case analysis

S w

906M23 SUN MICROSYSTEMS Scott Jacobs and Prescott C. Ensign prepared this case solely to provide material for class discussion. The authors do not intend to illustrate either effective or ineffective handling of a managerial situation. The authors may have disguised certain names and other identifying information to protect confidentiality. Ivey Management Services prohibits any form of reproduction, storage or transmittal without its written permission. This material is not covered under authorization from CanCopy or any reproduction rights organization. To order copies or request permission to reproduce materials, contact Ivey Publishing, Ivey Management Services, c/o Richard Ivey School of Business, The University of Western Ontario, London, Ontario, Canada, N6A 3K7; phone (519) 661-3208; fax (519) 661-3882; e-mail [email protected]. Copyright © 2006, Ivey Management Services Version: 2009-01-12 PRESS RELEASE The fourth quarter ended June 30, 2005. The 2005 fiscal year was over. On July 26, 2005, Sun Microsystems (Sun) unveiled fiscal year (FY) 2005 results: Q4 revenues were US$2.98 billion (2004 Q4 revenues were US$3.11 billion, but had included income of US$1.6 billion from a legal settlement with Microsoft); for the full fiscal year 2005, Sun Microsystems reported revenues of US$11.07 billion (FY2004 revenues were US$11.19 billion). Sun Microsystems Chief Financial Officer (CFO), Steve McGowan commented, “We achieved impressive operational improvements in fiscal 2005 . . . our 16th consecutive year of generating positive cash flow from operations.” Sun Microsystems employees, investors and analysts recognized that Bill Gates had helped out last year. But what explained this year? Perhaps Sun was righting itself? “Putting our cash to work, we’ve expanded our product portfolio and announced plans to acquire companies that deepen and broaden our systems strategy. We’ve maintained our R&D commitment and delivered crown jewels like Solaris 10 to the market,” said Scott McNealy, chairman and chief executive officer (CEO) of Sun Microsystems. “Big-time progress in FY05. The company is now in a position to take advantage of the investments we have made over the past few years and we believe there is more to come in FY06.”

For the exclusive use of R. Wang, 2020.

This document is authorized for use only by Rachel Wang in 2020.

Page 2 9B06M023

McNealy continued, “Our demand indicators for Q4 were positive. We have great partners, lots of cash, and a strong team across the board. FY05 was a year of stabilized revenue and earnings. Our opportunity for FY06 is sustained growth and profitability.” “Profitability?” was the incredulous reaction of many who had followed Sun’s struggle. A STRATEGIC CRISIS At the beginning of 2004, Sun Microsystems found itself at a serious inflection point (see Exhibits 2, 3 and 4). It had lost money in eight of the last 10 quarters and allowed its market share to slip again, this time from 12.1 per cent in 2002 to 10.3 per cent in 2003, Sun announced its largest product offering update in company history. The products were intended to stop the bleeding that had caused the company’s share price to plummet to only one-tenth of its high value in 2000, a decline that in 2003 raised speculation of a takeover. The chief executive officer and chairman of the board of directors was Scott McNealy, co- founder of the company originally named Stanford Unified Networks when it began in 1982. Since developing the vision that had endured throughout the company’s history, “The Network is the Computer,” McNealy had come under great scrutiny as many industry analysts relentlessly questioned Sun’s business-level strategy. Specifically, many doubted the effectiveness and competitiveness of the products that the company offered during its recent decline in profitability. On December 8, 2004, Scott McNealy delivered a keynote address at an industry show in San Francisco, California. He relayed some statistics about Sun’s Java platform: 579 million Java-enabled phones by the end of the year, almost one billion java smart cards and nearly 900 companies were now contributing to Java. He spoke of how the world had envisioned the computer 50 years prior and remarked, “It’s hard to imagine where we’ll be 50 years from now.” Some in the audience were scratching their heads wondering where Sun was headed in the very near future. Fiscal year 2005 had just closed and it was now July. With catchy newspaper headlines, The Sun Always Rises; Dawn of a New Era for Sun; and Will Sun Shine Again?, it was evident that Sun Microsystems was at a crossroads, and it was time for serious reflection on the part of Scott McNealy. The company, based in Santa Clara, California, had been through a lot in the last 10 years. It had learned from some experiences and was still learning from others.

For the exclusive use of R. Wang, 2020.

This document is authorized for use only by Rachel Wang in 2020.

Page 3 9B06M023

BOOM TO BUST In 1997, Thailand could no longer prop up its currency, the baht. High levels of debt and years of trade deficits had resulted in a spectacular crash in the value of the baht. The ensuing economic crisis engulfed most of Asia and spread throughout the entire world, slowing down economies from east to west. Most importantly for Sun, the crisis adversely impacted Japan, its largest single source of foreign revenue, at nine per cent. The global economic collapse continued in January of 1999 when Brazil devalued its currency, the real. The devaluation dashed any immediate growth plans and delayed future growth there indefinitely. Brazil had been one of the countries on which Sun had been counting to lead information technology (IT) spending in South America. The Brazilian and Asian financial crises provided compelling evidence that, while Sun’s diverse portfolio appeared to protect it from individual blips in economic stability, its product sales were highly sensitive to macroeconomic conditions. In 1998, the company reported that more than 45 per cent of its revenues was generated outside of the United States. By 2003, Sun’s reliance on foreign revenue had increased to more than 50 per cent, and the economic crises had unfolded into a worldwide economic slowdown — with some analysts using the label “recession.” However, economic conditions alone did not lead to Sun’s fall from profitability. Rather, the economic macro-environment was a catalyst for change towards affordable enterprise computing. In addition, technological advancements led to performance improvements in the x86 platform.1 Making use of these advancements along with utilizing various sources of leverage to lower cost structures, many firms were able to offer products similar to Sun’s but at a fraction of the price. Sun’s expensive high-margin products lagged in demand while the company attempted in vain to continue presenting its same value proposition. To compound the adverse effect of developments in the x86 hardware space, competitive pressure was also being applied to Sun by an inexpensive, competitive and very contrary x86-based software product, named Linux. This singular platform would change Sun’s competitive environment drastically. Open source software on inexpensive standard hardware began to unravel Sun’s offering: a product where software, hardware and service were bundled. 1x86 hardware stands for x86 central processing units (CPUs) developed by Intel. This CPU forms the base of PCs. History of the chip: 286, 386, 486, 586 (dubbed “Pentium), Pentium Pro, Pentium II, Celeron, Pentium III, Pentium IV, etc. Other manufacturers, such as Advanced Micro Devices (AMD) and Cyrix, have used the same CPU machine code with their own architecture to create equivalent chips, but with a lower price-to-performance ratio.

For the exclusive use of R. Wang, 2020.

This document is authorized for use only by Rachel Wang in 2020.

Page 4 9B06M023

BILLION-DOLLAR BETS In 2000, shares of Sun Microsystems hovered near US$60 per share. Around that same time, Scott McNealy, CEO, put Masood Jabbar in charge of managing Sun’s worldwide sales. McNealy was counting on Jabbar to formulate the right plan and strategy to capitalize on Sun’s momentum and reputation for innovation. McNealy needed Jabbar to grow the business and build shareholder wealth. In response, Jabbar developed a strategy that focused on five countries that were each potential billion-dollar-a-year markets for Sun’s server business. They were Brazil, Spain, China, India and Italy. Shareholders were optimistic about Sun’s potential for growth. Although shareholders responded positively to Sun’s future, managers of the U.S. business at Sun’s Palo Alto, California headquarters were getting tense. It seemed as though a meeting could not take place without attention being diverted from the U.S. business toward conversations about foreign interest rates, political regimes and foreign legal language. U.S. business managers simply could not have discussions on policy and strategy without some “foreign distractions.” Many of the U.S. managers involved began to feel that those managers representing Mexico, South America and Canada were muddying the waters with talk about fluctuating currency exchange rates and political instability. In fact, those territories accounted for less than a few per cent of Sun’s business at the time. The United States had nearly always accounted for roughly half of Sun’s multibillion-dollar annual sales. However, Jabbar’s “billion-dollar bets” required that these emerging areas get attention in order to shift the distribution of sales. And yet, international managers felt they were getting little cooperation and support to expand business in their emerging markets. Jabbar also knew that under the current structure, U.S. managers could not focus on their own strategy and planning, which meant that a considerable portion of Sun’s revenue could be threatened. The current organizational structure was just not working. Something at Sun had to change. Jabbar approached Bob MacRitchie, who at that time was already managing the South American, Mexican and Canadian lines of business under the umbrella of the U.S. organizational division. The infrastructure was already in place. What remained was an official transition that emphasized Sun’s commitment to pursuing its billion-dollar bets. “I just got on the phone with Bob and told him what we were doing,” Jabbar later said, “It took only a minute.” However, the transition itself was not so easy. The same U.S. business managers that previously decried the waste of precious resources on those underperforming emerging markets felt a sense of loss they had not anticipated — a loss of power.

For the exclusive use of R. Wang, 2020.

This document is authorized for use only by Rachel Wang in 2020.

Page 5 9B06M023

“Egos were involved here,” Jabbar explained, “It was hard to get those guys to give up the control.” Apart from being a functional decision, the decision was symbolic of the attention and commitment Sun would give to these emerging areas. Sun’s strategy was to develop these “billion-dollar bets.” The economy was booming and the demand for enterprise computing products was growing quickly. Sun did not have to chase customers. The sales, it seemed, were coming to them. However, neither Sun nor any other company could contend with the competitive pressure created by the volatile nature of the new global economy. ATTACK OF THE GIANT PENGUIN — LINUX “We’re saying that Linux will eat Unix,”2 was the prediction for 2003, issued by John Gantz, International Data Corporation’s (IDC’s) chief research officer. The irony was subtle — the Linux mascot was an innocuous penguin. Linux, an operating system named after its creator Linus Torvalds, took the enterprise computing industry by storm. Its open source, community approach to application creation was not only revolutionary in theory, but in practice as well. Linux is open source, inexpensive and scalable. Adoption of Linux products spread rapidly, especially internationally. In fact, countries outside the United States found Linux to be a viable alternative to Microsoft’s Windows operating system, which many critics felt was unsecure and not readily scalable. In 2003, Brazil’s government urged its federal agencies to adopt Linux in an effort to cut costs. Brazil represented no small piece of the pie; it imported more than US$1 billion more in software than it exported in 2001. In moving to Linux, Brazil was not alone. It joined China, Japan and South Korea in using Linux to get IT spending under control. Sun had resisted Linux for a long time since Sun designed its own operating system, Solaris. In addition, Sun’s past success was due in large part to its strategy of integrating proprietary software (Solaris), hardware and its own ‘SPARC’ Unix microprocessor.3 By developing not only its own microprocessors, but also its own operating system, the company could have complete control of the integration of the system hardware and software. This strategy ensured optimization and control over the entire design process. In effect, Sun created a lot of value for its customers through this integration. 2Unix software is an abbreviation for UNiplexed Information and Computing System, originally spelled “Unics.” It is an interactive time-sharing operating system developed at Bell Labs in 1969. In the 1990s, Unix was the most widely used multi-use general-purpose operating system in the world. Unix is presently offered by many manufacturers and is the subject of international standardization efforts. 3SPARC is an abbreviation for Scalable Processor ARChitecture, designed by Sun Microsystems in 1985. SPARC is not a chip per se, but a specification. The first standard product based on SPARC was produced by Sun and Fujitsu in 1986. In 1989, Sun transferred ownership of the SPARC specifications to an independent, non-profit organization, SPARC International.

For the exclusive use of R. Wang, 2020.

This document is authorized for use only by Rachel Wang in 2020.

Page 6 9B06M023

However, Sun’s strategic focus on its operating system, Solaris, and its SPARC microprocessors prevented it from seeing the threat that Linux presented. Eventually, an x86 or Unix server running Linux offered customers a close substitute to Sun’s products. In February of 2002, after years of denial, Sun Microsystems could no longer fend off the waves of criticism — it gave in to Linux. To demonstrate its commitment to Linux, Scott McNealy gave his keynote speech (in a penguin suit) at a Sun trade show to announce the company’s new Linux products. Typically, Sun attacked other platforms by suggesting that they were not as reliable or did not offer the features of its own Unix servers. However, with the advent of Linux, users found a new, reliable, inexpensive choice to expensive Sun products. The problem was serious. If Sun refused to offer Linux products, it would be cutting itself out of a considerable portion of the server market. Linux was dominating the inexpensive x86 server space. Yet, if Sun were to produce its own x86 products, it would have to purchase microprocessors from a third party for a new line of x86 hardware. Since there was essentially one company that was developing highly respected enterprise-level x86 microprocessors, Sun had few choices. Entering the x86 market would mean teaming up with a tried and true competitor — Intel. For Sun, doing business with Intel would not only mean accepting the rise of Linux, but also the commoditization of enterprise computing, at least at the low end. Moreover, Sun’s value proposition became increasingly diminished as the company lost control of the integration process. By using Intel chips, Sun would no longer offer value to customers through engineering and innovation as it had previously, but rather via an assembly of outsourced components. Although very similar to some of its competitors, that value proposition was a substantial departure from Sun’s original model. To counteract this change, Sun developed its own middleware, the Java Enterprise System, that would operate in Unix or Linux and offered a unique pricing model that made scalable enterprise computing more affordable. In 2003, the People’s Republic of China and Sun announced a deal. The arrangement was for 500,000 to one million copies per year of Sun’s Java Enterprise System. Most industry observers believed the blockbuster result of this deal would not be the revenue, but rather the market share. Indeed the revenue would likely not be significant and would even take quite a while to capture, given the extended time frame of the agreement. With Java Enterprise System, Sun was trying to inject value into its offerings through research and development, one of Sun’s competitive advantages. Continuing to partner with Intel, however, remained a strategic disadvantage. After years of competition, joining forces and integrating development initiatives

For the exclusive use of R. Wang, 2020.

This document is authorized for use only by Rachel Wang in 2020.

Page 7 9B06M023

with Intel seemed unrealistic. Sun’s future in the x86 space continued to look uncertain until November of 2003, when Sun announced a newly formed strategic alliance with an emerging x86 microprocessor developer — AMD. A NEW ENTRANT, OLD COMPETITORS AND BIG THREATS In 2003, Advanced Micro Devices, also known as AMD, agreed to collaborate with Sun in a new strategic venture. By providing its x86-based Opteron chips for use in Sun’s low-end servers, AMD stood to combat years of dominance by rival, Intel. AMD’s new Opteron technology took its x86 technology to 64 bits,4 a feat Intel had yet to master. In doing so, it became an industry leader. Along with that status came new friends. Sun was not the only firm to form an alliance with AMD — HP and IBM both struck similar deals, although seemingly less integrated than Sun’s alliance. Intel leveraged its technology, manufacturing capabilities and market share in the PC space to produce chips able to compete in the enterprise computing domain, at least in the low-end segment. And, it certainly did not suffer from diminishing relations with Sun. Dell capitalized in the low-end space by selling a large volume of servers running Linux on Intel. Finding operating efficiencies allowed Dell, the fourth largest server maker behind Sun, to turn the low-end server market into a battlefield by steadily dropping prices.5 Sun attempted to compete on the basis of price, but found it difficult to out-price companies like Dell and even IBM that reduced costs through sources of leverage, such as unit volume and consulting services respectively. As a result, Sun was losing the low-end battle against Dell, IBM and HP, and there continued to be little demand in the high-end for its expensive Unix servers. As if this were not bad enough, another competitor was posing a significant threat to Sun. Japan’s Fujitsu, which ranked as the fifth largest server maker, licensed Sun’s proprietary UltraSPARC chip designs to power its own SPARC64 chips found in its line of Unix servers. Fujitsu had great success in taking Sun’s SPARC chips and turning them into more powerful SPARC64 Fujitsu chips — and winning Sun customers in the process. Fujitsu was much quicker to market than Sun, which was reliant on Texas Instruments to churn out its chips. The situation presented a very difficult decision for Sun. Should it discontinue its relationship with Fujitsu or get in even tighter with the Japanese giant? Sun Microsystems desperately needed the revenue from the licensing agreement, so it was extremely reluctant to cease the licensing contracts. To complicate 464 bit is a computer architecture term, which refers to the bandwidth of the arithmetic logic unit, registers (high-speed memory locations in the CPU) and data bus (connections between and within the CPU, memory and peripherals). 5HP was the server market leader, followed by IBM.

For the exclusive use of R. Wang, 2020.

This document is authorized for use only by Rachel Wang in 2020.

Page 8 9B06M023

matters further, Fujitsu was not only licensing Sun’s chipset designs, it was also a leading reseller of Sun product. Fujitsu was more than a manufacturer; it served also as a reseller or channel partner. In fact, Fujitsu’s role was significant. Sales to Fujitsu accounted for a large portion of Sun sales in Japan, Sun’s leading foreign market in revenue terms. Rumors surfaced in 2003 about a possible alliance or even merger with the Japanese firm. However, for Sun, forming a tighter alliance with Fujitsu might cause more problems than it would address. For instance, if Sun were to begin using the SPARC64 chips, it might mean spurning its longstanding chip- producing partner, Texas Instruments. Given a very large installed base, such bridges were not to be burned hastily. Moreover, despite the fact that Sun and Fujitsu used the same instruction set for the microprocessors, Sun and Fujitsu products could not be used together without some reprogramming. Adopting the SPARC64 chips would mean massive reconstruction of its current products and the products it had in the pipeline, which suggested that Fujitsu would remain a competitor in the future. EXECUTIVE EXODUS A month-long executive exodus culminated with the Chief Operating Officer (COO) and President Ed Zander resigning. He and four other high-level officers — all with at least 15 years each at Sun — departed effective July 1, 2002. The naysayers on Wall Street were having a field day — “the Captain goes down with the ship.” McNealy was going to do it alone — he named no successor for the president and COO’s position. There were also now vacancies for the vice- president, two executive VPs and the CFO. On July 18, 2002, it was announced that Masood Jabbar would retire from Sun after 16 years of service. The stock was at a 52-week low, hovering around US$7 per share. And things did get worse — at least according to the stock market. The share price slowly but steadily, in the coming years, fell to a mark below US$3. It would be almost two years before a president and COO was named — the 38- year-old, pony-tailed Jonathan Schwartz received the nod from McNealy in April 2004. Schwartz had come to Sun in 1996 when it acquired Lighthouse Design, Ltd. — the firm Schwartz was then CEO of. ALL BETS ARE OFF Growth in the server industry was not flat by any means. IDC expected tech spending would increase in 2005 and beyond, regaining some of its lost momentum from the bubble burst of 2000.

For the exclusive use of R. Wang, 2020.

This document is authorized for use only by Rachel Wang in 2020.

Page 9 9B06M023

Scott McNealy’s vision of everything and everyone connected to the network was probably not that far from the eventual truth (see Exhibit 1). Yet, Sun’s ability to determine the long-term direction of computing was more refined than its value- added strategy. Most of Sun’s billion-dollar bets that were developed by Masood Jabbar failed to materialize. In fact, Sun Microsystems was now desperate, fighting for revenue each quarter. In retrospect, Sun might have put the cart before the horse. Instead of focusing on the evolution of its competitive environment, it was distracted by the economic euphoria that reigned in the late 1990s. McNealy admitted to as much, “Looking back, we probably hired too many people and signed too many leases, but we had a very natural and understandable desire to fill all the orders we could.” The network was still the computer — even increasingly so — but recent product offerings and industry conferences suggested Sun Microsystems and especially McNealy had succumbed to some inevitable truths about the competitive environment of the industry. One was that the x86 platform was not going away as McNealy had once envisioned. It became clear that Sun’s future strategy would include x86- and Linux-based products, as well as more compelling and competitive high-end Unix servers. To sell Sun product, McNealy was banking on a healthy return on research and development. He asserted, “Sun is doing things that Intel isn’t doing or AMD isn’t doing.” Innovation, however, was costly. Sun’s R&D budget had been around US$2 billion annually for the last few years, dwarfing its competitors. Innovation was a key component of Sun’s current strategy, but the results were not yet readily quantifiable. Still, McNealy claimed, “[Over the decade,] we’re going to spend $20 billion to $30 billion minimum on R&D.” Clearly, Sun was intent on continuing to add value through innovation rather than being relegated to a game of economies of scale and efficiencies. Charles Cooper, editor of CNETNEWS.com, remarked, “[Sun’s] strategy — if you can call it that — has been to throw a lot of stuff against the wall and wait to see what sticks.” It was still hard to tell what the firm and specifically Scott McNealy were learning from the devolution of Sun’s business. The well-known CEO admitted to strategic missteps but quipped “I try to make a mistake only once.” “If your strategy isn’t controversial, you have zero chance of making money.” “You have to have a wildly different strategy and you have to be right. It’s that second part that gets tricky.” McNealy recognized, “There’s lots more to do.” Now was the time to do it.

For the exclusive use of R. Wang, 2020.

This document is authorized for use only by Rachel Wang in 2020.

Page 10 9B06M023

Exhibit 1

SUN MICROSYSTEMS’ STRATEGIC DIRECTION A singular vision -- “The Network Is The Computer” -- guides Sun in the development of technologies that power the world’s most important markets. Sun’s philosophy of sharing innovation and building communities is at the forefront of the next wave of computing: the Participation Age. VISION: Everyone and everything connected to the network. Eventually every man, woman, and child on the planet will be connected to the network. So will virtually everything with a digital or electrical heartbeat -- from mobile phones to automobiles, thermostats to razor blades with RFID tags on them. The resulting network traffic will require highly scalable, reliable systems, from Sun. MISSION: Solve complex network computing problems for governments, enterprises, and service providers. At Sun, we’re tackling complexity through system design. Through virtualization and automation. Through open standards and platform-independent Java technologies. In fact, we’re taking a holistic approach to network computing in which new systems, software, and services are all released on a regular, quarterly basis. All of it integrated and pretested to create what we call the Network Computer.

Source: http://www.sun.com, accessed February 11, 2006.

For the exclusive use of R. Wang, 2020.

This document is authorized for use only by Rachel Wang in 2020.

Page 11 9B06M023

Exhibit 2

INCOME STATEMENT (US$ millions)

1996 1997 1998 1999 2000 2001 2002 2003 2004 2005

Revenue 7,094.8 8,598.4 9,790.8 11,726.3 15,721.0 18,250.0 12,496.0 11,434.0 11,185.0 11,070.0 COGS 3,972.0 4,320.5 4,693.3 5,648.4 7,549.0 10,041.0 7,580.0 6,492.0 6,669.0 6,481.0 Gross Profit 3,122.7 4,277.9 5,097.5 6,077.9 8,172.0 8,209.0 4,916.0 4,942.0 4,516.0 4,589.0

Operating Expenses Selling, General and Administrative 1,732.7 2,402.4 2,777.3 3,173.0 4,137.0 4,544.0 3,812.0 3,329.0 3,317.0 2,919.0 Research and Development 657.1 826.0 1,190.2 1,383.2 1,642.0 2,093.0 1,835.0 1,841.0 1,996.0 1,785.0 Other 57.9 23.0 - - - 261.0 517.0 2,496.0 393.0 262.0 Operating Income 675.0 1,026.5 1,130.1 1,521.8 2,393.0 1,311.0 (1,248.0) (2,724.0) (1,190.0) (377.0)

Other Income and Expenses Net Interest Income and Other 33.9 94.7 46.1 83.9 378.0 273.0 200.0 71.0 1,627.0 193.0 Earnings Before Taxes 708.9 1,121.2 1,176.2 1,605.7 2,771.0 1,584.0 (1,048.0) (2,653.0) 437.0 (184.0) Income Taxes 232.5 358.8 413.3 574.4 917.0 603.0 (461.0) 776.0 825.0 (77.0) Earnings After Taxes 476.4 762.4 762.9 1,031.3 1,854.0 981.0 (587.0) (3,429.0) (388.0) (107.0) Accounting Changes (54.0) Net Income 476.4 762.4 762.9 1,031.3 1,854.0 927.0 (587.0) (3,429.0) (388.0) (107.0) Diluted EPS Continuing Ops 0.2 0.3 0.2 0.3 0.6 0.3 (0.2) (1.1) (0.1) (0.0) Diluted EPS 0.2 0.3 0.2 0.3 0.6 0.3 (0.2) (1.1) (0.1) (0.0) Shares 3,147.0 3,111.0 3,154.0 3,256.0 3,378.0 3,417.0 3,242.0 3,190.0 3,277.0 3,368.0

Source: company files

For the exclusive use of R. Wang, 2020.

This document is authorized for use only by Rachel Wang in 2020.

Page 12 9B06M023

Exhibit 3

CASH FLOWS (US$ millions)

1996 1997 1998 1999 2000 2001 2002 2003 2004 2005

Cash from Operating Activities Net Income 476.4 762.4 762.9 1,031.3 1,854.0 927.0 (587.0) (3,429.0) (388.0) (107.0) Depreciation/Amortization 284.1 341.7 439.9 627.0 776.0 1,229.0 1,092.0 918.0 730.0 671.0 Deferred Taxes - - - - - - (673.0) 706.0 620.0 (315.0) Other (72.2) 0.9 323.7 858.6 1,124.0 (67.0) 1,048.0 2,842.0 1,264.0 120.0 Cash from Operations 688.3 1,105.1 1,526.5 2,516.9 3,754.0 2,089.0 880.0 1,037.0 2,226.0 369.0

Cash from Investing Activities Cap Expenditures (295.6) (554.0) (830.1) (738.7) (982.0) (1,292.0) (559.0) (373.0) (249.0) (257.0) Purchase of Business - (23.0) (244.0) (130.3) (89.0) (18.0) (49.0) (30.0) (201.0) (95.0) Other 170.9 33.3 (94.8) (1,227.1) (3,154.0) (244.0) 647.0 (125.0) (1,861.0) (73.0) Cash from Investing (124.7) (543.7) (1,169.0) (2,096.1) (4,225.0) (1,554.0) 39.0 (528.0) (2,311.0) (425.0)

Cash from Financing Activities Net Issuance of Stock (467.5) (374.8) (190.8) (242.7) (285.0) (899.0) (354.0) (317.0) 239.0 218.0 Net Issuance of Debt - - - - 1,500.0 - (13.0) (201.0) (28.0) (252.0) Dividends - - - - - - - - - - Other 18.9 (55.3) (4.6) 88.7 4.0 (13.0) - - - - Cash from Financing (448.6) (430.1) (195.5) (154.0) 1,219.0 (912.0) (367.0) (518.0) 211.0 (34.0) Currency Adjustments - - - - - - - - - - Change in Cash 115.0 131.3 162.1 266.7 748.0 (377.0) 552.0 (9.0) 126.0 (90.0)

Free Cash Flow Cash from Operations 688.3 1,105.1 1,526.5 2,516.9 3,754.0 2,089.0 880.0 1,037.0 2,226.0 369.0 Capital Expenditures (295.6) (554.0) (830.1) (738.7) (982.0) (1,292.0) (559.0) (373.0) (249.0) (257.0) Free Cash Flow 392.7 551.1 696.4 1,778.2 2,772.0 797.0 321.0 664.0 1,977.0 112.0

Source: company files

For the exclusive use of R. Wang, 2020.

This document is authorized for use only by Rachel Wang in 2020.

Page 13 9B06M023

Exhibit 4

BALANCE SHEET (US$ millions)

1996 1997 1998 1999 2000 2001 2002 2003 2004 2005

Assets Cash and Equivalent 528.9 660.2 822.3 1,089.0 1,849.0 1,472.0 2,024.0 2,015.0 2,141.0 2,051.0 Short-term Investments 460.7 452.6 476.2 1,576.1 626.0 387.0 861.0 1,047.0 1,460.0 1,345.0 Account Receivable 1,206.6 1,666.5 1,845.8 2,286.9 2,690.0 2,955.0 2,745.0 2,381.0 2,339.0 2,231.0 Inventory 460.9 438.0 346.5 307.9 557.0 1,049.0 591.0 416.0 464.0 431.0 Other Current Assets 376.6 511.2 656.8 856.5 1,155.0 2,071.0 1,556.0 920.0 899.0 1,133.0

Total Current Assets 3,033.7 3,728.5 4,147.5 6,116.4 6,877.0 7,934.0 7,777.0 6,779.0 7,303.0 7,191.0 Net Property, Plant, and Equipment 533.9 799.9 1,300.6 1,608.9 2,095.0 2,697.0 2,453.0 2,267.0 1,996.0 1,769.0 Intangibles 2,041.0 2,286.0 417.0 533.0 554.0 Other Long-term Assets 233.3 168.9 262.9 695.1 5,180.0 5,509.0 4,006.0 3,522.0 4,671.0 4,676.0

Total Assets 3,800.9 4,697.3 5,711.1 8,420.4 14,152.0 18,181.0 16,522.0 12,985.0 14,503.0 14,190.0

Liabilities and Stockholders' Equity Accounts Payable 325.1 468.9 495.6 753.8 924.0 1,050.0 1,044.0 903.0 1,057.0 1,167.0 Short-term Debt 87.6 100.9 47.2 1.7 7.0 3.0 205.0 - 257.0 - Taxes Payable 134.9 118.6 188.6 402.8 422.0 90.0 Accrued Liabilities 801.6 963.0 1,126.5 1,646.6 2,117.0 1,862.0 1,739.0 1,506.0 1,930.0 1,727.0 Other Short-term Liabilities 140.2 197.6 265.0 422.1 1,289.0 2,141.0 2,069.0 1,720.0 1,869.0 1,872.0

Total Current Liabilities 1,489.3 1,849.0 2,122.9 3,227.0 4,759.0 5,146.0 5,057.0 4,129.0 5,113.0 4,766.0 Long-term Debt 60.2 106.3 1,720.0 1,705.0 1,449.0 1,531.0 1,175.0 1,123.0 Other Long-term Liabilities 74.6 381.6 364.0 744.0 215.0 834.0 1,777.0 1,627.0

Total Liabilities 1,549.4 1,955.3 2,197.4 3,608.6 6,843.0 7,595.0 6,721.0 6,494.0 8,065.0 7,516.0 Total Equity 2,251.5 2,741.9 3,513.6 4,811.8 7,309.0 10,586.0 9,801.0 6,491.0 6,438.0 6,674.0

Total Liabilities and Equity 3,800.9 4,697.3 5,711.1 8,420.4 14,152.0 18,181.0 16,522.0 12,985.0 14,503.0 14,190.0 Source: company files

For the exclusive use of R. Wang, 2020.

This document is authorized for use only by Rachel Wang in 2020.