TSMC dynamic capabilities by period

In order to examine TSMC’s dynamic capabilities, the period from 1998 to 2009 has been divided into three sections. The division between periods corresponds with a relatively low point in TSMC’s profitability as measured by its ROA (fig. 1) and ROE (fig. 2). These low points, 2001 and 2005, both coincided with global economic shocks, environmental disasters, and political upheaval. As the definition of dynamic capabilities states that dynamically capable firms have the ability to systematically solve problems, TSMC’s response to the events that caused the downturns in their profitability will be examined in terms of their ability to sense opportunities and threats in these events, to make timely and market-oriented decisions, and to change its resource base in order to restore its level of profitability.

32

Table 5 gives a brief summary of TSMC’s use of its dynamic capabilities for each of the three periods (P1, P2, and P3).

Table 5: Dynamic Capabilities by Period TSMC

Propensity to P1: 1998-2001 P2: 2002-2005 P3: 2006-2009 Sense Sense Threats -IC market downturn

and decline in orders

-4 phase plan for Philips exit

Period 1 ranges from 1998 to the end of 2001, ROA and ROE increased for two years and then decreased at the end of the period. ROA went from 12.36% in 1998 to 19.09% in 2000, and then dropped to 4.31% in 2001. ROE rose from 18.25% in 1998 to 24.87% in 2000, and then dropped to 5.22% in 2001. As was the case for SEC, these changes followed the general trend of, but were significantly higher than, the industry average. The fact that SEC’s profitability remained higher than the industry average is an indication that it was effective in dealing with the environmental shocks that affected the entire industry.

Sensing opportunities: IC industry growth and increased orders.

In 1999, TSMC sensed that, “[a]fter a slow period in the IC industry, customer orders took a strong upward turn in the second quarter of 1999” (TSMC, 1999). Similarly in 2000

33

TSMC anticipated a “steady long-term growth of the IC industry” and a “strong immediate demand from key customers (TSMC, 2000).

Making timely and market-oriented decisions: IC industry growth and increased orders.

The decision to create a new type of semiconductor firm, the “pure-play” foundry was based on the idea of fulfilling the demand for semiconductor chips without competing with customers. As part of this creation of a new niche in the industry, TSMC decided to embark upon its “Virtual Fab” strategy “to provide customers with the benefits of an in-house fabrication plant without the associated expense or organizational complexities” (TSMC, 1998). By deciding to serve as a link between upstream chip designers and downstream users, TSMC has managed to build networks of customers and suppliers (C. W. Lee, Hayter, &

Edgington, 2008), and its success has become “a model for many new entrants to the market”

(TSMC, 1999). When TSMC sensed the opportunity of increased customer orders in 1999 and 2000, it made the decision to increase capacity to meet the growing demand (TSMC, 1999, 2000).

Changing resource base: IC industry growth and increased orders.

TSMC used these resources to create a new niche in the industry, its Virtual Fab, “to provide customers with the benefits of an in-house fabrication plant without the associated expense or organizational complexities” (TSMC, 1998).

After TSMC made the decision to increase capacity to meet the growing demand in 1999 and 2000, it did so by increasing the capacity from its existing fabs 4, 5, and 6, as well as adding capacity from affiliates, new fab construction, mergers, and joint ventures (TSMC, 1999, 2000). Increased capacity from its affiliates came from WaferTech, Vanguard

International Semiconductor Corporation, and TSMC-Acer Semiconductor Manufacturing

34

Corporation (TASMC) and mergers and joint ventures included Philips in Singapore, TASMC, and Worldwide Semiconductor Manufacturing Corporation (TSMC, 1999). They also merged with TSMC-ACER (TSMC, 2000). New fab capacity was accomplished when TSMC “expanded advanced manufacturing capacity aggressively at nine internal or affiliated 8-inch fabs” and “continued the construction of two of the industry's first production scale 12-inch facilities” (TSMC, 2000).

Sensing threats: IC market downturn and decline in orders.

In 2001 TSMC sensed a threat from the market in the form of an industry downturn and a decline in orders (TSMC, 2001). This took the form of a 30% contraction in the global IC market (TSMC, 2001).

Making timely and market-oriented decisions: IC market downturn and decline in orders.

When in 2001 TSMC sensed a threat from the market in the form of an industry downturn and a decline in orders, the decision was made to slow the pace of capital investment (TSMC, 2001).

Changing resource base: IC market downturn and decline in orders.

When in 2001 TSMC made the decision was made to slow the pace of capital investment in the face of an IC market downturn and decline in orders, it did so by a

significant amount. As a result, “capital expense for 2001 totaled US$2.2 billion, a decrease of 40 percent from previous capital expense in 2000” (TSMC, 2001). After this decrease, TSMC stated that most of the remaining capital expense went to “increasing the capacity of our 0.18um, 0.15um and 0.13um processes” (TSMC, 2001).

35

5.2.2.2 Period 2: 2002-2005.

Period 2 corresponds to the years from the beginning of 2002 to the end of 2005.

During this time TSMC’s ROA rose steadily from the low point of 4.31% in 2001 to 18.93%

in 2004, then fell off slightly in 2005 to 18.44%. TSMC’s ROE followed a similar path, rising steadily from 5.22% in 2001 to 23.14% in 2004, and then falling off slightly to 21% in 2005.

Unlike SEC, TSMC’s measures did not drop in 2003, but instead followed the general curve of the industry average for the first three years, and only fell of slightly in the last year.

Throughout the period, TSMC’s ROA and ROE exceeded the industry average.

Sensing opportunities: China market.

Another opportunity that TSMC has sensed is the opening of the semiconductor industry in China. As early as 2001 TSMC saw the potential for expansion in this new and fast-growing market market. In order to achieve a competitive position in this, TSMC took steps to take advantage of this opportunity (TSMC, 2001).

Making timely and market-oriented decisions: China market.

TSMC decided that it would “become to become a key participant in the Mainland China domestic semiconductor market” (TSMC, 2002, 2003).TSMC made the decision to begin a series of steps, that eventually led to the establishment of TSMC (Shanghai) Company Limited in China (TSMC, 2003).

Changing resource base: China market.

When TSMC decide to enter the emerging semiconductor market in China, it began by allocating resources to set up an office in Shanghai (TSMC, 2001). It followed up on this in 2002 by submitting an application to the Taiwan government to invest in a semiconductor fabrication plant in China (TSMC, 2002), and then in 2003 it established TSMC (Shanghai)

36

Company Limited in China (TSMC, 2003). It equipped the plant with used tools and machinery relocated from TSMC's Taiwan fabs. As of December 31, 2004, TSMC had invested a total of US$276 million in TSMC (Shanghai) (TSMC, 2004).

Sensing opportunities: near-future demand and foundry growth.

In 2003 and 2004, TSMC also sensed strong near-future demand and foundry growth, and in 2005 estimated that up to 30% of global semiconductor revenue would come from dedicated foundries (TSMC, 2003, 2004, 2005).

Making timely and market-oriented decisions: near-future demand and foundry growth.

In 2003, sensing growing near-future demand, TSMC further decided to accelerate its capacity expansion (TSMC, 2003), and continued to do so in 2004 when it saw the potential for future foundry growth (TSMC, 2004), and again when it predicted that up to 30% of global semiconductor revenue would come from dedicated foundries in 2005 (TSMC, 2005).

Changing resource base: near-future demand and foundry growth.

TSMC followed up its decision to expand capacity by ramping up fab 12, and by bringing fabs 10 and 14 into production by 4th quarter of 2004 (TSMC, 2003), as well as expanding capacity at fab 14 (TSMC, 2004). To achieve this capacity expansion made a capital investment of approximately US $2.6-2.8 billion (TSMC, 2005).

Sensing threats: excess capacity and falling prices.

In 2004 and 2005, TSMC sensed the twin threats of excess capacity and falling prices (TSMC, 2004, 2005),

37

Making timely and market-oriented decisions: excess capacity and falling prices.

Later, in 2004 and 2005, when it sensed the twin threats of excess capacity and falling prices (TSMC, 2004, 2005), TSMC decided to focus “on creating values to better serve customers’ requirements” (TSMC, 2005).

Changing resource base: excess capacity and falling prices.

In order to better serve customers’ requirements TSMC decided to “continue to focus on high-growth segments” (TSMC, 2004).

5.2.2.3 Period 3: 2006-2009.

Period 3 corresponds to the years from the beginning of 2006 to the end of 2009. As is the case with SEC, this period shows a significantly different trend than the previous periods.

Rather than showing a gradual climb with a drop (large in 2001, small in 2005), after the first year TSMC’s ROA and ROE both show a gradual decline. In the first year, TSMC’s ROA went from 18.44% in 2005 to 22.14% in 2006, then dropped steadily to 15.45% in 2009.

ROE followed a similar pattern, rising from 21% in 2005 to 25% in 2006, then steadily falling to 18.02% in 2009. Unlike those of either SEC or the industry average, TSMC’s measures did not experience a recovery between 2008 and 2009.

Sensing opportunities: consolidate holdings.

Despite its early dependence on the transfer of technology from outside sources, TSMC eventually met and even surpassed its partners in developing its wafer fabrication technology. It eventually sensed that it could operate more efficiently by consolidating its holdings by buying out Philips (the original partner MNC) and increasing its holdings in Vanguard International Semiconductor Corporation (TSMC, 2007, 2008).

38

Making timely and market-oriented decisions: consolidate holdings.

Later, as TSMC’s technological abilities began to surpass that of its partners and it began to sense that it could operate more efficiently by consolidating its holdings, the decision was made to buy out Philips, as well as to increase its holdings in another partner, Vanguard International Semiconductor Corporation (TSMC, 2007, 2008).

Changing resource base: consolidate holdings.

When the decision was made to buy out Philips in 2007, TSMC initiated a four-phase plan for Philips exit, which was culminated in 2008 with the repurchase of “a total of 495,549 thousand common shares in the open market of the Taiwan Stock Exchange, accounting for approximately 1.92% of its total outstanding shares, at an average price of NT$61.40 per share. The repurchased shares were cancelled subsequently” (TSMC, 2007, 2008). TSMC also decided to increase its holdings in another partner, Vanguard International

Semiconductor Corporation (TSMC, 2007, 2008) to 37%.

Sensing opportunities: global economy.

After the global economic downturn that began in 2007 and continued through the first half of 2009, TSMC sensed that “the global economy is on its gradual recovery course and the outlook for semiconductor industry in 2010 appears robust” (TSMC, 2009).

Making timely and market-oriented decisions: global economy.

Even after the recent global economic slowdown in 2008 and 2009, TSMC responded quickly to the resurgence in demand by again “ramping up production capacity and capturing the pursuant recovery” (TSMC, 2009). After sensing the opportunity for renewed demand, TSMC quickly made the decision to seize “opportunity by expanding capacity early to meet subsequent urgent demand from customers” (TSMC, 2009).

39

Changing resource base: global economy.

In its 2009 Annual Report, TSMC announced plans to continue capacity expansion to capture the resurgence in demand after the slowdown of late 2008 and early 2009. “Total monthly capacity of the Company’s 12-inch wafer fabs was increased from 154,300 wafers in December 31, 2008 to 171,400 wafers in December 31, 2009. Overall, TSMC increased its annual production capacity by approximately 0.6 million 8-inch equivalent wafers in 2009”

(TSMC, 2009).

Sensing threats: IC market growth slowing.

In 2006 TSMC sensed that the growth of the overall IC market was slowing (TSMC, 2006).

Making timely and market-oriented decisions: IC market growth slowing.

In 2006, after it sensed that the growth of the overall IC market was slowing, TSMC decided to “expand into new CMOS logic IC product markets” in order to sustain growth (TSMC, 2006).

Changing resource base: IC market growth slowing.

In 2006, after it sensed that the growth of the overall IC market was slowing, TSMC decide to “expand into new CMOS logic IC product markets” by proving an “increasingly broad portfolio of CMOS logic and derivative technologies to address memory, analog, high performance logic or image sensor applications,” as well as continuing its “efforts in

strengthening ability to create a much deeper and broader relationship with each customer”

(TSMC, 2006). In 2007, TSMC announced that it “expanded its R&D in mainstream and derivative technologies, advanced CMOS and system-on-chip (SoC). R&D expenditure

40

reached NT$17.9 billion, while R&D staff grew by 14.5% during the same period” (TSMC, 2007).

5.3 Conclusion of Identification of Dynamic Capabilities in SEC and TSMC

Through the three periods studied from 1998 to 2009, the profitability of SEC and TSMC, measured by ROA and ROE, remained well above that of the industry average. The two firms showed a strong propensity to sense threats and opportunities in the environment and markets, and then to make timely and market-oriented decisions to address those threats and opportunities. They also showed a strong propensity to change their resource bases based on those decisions. Although their profitability measures went through periods of downturn, in the first two periods they showed the tendency to recover from those downturns after only one year. In the last period, however, the downturns, though generally less severe than previous ones, appear to be taking longer to recover from. Although both firms remain well ahead of the industry average in terms of profitability, this slowing of the rate of recovery may have implications towards their abilities to maintain a competitive advantage in their industry.

VI. Analysis of Financial Data: Samsung Electronics Corporation and TSMC

The case studies of Samsung Electronics Company (SEC) and TSMC demonstrate that they both possess dynamic capabilities according to the definition adopted by this study:

“A dynamic capability is the firm’s potential to systematically solve problems, formed by its propensity to sense opportunities and threats, to make timely and market-oriented decisions, and to change its resource base” (Barreto, 2010). Having demonstrated that the firms have dynamic capabilities, however, leaves the question as to how to measure whether these capabilities lead to success. It is Wu’s contention that “dynamic capabilities influence competitiveness, and subsequently financial performance” (Wu & Wang, 2007). Based on

41

this idea, the next section of this study will examine the financial performance of both SEC and TSMC to determine if their dynamic capabilities have translated into financial

performance.

6.1 Financial Performance of the Firms Relative to National Economic Trends 6.1.1 SEC sales growth rate v Korean GDP growth rate.

To determine the financial performance of SEC a comparison of its annual sales growth was made with Korea’s GDP growth rate from 1997 to 2009 (fig. 4). The figure clearly shows that SEC has averaged a sales growth rate surpassing the rate of growth of the Korean economy at large. Since a nation’s GDP is the “total expenditure on the economy’s output of goods and services” (Mankiw, 2002), its rate of increase can be thought of as an overall average of the growth rates of all the firms in the economy. If a firm’s growth on average is higher than that of the national GDP, then it can be seen as doing better than average for firms in that country.

In each year except for two, SEC’s sales growth has outpace that of the national GDP, and often by a substantial amount (the mean growth rate for 1997-2009 was almost 17% for SEC sales and less than 5% for the Korean GDP). The years in which SEC’s sales growth rate slowed, or lagged behind the GDP growth rate, were years in which there were

significant outside influences at work. In 2001, for example, SEC’s sales growth rate (6.7%) was only slightly higher than the GDP growth rate (4.81%). However, this was the year after the dot com bubble burst, as well as a year in which there was significant international political unrest. As Jong-Yong Yun, Vice Chairman and CEO of SEC, stated in SEC’s 2001 Annual Report, “Nothing could have prepared us for the challenges that the year would bring, including a devastating crash in memory prices, a stagnant IT sector, and a global economic chill that turned even colder in the wake of the terrorist attacks in the U.S.” (SEC, 2001).

42

SEC responded by “Rapid restructuring and innovations in the way we do business” (SEC, 1998). This restructuring was dubbed “select and focus” and included the decisions to drastically cut personnel and exit from or sell off business units (SEC, 1998). Although SEC’s sales growth was slow that year, it is further evidence of their ability to respond quickly and effectively to threats that by the next year they posted a 28 percent sales growth, the second highest in more than ten years.

In 2003, SECs sales growth rate again slipped, but as was the case in 2001, this occurred against the backdrop of international concerns about terrorism, and the US invasion of Iraq. Jong-Yong Yun pointed out that 2003 was an “unfavorable business environment”

and that “uncertainty prevailed” (TSMC, 2003). In addition, Yun pointed out that they

“emphasize profits rather than sales growth” (TSMC, 2003), and indeed in 2003 SEC posted an almost 14% profit margin that year.

The lowest sales growth rate that SEC experienced between 1997 and 2009 was in 2005. This was a year that SEC’s Vice Chairman and CEO Yun referred to as “eventful,” in which “[m]any parts of the world struggled with natural disasters including earthquakes and hurricanes” and in which “[t]he global economy was hit with record high oil prices of over USD 60 per barrel” (SEC, 2005). He also cited “increased competition created predatory pricing in global markets” and “a sharp surge in raw material prices and the appreciation of the Korean won against major global currencies” in the domestic market as challenges (SEC, 2005). Nevertheless, SEC “posted annual average operating profits for the last four years of KRW 8 trillion,” (SEC, 2005) which indicates that while sales growth may have been slow, the overall financial performance of the firm was sound.

After that SEC’s sales growth rate made a continuous climb, matching GDP growth rate in 2006, and substantially surpassing it in the next three years.

43

Growth Rates 1998-2009: Korean GDP v SEC Sales

Korea GDP Growth -7.28 8.81% 7.28% 4.81% 8.25% 4.58% 7.51% 4.88% 6.04% 7.25% 4.95% 3.25%

SEC Sales Growth 13.62 24.51 35.65 6.70% 28.26 8.81% 26.45 -1.63 5.95% 15.31 23% 14.59 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009

Figure 4: Growth rates of SEC sales and Korean GDP 1998-2009.

Source: Penn World Table 6.3 (Heston, 2009), SEC Annual Reports.

Although sales growth rate is just one measure of a firm’s financial performance, combined with other measures—which will be discussed in following sections—SEC’s consistent performance above the Korean GDP in most years examined, as well as its rapid recovery from years in which growth was slow, gives evidence as to its effective use of its dynamic capabilities.

6.1.2 TSMC sales growth rate v Taiwanese GDP growth rate.

Like SEC, TSMC has generally enjoyed a rate of sales growth that has surpassed the growth of the Taiwanese GDP, which can be seen in figure 5 (between 1999 and 2009, TSMC had a mean sales growth rate of almost 22%, while Taiwan’s GDP averaged about 5%

growth). In 1999, TSMC’s sales grew more than 45% from the previous year, and in 2000 they skyrocketed more than 127%. This was due to an overall surge in the global

44

semiconductor industry, but many of TSMC’s moves to increase capacity and form partnerships and mergers (discussed in previous sections) also contributed to the firm’s performance.

The following year, however, brought negative sales growth, but as in the case of SEC, much of this can be attributed to the industry-wide contraction of over 30%. TSMC’s sales growth for the first half of the year was very slow, but recovered partially in the second

The following year, however, brought negative sales growth, but as in the case of SEC, much of this can be attributed to the industry-wide contraction of over 30%. TSMC’s sales growth for the first half of the year was very slow, but recovered partially in the second