. Mean plot for gross margin
Mean of gross margin (Black)
Component suppliers Lead firmsCMs/ODMs
35%
30%
25%
20%
15%
10%
Mean of ROA (Red)
1.00%
0.80%
0.60%
0.40%
0.20%
0.00%
Figure 3. Mean plots for ROA and gross margin
100 N. Shin et al.
Downloaded by [Namchul Shin] at 05:10 20 February 2012
4.2 Comparison of Active and Passive Component Suppliers
Table 5 shows that value captured by active and passive component suppliers is
significantly different in terms of gross margin and gross profit: active component suppliers
gain higher profits than passive component suppliers. All three test statistics of ANOVA,
non-parametric
x
2
and median tests are significant at a level of 0.001. However, the
differences are not significant for net margin and ROA. Passive component suppliers
perform better than active component suppliers when ROE is employed as a performance
measure (except for the median test).
Our results also show that active component suppliers invest heavily in R&D, compared
to passive component suppliers. Such heavy investment in R&D enables active component
suppliers to introduce new components to market and charge a high premium, thus earning
higher gross profits. However, these higher margins are negated by the cost of conducting
R&D, so their bottom-line financial performance measured as net margin and ROA is not
significantly different from passive component suppliers, and their returns on equity are even
lower than the ones for passive component suppliers.
19
Mean of gross margin (Black)
Component suppliers Lead firmsCMs/ODMs
35%
30%
25%
20%
15%
10%
–2.00%
–4.00%
–6.00%
–8.00%
Mean of ROE (Red)
Figure 4. Mean plots for ROE and gross margin
19
We ran a regression analysis to examine the impact of R&D on gross profit, ROA and ROE of active and passive
component suppliers. The analysis was conducted with R&D and a dummy for active component suppliers, along with
an interaction term of active component suppliers and R&D. We found a significant coefficient on the interaction term
for gross profit, but not for ROA and ROE. These results imply that R&D has a stronger impact on performance as
measured by gross profit (i.e. value capture), but not by ROA and ROE, in active component suppliers as compared to
passive component suppliers. These findings are consistent with our theoretical speculation that active component
suppliers capture higher gross profits from their R&D investment, but the cost of conducting R&D negatively affects
their bottom-line financial performance. We also examined the impact of R&D on gross profit, ROA and ROE of lead
firms and passive component suppliers. We found similar results: a significant coefficient on the interaction term of
lead firms and R&D for gross profit, but not for ROA and ROE. These results also imply that R&D has a stronger impact
on gross profit, but not on ROA and ROE, in lead firms as compared to passive component suppliers. Overall, these
findings are consistent with the results of our earlier work conducting a multivariate analysis for lead firms and non-lead
firms (Shin et al., 2009).
Value Capture in the Global Electronics Industry 101
Downloaded by [Namchul Shin] at 05:10 20 February 2012
Table 3. Post-spinoff performance of Acer and Motorola
Company
Spinoff
year Year
ROA
(%) ROE (%)
Gross
margin (%)
Net
margin (%)
Sales
(millions)
Firms
spun off
Acer 2001 2003 6.2 11.3 13.4 4.6 4,622.5 Wistron
2002 7.9 13.0 13.6 8.1 3,089.0
2001 0.9 1.7 13.3 0.9 3,232.1
2000 5.1 10.7 9.8 4.3 4,760.6
Motorola 2004 2006 9.5 21.4 30.8 8.5 42,879.0 Freescale
Semiconductor, Inc.2005 12.8 27.5 32.0 12.4 36,843.0
2004 5.0 11.5 33.5 4.9 31,323.0
2003 2.8 7.0 33.1 3.3 27,058.0
Table 4. Comparison of post-spinoff perform ance (three-year average)—Acer vs. Wistron and Motorola vs. Freescale
Semiconductor, Inc.
Years ROA (%) ROE (%) Gross margin (%) Net margin (%) Sales (millions)
Acer 2003– 2005 5.57 11.93 9.50 4.23 6,144.2
Wistron 2.63 5.63 6.13 1.10 3,676.2
Motorola 2005– 2007 7.43 16.30 30.00 6.93 38,781.3
Freescale 25.41 226.99 39.13 218.42 6,066.4
Table 5. ANOVA, non-parametric
x
2
(Kruskal–Wallis) and median test results (2000 –2005)
NMean F
x
2
Median test (
x
2
)
Gross margin Active 161 37.38% 25.544*** 25.293*** 11.698***
Passive 54 24.69%
Ln(gross profit) Active 158 6.80 21.150*** 18.837*** 14.312***
Passive 54 6.09
Net margin Active 174 21.06% 2.632 0.203 0.111
Passive 63 5.26%
ROA Active 166 22.75% 2.371 2.673 1.884
Passive 63 5.17%
ROE Active 163 23.77% 6.854** 4.369** 1.825
Passive 61 16.63%
R&D/sales Active 149 14.37% 50.946*** 43.131*** 35.640***
Passive 30 4.06%
Ln(R&D) Active 149 5.70 41.966*** 34.913*** 22.745***
Passive 30 4.13
S&GA/sales Active 161 13.10% 0.001 0.016 0.627
Passive 53 13.07%
Notes: ROA ¼return on assets; ROE ¼return on equity; S&GA ¼selling and general administration expense. The
log transformation of net profit is not used because the number of observations with a negative value is high.
***p,0.001; **p,0.01.
102 N. Shin et al.
Downloaded by [Namchul Shin] at 05:10 20 February 2012
4.3 Comparison of Firms in Advanced and Emerging Economies (Insiders and Outsiders)
Table 6 shows that firms based in advanced economies capture higher value in terms of
gross margin and gross profit, compared to firms based in emerging economies. All three
test statistics of ANOVA, non-parametric
x
2
and median tests are significant at a level of
0.001. However, the differences are trivial for net margin, ROA and ROE in the ANOVA F-
and Kruskal–Wallis
x
2
-tests.
As shown in Table 6, insiders (firms in advanced economies) spend more money on
R&D and S&GA, compared to outsiders (firms in emerging economies). These firms earn
high levels of profits by recognizing the highly diverse needs of individual markets and
continually doing design-driven innovations.
20
However, these higher gross margins are
offset by the costs of R&D, selling and marketing, and brand building activities, so their
returns are not significantly different from the ones for firms in emerging economies.
In order to see if lead firms and component suppliers are concentrated more in
advanced economies, rather than in emerging economies, we examine the distribution of
lead firms, CMs/ODMs and component suppliers in the two economies. Table 7 (the rows of
per cent within type) shows that there are more lead firms and component suppliers in
advanced economies than in emerging economies (74.6 per cent vs. 25.4 per cent and 79.4
Table 6. ANOVA, non-parametric
x
2
(Kruskal–Wallis) and median test results (2000 –2005)
NMean F
x
2
Median test (
x
2
)
Gross margin Insiders 425 31.32% 49.789*** 53.720*** 32.624***
Outsiders 84 17.42%
Ln(gross profit) Insiders 423 6.90 24.803*** 25.114*** 10.507***
Outsiders 83 6.15
Net margin Insiders 443 20.19% 3.647
þ
1.625 0.918
Outsiders 168 3.36%
ROA Insiders 438 20.20% 0.632 2.514 5.050**
Outsiders 112 1.95%
ROE Insiders 425 25.65% 0.493 2.651 4.325**
Outsiders 110 2.72%
R&D/sales Insiders 344 10.64% 78.201*** 77.418*** 70.219***
Outsiders 94 3.69%
Ln(R&D) Insiders 343 5.66 66.933*** 65.079*** 52.629***
Outsiders 92 4.20
S&GA/sales Insiders 421 14.74% 62.624*** 76.384*** 48.844***
Outsiders 79 7.27%
Notes: ROA ¼return on assets; ROE ¼return on equity; S&GA ¼selling and general administration expense. The
log transformation of net profit is not used because the number of observations with a negative value is high.
***p,0.001; **p,0.01;
þ
p,0.10.
20
Design-driven innovations could be either market- or technology-driven. While lead firms focus on market-driven
innovations, that is, tailoring products to markets, component suppliers focus on technology-driven innovations. Some
lead firms, such as Apple, do both market- and technology-driven innovations.
Value Capture in the Global Electronics Industry 103
Downloaded by [Namchul Shin] at 05:10 20 February 2012
per cent vs. 20.6 per cent). On the other hand, CMs/ODMs are located more in emerging
economies than in advanced economies (54.5 per cent vs. 45.5 per cent). Table 7 (the rows
of per cent within economies) also shows that advanced economies have relatively more
lead firms and component suppliers than CMs/ODMs (88.5 per cent vs. 11.5 per cent),
compared to emerging economies (65.1 per cent vs. 34.9 per cent). These findings suggest
that the three types of firms are not equally distributed across the two economies: that is,
lead firms and component suppliers are largely concentrated in advanced economies. The
test statistics of the Pearson
x
2
and likelihood ratio tests are significant at a level of 0.001.
Figure 5 depicts the distribution of lead firms, CMs/ODMs and component suppliers in
Table 7. Advanced and emerging economies (insiders and outsiders) by types of firms (2000– 2005)
Economies
Advanced economies Emerging economies Total
Type Lead firm Count 150 51 201
% within type 74.6% 25.4% 100%
% within economies 33.8% 29.1%
CM/ODM Count 51 61 112
% within type 45.5% 54.5% 100%
% within economies 11.5% 34.9%
Component supplier Count 243 63 306
% within type 79.4% 20.6% 100%
% within economies 54.7% 36.0%
Total Count 444 175 619
% within type 71.7% 28.3% 100%
% within economies 100% 100%
0
50
100
150
200
250
300
Component
supplier CM/ODM Lead firm
Count
Type of firm
Advanced
Economies
Emerging
Economies
Figure 5. Distribution of types of firms in advanced and emerging economies
104 N. Shin et al.
Downloaded by [Namchul Shin] at 05:10 20 February 2012
advanced and emerging economies. It illustrates that advanced economies conform to the
smiling curve, but emerging economies do not.
5. Discussion and Conclusions
This research empirically analyzes the pattern of value capture in the global electronics
industry. It demonstrates that lead firms and component suppliers, particularly suppliers of
key components, capture most of the value created from a successful product in global
production networks in which production and product development are outsourced to CMs
and ODMs. Compared to CMs and ODMs, lead firms and component suppliers are in a
position to build up higher entry barriers by deploying and integrating such resources as
marketing, branding and intellectual property, thereby capturing higher profits.
The contribution of this research is threefold: first, it provides theoretical support for the
smiling curve concept by showing that its predictions are consistent with theory from the
resource-based view, dynamic capabilities and industrial organization. Although the concept
has been used in prior research, it has not previously been analyzed in relation to existing
theory. This research did so and found strong support.
Second, the research provides empirical evidence for the theoretical predictions. By
applying the concept of the “smiling curve” to analyze value capture among lead firms,
component suppliers and CMs/ODMs, it shows which firms are most likely to profit in global
value chains. There has been limited empirical testing of the smiling curve argument in prior
research.
Third, this research also sheds light on globalization of production networks by showing
the importance of value chain position for capturing higher profit margins in today’s global
electronics industry. For higher profits, a firm can either move downstream and
develop brands or move upstream and develop innovative components. This is not easy
in practice, as CMs and ODMs generally do not possess capabilities in either R&D or
marketing, and face the potential loss of their CM/ODM business if they try to compete with
their own customers. However, a few have made the transition from CM to brand name
vendor (e.g. HTC) or divested their CM/ODM businesses to concentrate on their own brands
(Acer, Asustek).
The “smiling curve” predictions are right if value is defined in terms of gross margins, but
the cost of sustaining a position on either end of the curve (R&D for component suppliers and
sales/marketing for brand name firms) is so high that returns on investment are similar
across the curve. This is what basic economics would predict—if one segment or company
is more profitable than others, then investors will bid up the price until its returns are normal.
What is surprising is that the industry continues to sustain negative returns on equity on
average. Perhaps many money-losing firms remain in business in the hope of developing a
breakthrough product and turning their losses into gains.
From a national policy view, if the goal is to employ high-paid scientists, engineers and
marketing people, then it makes sense to try to move into the upstream and downstream
parts of the value chain, as higher margins captured from such positions in the value chain
can support R&D and marketing activities. However, many developing countries cannot
reasonably aspire to such a goal. It is important for policymakers in these countries to
remember that the companies in the middle of the value chain still make gross profits and
provide jobs for low- to moderate-skilled workers and some engineers and managers.
Value Capture in the Global Electronics Industry 105
Downloaded by [Namchul Shin] at 05:10 20 February 2012
In addition, they provide an opportunity for learning and, possibly, for moving to a better
position along the value chain.
The research suggests several directions for future work. This research categorizes
firms into pure lead firm, CM/ODM or component supplier. Although value chains in the
electronics industry have steadily disintegrated over the past several decades, there are still
major firms, especially in Japan and Korea, with highly integrated operations. Since those
firms can have mixed sales figures such as sales from brand products, from contract
manufacturing and from components, it would be interesting to replicate the present analysis
using sales percentages of different operations for firms to see if the current results still hold.
Future research could also focus on a particular industry such as the semiconductor
industry where fabless chip companies such as Qualcomm and Nvidia outsource
chip manufacturing to contract chip manufacturers (foundries). It would be interesting to
examine who captures the most value in the global semiconductor industry by comparing
fabless chip companies and contract chip manufacturers. Future research could also narrow
down the scope into a particular country, such as Taiwan or China, and examine if the
pattern of the value capture evidenced in this study holds for the country. Although this
research discusses the impacts of innovation (R&D) and branding on value capture by firms
in the global industry, it does not control for country economic variables in the analysis.
Therefore, future research could provide additional understanding about value capture in the
global electronics industry by incorporating such variables into an analysis.
Acknowledgements
This research has been supported by grants from the US National Science Foundation and
the Alfred P. Sloan Foundation. Any opinions, findings and conclusions or recommendations
expressed in this material are those of the authors and do not necessarily reflect the views of
the National Science Foundation or the Sloan Foundation. The authors would like to thank
five reviewers for their valuable comments and suggestions for improvement of this paper.
References
Barney, J. (1991) Firm resources and sustained competitive advantage, Journal of Management, 17, pp. 99 –119.
Baumol, W. J., Panzer, J. C. and Willig, R. D. (1982) Contestable Markets and the Theory of Industrial Structure (New York: Harcourt
Brace Jovanovitch).
Brusoni, S., Prencipe, A. and Pavitt, K. (2001) Knowledge specialization, organizational coupling, and the boundaries of the firm: why do
firms know more than they make?, Administrative Science Quarterly, 46(4), pp. 597 –621.
Chesbrough, H. (2003) Open Innovation: The New Imperative for Creating and Profiting from Technology (Boston: Harvard Business
School Press).
Conover, W. J. (1980) Practical Nonparametric Statistics. 2nd edn (New York: John Wiley).
Coombs, J. E. and Bierly III, P. E. (2006) Measuring technological capability and performance, R&D Management, 36, pp. 421– 438.
Electronic Business (2006) EB top 300 revenue.
Everatt, D., Tsai, T. and Chang, B. (1999) The Acer Group’s China manufacturing decision. Richard Ivey School of Business Case Series
#9A99M009, University of Western Ontario.
Gereffi, G. (1994) The organization of buyer-driven global commodity chains: how U.S. retailers shape overseas production networks, in:
G. Gereffi & M. Korzeniewicz (Eds), Commodity Chains and Global Capitalism (London: Praeger).
Gereffi, G. (1999) International trade and industrial upgrading in the apparel commodity chain, Journal of International Economics, 48(1),
pp. 37– 70.
106 N. Shin et al.
Downloaded by [Namchul Shin] at 05:10 20 February 2012
Gereffi, G. (2001) Shifting governance structures in global commodity chains, with special reference to the Internet, American Behavioral
Scientist, 44(10), pp. 1616– 1637.
Gereffi, G., Humphrey, J. and Sturgeon, T. (2005) The governance of global value chains, Review of International Political Economy,
12(1), pp. 78– 104.
Grant, R. M. (1991) The resource-based theory of competitive advantage: implications for strategy formulation, California Management
Review, 33(3), pp. 114– 135.
Iansiti, M. and Clark, K. B. (1994) Integration and dynamic capability: evidence from product development in automobiles and mainframe
computers, Industrial and Corporate Change, 3(3), pp. 557– 605.
Linden, G., Kraemer, K. L. and Dedrick, J. (2009) Who captures value in a global innovation network? The case of Apple’s iPod,
Communications of the ACM, 52(3), pp. 140 –144.
Morrison, A., Pietrobelli, C. and Rabellotti, R. (2008) Global value chains and technological capabilities: a framework to study learning and
innovation in developing countries, Oxford Development Studies, 36(1), pp. 39 –58.
Mudambi, R. (2007) Offshoring: economic geography and the multinational firm, Journal of International Business Studies, 38(1), p. 206.
Mudambi, R. (2008) Location, control and innovation in knowledge-intensive industries, Journal of Economic Geography, 8(5),
pp. 699– 725.
Norusis, M. (2004) SPSS 13.0 Guide to Data Analysis (Upper Saddle River, NJ: Prentice Hall).
Penrose, E. (1959) The Theory of the Growth of the Firm (New York: John Wiley).
Porter, M. (1980) Competitive Strategy (New York: Free Press).
Pyndt, J. and Pedersen, T. (2006) Managing Global Offshoring Strategies: A Case Approach (Copenhagen: Copenhagen Business School
Press).
Ricardo, D. (1891) Principles of Political Economy and Taxation (London: G. Bell).
Schumpeter, J. A. (1934) Theory of Economic Development (Cambridge, MA: Harvard University Press).
Shih, S. (1996) Me-Too is Not My Style: Challenge Difficulties, Break through Bottlenecks, Create Values (Taipei: The Acer Foundation).
Shin, N., Kraemer, K. L. and Dedrick, J. (2009) R&D, value chain location and firm performance in the global electronics industry, Industry
and Innovation, 16(3), pp. 315– 330.
Siegel, S. and Castellan, N. J. (1988) Nonparametric Statistics for the Behavioral Sciences (New York: McGraw-Hill).
Smakman, F. (2003) Local industry in global networks: changing competitiveness, corporate strategies and pathways of development in
Singapore and Malaysia’s garment industry. PhD Dissertation, Utrecht University, the Netherlands.
Spencer, J. W. (2003) Firms’ knowledge-sharing strategies in the global innovation system: empirical evidence from the flat panel display
industry, Strategic Management Journal, 24(3), pp. 217– 233.
Sturgeon, T. (2003) What really goes on in Silicon Valley? Spatial clustering dispersal in modular production networks, Journal of
Economic Geography, 3, pp. 199– 225.
Teece, D. J. (2007) Explicating dynamic capabilities: the nature and microfoundations of (sustainable) enterprise performance, Strategic
Management Journal, 28, pp. 1319– 1350.
Teece, D. J. and Pisano, G. (1994) The dynamic capabilities of enterprises: an introduction, Industrial and Corporate Change, 3(3),
pp. 537– 556.
Teece, D. J., Pisano, G. and Shuen, A. (1997) Dynamic capabilities and strategic management, Strategic Management Journal, 18,
pp. 509– 533.
The Yearbook of World Electronics Data. Vol. 2: America, Japan & Asia Pacific (2003).
Value Capture in the Global Electronics Industry 107
Downloaded by [Namchul Shin] at 05:10 20 February 2012
