6

                              

Arbitrage

Exploiting Differences

Globalization is about producing where it is most cost effective, sourcing capital from where it’s cheapest and selling it where it is most profitable.

—N. R. Narayana Murthy, Infosys, August 2003

THE THIRD OF OUR AAA STRATEGIES for dealing with distances and crossing borders successfully is arbitrage. Arbitrage is a way of exploiting differences. It implies seeking absolute economies, rather than the scale economies gained through standardization. It treats differences across borders as opportunities, not as constraints.

This chapter begins by underscoring the absolute importance of arbitrage. It then uses the CAGE framework to unpack the cultural, administrative, geographic, and economic bases of arbitrage. To illustrate the variety of arbitrage strategies, I use a complex case drawn from the pharmaceutical industry, where both the administrative and economic bases of arbitrage are important. The chapter concludes with further discussion of how to use the ADDING Value scorecard to analyze arbitrage, and of some of the managerial challenges that arise in exploiting arbitrage opportunities.

The Absolute Importance of Arbitrage

Arbitrage, of course, is the original cross-border strategy. Many of the great traders throughout history got their start by trading luxuries that were subject to extreme differences in absolute costs and availability. Thus, Europe’s spice trade with India developed because spices could (initially) be sold in Europe for several hundred times what they cost in India. Furs and fish that were abundantly available only in North America helped create a transatlantic trade and, incidentally, led to the colonization of that continent. Similar, essentially geographic differences drove the global whaling fleets of the late eighteenth century (which, with their floating factory ships, can be said to have originated offshore manufacturing), as well as the vertically integrated agricultural and mining companies that emerged early in the nineteenth century.

The freestanding enterprises that dominated British foreign direct investment at the end of the nineteenth century attempted to arbitrage across differences in administrative structure (and power) by pursuing foreign investment opportunities under British law. Additionally, exports of light manufactures (e.g., garments) became important in the nineteenth century. These, too, involved arbitrage, but it was an arbitrage across economic differences, rather than geographic or administrative ones.

Despite this long history, arbitrage is often glossed over in contemporary discussions of globalization and strategy. Take Wal-Mart, for example: most of the general discussion about its internationalization revolves around its international store network. Wal-Mart has more than 2,200 international stores, which together generated $63 billion in sales (one-fifth of the company total) and $3.3 billion in operating income (closer to one-sixth of the total) in 2006. What attracts significantly less attention is Wal-Mart’s global sourcing effort, particularly from China. In 2004, the company claimed to buy about $18 billion worth of goods directly from China, not counting merchandise in its stores obtained indirectly from there via suppliers. Even if one takes just the $18 billion figure and applies the usual estimates of how much that reduces Wal-Mart’s costs, the implied savings approach $3 billion—that is, they are comparable to the operating income generated by the international stores, but on a much smaller investment base.¹ And going by the results of a store check I conducted in a small sample of Wal-Mart stores in 2004, Wal-Mart’s total procurement, direct and indirect, of Chinese-made goods may be two to three times as much as this official figure, implying that savings from sourcing from China are substantially greater than the operating income from the international stores! In that sense, buying Chinese goods cheaply and reselling them at a profit in the United States is a far more important part of Wal-Mart’s cross-border strategy than its international store network.

As a second example—one concerning a company that should have paid more attention to arbitrage sooner than it did—consider the sad case of Lego, the Denmark-based manufacturer of children’s building blocks and associated paraphernalia. Lego’s performance began to suffer in the late 1990s due to excessive diversification and aggressive competition in its core business, particularly from Canada-based Mega Brands, Inc., which started to sell much cheaper blocks (MEGA Bloks) sourced from China. But Lego continued injection-molding its own blocks in Denmark and Switzerland, which resulted in products that were up to 75 percent more expensive and inferior in financial performance (figure 6-1).² Lego’s performance has since rebounded as it has refocused on its core business and outsourced most of its production to contract manufacturer Flextronics, which is moving production offshore. However, Lego faces a much more serious, established competitor in MEGA Brands—in a category that Lego created and of which it was the namesake.

These examples suggest an asymmetry between the attention devoted—by commentators and sometimes even managers—to arbitrage opportunities versus other reasons for extending operations across borders. There are multiple reasons why arbitrage often doesn’t get the attention that it deserves—and they need to be identified before they can be corrected.

First, there seems to be a general sense that the activities underlying the traditional forms of arbitrage—hunting, fishing, farming, digging, weaving, and so on—are, well, backward. When it comes to the glamorous task of making money across borders, haven’t we progressed beyond hunting and gathering? If you are inclined to agree, think harder about what it actually means to purchase tens of billions of dollars of goods a year from China—in a world in which distance still matters—and to use them to feed a lean-mean sales machine in the United States. This challenge and opportunity has led Wal-Mart’s Global Procurement Center in Shenzhen, China, to develop some very sophisticated capabilities.

Second, there is a belief that arbitraging fundamental factors, such as capital or labor, offers only limited opportunities for competitive advantage.³ After all, can’t these factors be sourced globally at the click of a mouse, rendering them a weak competitive reed to lean upon? My response is to point to the reality of semiglobalization discussed earlier in this book, particularly in chapter 1, as well as the implication that even apparently unspecialized factors such as labor and capital are specialized at the level of location (even if in no other way). Sourcing from China by many companies, including Wal-Mart, hasn’t yet lifted Chinese labor costs to U.S. levels and is unlikely to do so for decades—although it has played a role in raising them. I will return to this issue of sustainability in the last section of this chapter.

A third stereotype, related to the second one, involves the notion that the profit potential from arbitrage is quite limited. My response? Look at the above calculations concerning Wal-Mart or at other sectors, such as Indian software services, which I discuss at greater length later on. For now, note that TCS, the Indian leader, has averaged a return on capital employed of more than 100 percent over the last five years while growing revenues at a 30 percent-plus rate. Although TCS is beginning to aggregate at a regional level, its core strategy has historically been one of labor arbitrage.

Fourth—and this gets to the issue of why companies don’t do more to talk up arbitrage opportunities even when they recognize them—there is a great deal of political sensitivity about arbitrage, particularly labor arbitrage, even though it is happening all around us. One does not have to be a conspiracy theorist to guess that this has something to do with the claim by Wal-Mart—a company with a supply chain and information systems unmatched in its industry—not to know the total volume of goods from China flowing through its store network. Again, the management of this challenge, among others, will be discussed at more length later in this chapter.

Finally, much of the discussion about arbitrage focuses—as did the Wal-Mart example—on obtaining labor-intensive goods (or services) from emerging markets and selling them in developed ones. This is a very important form of arbitrage, but it is far from the only one. If we are to give arbitrage its due, then we have to broaden our view of it.

One way to stretch our thinking about arbitrage possibilities is to cite exotic examples. Let’s consider some, mostly from recent headlines. Zhang Yin, the world’s richest self-made woman, with a net worth of over $3 billion, got her start importing waste paper from the United States and recycling it.⁴ Bumrungrad Hospital in Thailand, a pioneer in medical tourism, annually treats close to half a million foreign patients in its five-star facilities.⁵ A number of East European countries also attract many patients across borders in distinct specialties: the Czech Republic in cosmetic surgery, Latvia in knee surgery, Hungary in dentistry, and Slovenia in fertility treatments.⁶ Portuguese investors are contemplating building enormous retirement complexes for wealthier North Europeans.⁷ About 3,500 very wealthy individuals from all over the world have become Swiss citizens to benefit from local laws that set tax payments as a multiple of housing costs, without accounting for foreign wealth and income.⁸ LanChile has outperformed the averages for the airline industry with a strategy that capitalizes on Chilean exports of perishables such as salmon, fruit, and flowers: cargo accounts for 40 percent of its revenues, compared with 5 percent or less for large U.S. carriers.⁹ Some of Africa’s better boarding schools in countries such as Ghana, Kenya, and South Africa draw students, mostly of African extraction, from overseas.¹⁰ Remittances from emigrants account for more than 20 percent of GNP in a number of small countries, such as Moldova and Nicaragua.¹¹ And importing used cars is a bigger business, in terms of number of vehicles, than the new-car business in countries as diverse as Bulgaria, Jamaica, New Zealand, and Nigeria.¹²

CAGE and Arbitrage

The preceding examples all represent departures from—or at least variations on—the usual notion of low-cost manufactures from emerging markets being sold in developed ones. In particular, a number of examples illustrate the increasing incidence of cross-border arbitrage in services. But they do have a motley character. For a more comprehensive perspective on arbitrage, look at it through the lens of the CAGE framework since each type of difference between countries highlights a potential basis for arbitrage.¹³

Cultural Arbitrage

Favorable effects related to country or place of origin have long supplied a basis for cultural arbitrage. For example, French culture or, more specifically, its image overseas has long underpinned the international success of French haute couture, perfumes, wines, and foods.

But cultural arbitrage can also be applied to newer, more plebeian products and services. Consider, for example, the extraordinary international dominance of U.S.-based fast-food chains, which, at the end of the 1990s, accounted for twenty-seven of the top thirty fast-food chains worldwide and for over 60 percent of global fast-food sales.¹⁴ In their international operations, these chains exploit—to varying degrees—the global spread of American popular culture by serving up slices of Americana (at least as it’s perceived locally) along with their food. An even more extreme example is supplied by Benihana of Tokyo, the “Japanese steakhouse.” Although there was a predecessor outlet in Tokyo, the company’s Web site lists its first restaurant as opening on Broadway in New York. Benihana serves up a theatrical version of teppan-yaki cooking that the company describes as “eatertainment” and others as ersatz Japanese—and still has only one outlet in Japan, out of more than one hundred worldwide (heavily focused on the United States).

Nor are such “country-of-origin” advantages reserved for rich nations. Poor countries, too, can be important platforms for cultural arbitrage. Examples include Haitian compas music, Jamaican reggae, and dance music from the Congo, all of which enjoy image advantages in their respective regions.

We often hear claims that the scope for cultural arbitrage is decreasing over time, as the world becomes more consistently “vanilla.” But this clearly does not apply to all countries and product categories, as attested to by the launch of a number of successful place-branding consultancies in recent years. Or to be more specific, the persistent association of Brazil with football, carnival, beaches, and sex—all of which scream youth—is a case of cultural-arbitrage potential that companies have just begun to recognize. Thus, Inbev of Belgium, the world’s largest brewer in terms of volume, is now taking Brazil’s Brahma beer global—although the export version has a different formulation from the Brazilian one, a fancier bottle, and a premium positioning. According to Devin Kelley, Inbev’s vice president for global brands, Inbev saw the beer as a product that captured the essence of Brazil—even before considering what it tasted like. “The emotional context of Brahma, at the heart and soul of this incredible country called Brazil, was the single most important factor.”¹⁵

In fact, new opportunities for cultural arbitrage are appearing all the time. Thus, the push by the European Union to tighten rules for geographical designations on food products such as “Parma ham” and “Cognac brandy” would reinforce the natural advantages of particular countries or places of origin. What’s more, as Finland’s recently developed reputation for excellence in information technology indicates, in certain product categories, such advantages can now be created much faster than before: in years rather than decades or centuries. Meanwhile, reductions in other dimensions of the CAGE differences—tariffs or transport costs, for example—can also increase the viability of cultural arbitrage. For example, selling products or services to diaspora based on “back-home” appeal has become easier than ever before.

Administrative Arbitrage

Legal, institutional, and political differences from country to country open up another set of strategic arbitrage opportunities. Tax differentials are, perhaps, the most obvious example. Through the 1990s, to cite just one case, Rupert Murdoch’s News Corporation paid income taxes at an average rate of less than 10 percent, rather than the statutory 30 to 36 percent of the three main countries in which it operated: Britain, the United States, and Australia. By comparison, major competitors such as Disney were paying close to the official rates.

These tax savings were critical to News Corporation’s expansion into the United States, given the profit pressures on the company: net margins consistently less than 10 percent of sales in the second half of the 1990s and an asset-to-sales ratio that had ballooned to three to one. By placing its U.S. acquisitions into holding companies in the Cayman Islands, News Corporation could deduct interest payments on the debt used to finance the deals against the profits generated by its newspaper operations in Britain. Overall, the company has incorporated approximately one hundred subsidiaries in havens with no or low corporate taxes and limited financial disclosure laws. The intangibility of its informational assets has helped in this regard. As one accounting authority put it, “There’s absolutely no reason why a piece of paper, which is the right to show something, couldn’t sit anywhere, so it could be sitting in the Cayman Islands.”¹⁶

Most companies that cross borders do pay attention to international tax differentials and other administrative bases of arbitrage because of the large implications for value. However, they tend to be wary of discussing such considerations because the administrative gray areas that underpin them might substantially be curtailed or even eliminated. Thus, in a phenomenon known as round-tripping, many mainland Chinese businesspeople channel investment funds through foreign parties and then back into China, often through Hong Kong, in order to secure better legal protection, tax concessions, or otherwise favorable treatment. In fact, one-third or more of the FDI ostensibly flowing into China is estimated to have originated in China! And the tiny island nation of Mauritius (population 1.2 million) is, in most years, the top “source” of FDI flowing into India (population 1 billion) because of a tax treaty as well as, to a lesser extent, cultural links (two-thirds of Mauritians are of Indian extraction). More broadly, enclaves, tax havens, free-trade areas, export-processing zones, cross-border cities, and the like tend to be hot spots for administrative arbitrage. And some do very well out of it. The richest country in the world in 2006 was Bermuda, with an average GDP per person of almost $70,000, or 60 percent more than the average for the United States.

Much of what goes on under the rubric of administrative arbitrage is legal or at least semilegal—even when it has an odor to it, as in the relocation of economic activity, ranging from manufacturing activities to waste dumps, to exploit lax environmental rules. But cross-border criminal activity—drug production and distribution, human trafficking, illegal arms dealing, other forms of smuggling, counterfeiting, money laundering, to cite the major categories—also tends to involve some component of arbitrage, especially administrative arbitrage.¹⁷ The size of such arbitrage opportunities helps explain why the cross-border component of total criminal activity probably exceeds the 10 percent presumption—although it is obviously impossible to be certain.

The kinds of companies discussed in this book tend to work within or around the rules instead of breaking them. They can and do, however, try to use their political leverage to change rules that they do not like. Thus, in late 2006, the Confederation of British Industry warned that the burden of taxation in the United Kingdom could cause an exodus of corporations—clearly a pitch for limiting taxes and compliance burdens.¹⁸ A somewhat different kind of example is provided by companies using powerful home governments to pressure foreign governments into granting favorable treatment. Enron, for example, enlisted the help of the U.S. State Department, which obligingly threatened to cut off development assistance to Mozambique—one of the poorest countries in the world—if it granted a gas deal to a South African competitor instead of to an Enronled consortium.

Sordid? Yes, especially since the story involves Enron. But such stories remind us that companies help shape the administrative rules of the game: that they can be rule makers instead of simply rule takers, and that power differentials—at the governmental level as well as the company level—matter.

Geographic Arbitrage

Considering all that has been said and written about the alleged “death of distance,” it’s not surprising that few strategy gurus take geographic arbitrage very seriously. Yes, it’s true that transportation and communication costs have dropped sharply in the last few decades. But that drop does not necessarily translate into a decrease in the scope of geographic arbitrage strategies.

Consider the case of air transportation, the cost of which has declined more than 90 percent in real terms since 1930—a steeper drop than experienced by other, older modes of transportation over the same period. In fact, thanks to air transport, new opportunities for geographic arbitrage have been created. For example, in the international flower market in Aalsmeer, Netherlands, more than 20 million flowers and 2 million plants are auctioned off every day, with customers in the United States or Europe buying blooms flown in from, say, Colombia on the day they arrive. While this is a special example, the trade-related boom that transport companies—all of which can be thought of as geographic arbitrageurs—experienced between 2003 and 2006 reminds us that geographic distance still matters: if it didn’t, they would be facing bleak futures. The example of LanChile emphasizing cargo has already been cited. Note that the boom also extends to purely domestic transporters—for example, U.S. railways carrying Chinese goods from West Coast ports to other parts of the country—since geographic distance continues to matter within as well as between countries.¹⁹

Although communication costs have dropped even more sharply than transportation costs, they haven’t eliminated opportunities for geographically based arbitrage either. Thus, Cable & Wireless (C&W), the U.K-based telecommunications company, generated 37 percent of its revenues but 74 percent of its earnings from its international operation in 2005–2006.²⁰ High international profits involve taking advantage of residual distance by serving thirty-three relatively small markets around the world—many of them islands, whose communications links with the outside world are still dominated by C&W.

In fact, the overall evolution of international telephony has been greatly influenced by administrative arbitrage over residual administrative distance, even as some of the effects of geographic distance have weakened. Basically, regulatory regimes that prop up prices have consistently lagged advances in technology. In the days of telecom monopolies, a customer living outside the United States might call a personal computer in the United States, which would then call back the customer and the destination number (in a third country) and connect them, taking advantage of typically lower U.S. outbound rates. And currently, services such as Skype arbitrage the difference between distance-sensitive, administered pricing of long-distance calls, and the distance-invariant costs of Internet protocol (IP) telephony.

The geographic arbitrageurs that have lost some ground in recent decades are the great general trading companies of the past, which traditionally took advantage of large international variations in prices for a broad array of products by getting them from country A to country B. Lower transportation costs and greater connectivity have made it much easier for manufacturers and retailers to exploit these opportunities themselves.

Nevertheless, the savviest trading companies have found ways of staying in business. Thus, instead of simply trading on its own account, Hong Kong–based Li & Fung derives most of its revenue from more sophisticated geographic (and economic) arbitrage. It uses its offices in forty countries to set up and manage multinational supply chains for clients—or what might better be described as supply nets. For example, a down jacket’s filling might come from China, the outer shell fabric from Korea, the zippers from Japan, the inner lining from Taiwan, and the elastics, label, and other trim from Hong Kong. Dyeing might take place in South Asia, stitching in China, and quality assurance and packaging in Hong Kong. The product might then be shipped to the United States for delivery to a retailer such as The Limited or Abercrombie & Fitch, to which credit risk matching, market research, and even design services might also be provided.²¹

What’s all this activity about? It’s about creating multiple possibilities for arbitrage by slicing up the value chain ever more finely across geographies—or engaging in what economists have recently labeled “trade in tasks.”²² Thus, the major impact of decreasing transport and communications costs has not been on geographic arbitrage per se, but on the scope for economic arbitrage, which they have greatly increased and to which we turn next.

Economic Arbitrage

In a sense, all arbitrage strategies that add value are “economic.” But I use the term here to refer to the exploitation of economic differences that don’t derive directly from cultural, administrative, or geographic differences. These factors include differences in the costs of labor and capital, as well as variations in more industry-specific inputs (such as knowledge) or in the availability of complementary products.

The best-known type of economic arbitrage is the exploitation of cheap labor, which is common in labor-intensive, capital-light manufacturing (e.g., garments). What is worth emphasizing here is that high-tech companies can use that strategy just as effectively.

Consider the case of Embraer, the Brazilian firm that is one of the world’s two major suppliers of regional jets. While many factors, including managerial and technical excellence, contribute to Embraer’s success, labor arbitrage has also played a key role. To be specific, Embraer’s employment costs came to $26,000 per employee in 2002, versus an estimated $63,000 for the regional jet business of its archrival, Montreal-based Bombardier. If Embraer had had Bombardier’s employment cost structure, its operating margin would have fallen from 21 percent of revenues to 7 percent, and its net income would have turned negative. Not surprisingly, Embraer has focused its operations on final assembly, the most labor-intensive part of the production process, and has outsourced other operating activities to its supplier partners in richer countries with higher labor costs.²³ And labor arbitrage is also one of the bases of the challenge to Bombardier and Embraer being mounted by China Aviation Industry Corp I, a state-run manufacturing group, that is, with the help of a network of international suppliers, developing large regional jets to be offered at 10–20 percent lower list prices.²⁴

Capital cost differentials might seem, at first blush, to offer slimmer pickings than labor cost differentials—after all, the former are measured in single percentage points rather than in multiples of up to ten or twenty like the latter. But most companies (at least in the United States) earn returns within two or three percentage points of their cost of capital, so such differences are consequential, especially in capital-intensive industries. The Cemex case supplies a potent example of arbitrage in financing (chapter 3).

While we generally focus on economic arbitrage in the context of operations and financing, it can also be exploited in other functional activities. Consider Starent Networks, a company founded in August 2000 in Tewksbury, Massachusetts, with the mission of switching wireless networks to all-IP (Internet protocol) telephony. Soon after its founding, the company ran into what founder Ashraf Dahod describes as the “nuclear winter” of the meltdown in the telecommunications sector.²⁵ The company survived—and thrived—by migrating its product development function to India. And Starent is far from an unusual example: more U.S. companies seem to have offshored product development than call centers or help desks, even though the offshoring of call centers and help desks has attracted much more attention.²⁶

                              

The preceding discussion rounds out our understanding of each of the four broad dimensions of distance embedded in the CAGE framework as a potential basis for arbitrage. Even more numerous than the bases for arbitrage are the variety of arbitrage strategies. To further broaden our thinking about arbitrage, let’s consider a detailed example that illustrates this variety.

Varieties of Arbitrage: The Example of Indian Pharmaceuticals

When people think of the pharmaceutical industry, they typically think of “Big Pharma”: a dozen-odd multinational firms headquartered in the United States and Europe that account for about one-half of the worldwide market in terms of value.²⁷ Big Pharma firms historically generated high returns by developing and marketing drugs protected by patents—particularly blockbuster drugs, defined as ones that generate more than $1 billion in annual revenues.

In recent years, however, Big Pharma has come under a great deal of pressure: Accenture calculates that the overall market value of the pharmaceutical sector—mostly accounted for by Big Pharma—dropped from more than $2 trillion in 2000 to less than $1.5 trillion by 2005.²⁸ Big Pharma’s problems are various and include declining R&D productivity and general bloat. As my retired Harvard colleague Mike Scherer puts it, “[High] prices drive costs.”²⁹ But the challenge I concentrate on here is that from copycat generic drugs. Although generics have long threatened drugs coming off patent, soaring drug costs and buyer consolidation, among other structural changes, have recently intensified their impact on branded pharmaceuticals. Thus, according to Medco Health Solutions, three big drugs that came off patent in 2005 had 87 percent of their prescriptions switched over to generics within one month.³⁰

Generic drugs must meet the same quality standards as branded drugs, but are typically sold—after a six-month period of exclusivity for the first generic in the United States—at prices that are 20 to 80 percent lower than their branded counterparts. Generic drugs account for between 10 and 15 percent of the pharmaceutical market by value and significantly more by volume.³¹ Moreover, they are thought likely to attack another 30 percent of the current market in the United States alone over the next five years, as key drugs go off patent.³²

There are many generic-drug manufacturers worldwide—about 150 significant ones by one count. The largest, Teva of Israel, had $5.3 billion in sales in 2005. Teva’s success is rooted in administrative arbitrage: according to Eli Hurvitz, who ran it for twenty-six years, it owes its existence to the Arab boycott of companies doing business with Israel.³³ In response, Israel let local companies copy drugs patented overseas if their owners didn’t market them locally—which is how Teva built up its process expertise.

A similar administrative loophole underlies the success of a more recent wave of Indian challengers in generic drugs. Pharmaceutical manufacturing in India long reflected a policy that recognized process patents but not product patents, and thereby rewarded the reverse engineering of imported drugs. Since 2005, Indian patent laws have adjusted toward international norms as a result of the country’s accession to the World Trade Organization (WTO). But because of this history, low labor costs and buyer willingness to pay, and cutthroat domestic competition, the larger Indian manufacturers have developed low-cost manufacturing capabilities that have let them build up a significant position in generic drugs. One indication is that Indian companies account for 25 percent of the Abbreviated New Drug Applications (ANDAs) filed with the U.S. Food and Drug Administration (FDA) to launch generic drugs. Very diverse strategies underlie this level of market penetration—even if one focuses just on the top ten Indian firms out of several thousand.

Some Indian firms have continued to focus on imitating drugs coming off patent or drugs that are still under patent in some places but can be marketed in other, unregulated markets. The first approach is the one traditionally followed by generic drugs competitors. The second approach is illustrated by the second-largest Indian pharmaceutical firm, Cipla. In 2000, Cipla announced generic anti-HIV antiretrovirals that reduced the annual price of treatment from $11,000 per patient to $400 per patient. Cipla’s products are thought to account for one-third of the anti-HIV/AIDS medications taken in Africa, and the company stands to gain additional markets if other governments invoke WTO provisions that let them declare a national emergency and license the production or sale of drugs without the permission of the patent holders—as happened in Thailand at the end of January 2007.³⁴

Other Indian firms have started collaborating with Western firms by either in-licensing the latter’s products—usually with a view to manufacturing and marketing them in India—or manufacturing active pharmaceutical ingredients and intermediates that are then marketed by Western firms outside India. One example of a company that has pursued both approaches is Nicholas Piramal, India’s eighth-largest pharmaceutical firm, which has essentially avoided generic exports—and associated frictions with Big Pharma—in order to focus on such partnerships. It has licensed in drugs from several Western firms and has emphasized custom manufacturing for and R&D partnerships with them (see below).

Yet other Indian pharmaceutical firms, such as its largest, Ranbaxy, have come to focus on innovating or, more broadly, pioneering.³⁵ Like most other Indian firms, Ranbaxy built up overseas sales—now 80 percent of its total—with generic exports, but in recent years, it has pushed the envelope in a number of ways. To benefit from a statutory six-month period of exclusivity in the United States, Ranbaxy has been particularly aggressive in its attempts to be the first to manufacture generic versions of drugs going off patent. This approach has entailed extensive litigation, sometimes unsuccessful (e.g., its challenge to Pfizer’s patent on the world’s top-selling drug, the anticholesterol drug Lipitor), but other times offering rich rewards (e.g., the anticholesterol drug simavastatin). Another set of innovation initiatives has aimed to improve off-patent drugs (e.g., through novel delivery systems) to develop so-called branded supergenerics. Thus, in 1999, Ranbaxy licensed its once-a-day formulation of ciprofloxacin, an antibiotic, on a worldwide basis to Bayer. More recently, the company has emphasized patented inhalation devices and transdermal patches. While subject to a more complex approval process, supergenerics benefit from three years of marketing exclusivity in the United States.

Another, even more important innovative approach employed by Ranbaxy—and other larger Indian firms—has been to invest in developing entirely new drugs. In total, Indian firms are estimated to have as many as three dozen “new chemical entities” at relatively advanced stages of development. But the cost (including failures) of discovering and developing a new drug is estimated, in the West, to exceed $1 billion—more than the annual turnover of all Indian pharmaceutical companies but Ranbaxy. Thus, most Indian firms that are attempting to develop new drugs—for example, Dr. Reddy’s, the third-largest—have been explicit about their intent to out-license promising drug candidates as a way of defraying the costs and risks of clinical trials and launch.

Out-licensing also suggests a variety of related strategies involving focus on value-chain activities other than drugs manufacturing:

• Contract R&D: Instead of simply engaging in contract manufacturing, a number of Indian firms are also undertaking contract R&D for Western manufacturers. Such an approach focuses on the largest arbitrage differentials in the sector: Pfizer estimates that Indian chemists make about $5 an hour, versus more than $50 an hour for U.S. scientists. Thus, in early 2007, Nicholas Piramal and Eli Lilly signed an agreement under which the former will be responsible for the global design and execution of preclinical and early-stage clinical work for some of the latter’s new drugs.

• Clinical trials: A new medicine must go through clinical trials—the final and most expensive stage of trials—on a carefully selected sample of patients. These trials have also attracted great attention from drug-industry arbitrageurs. More than 40 percent of all clinical trials are now conducted in poor countries.³⁶ India attracts particular attention because of a large supply of patients, many of whom are “treatment-naive” (i.e., don’t consume lots of pharmaceuticals), and of English-speaking doctors.³⁷

• IT-enabled services: India has been successful as a destination for IT-enabled services, accounting for nearly half of total offshoring activity in 2005.³⁸ As a result, the pharmaceutical sector has shown great interest in exploiting its potential to contain the surging costs of data management and informatics support during the drug development process in areas such as data entry, database management, and trial study design, customer support services, and data analytics.

This characterization is far from complete—one could distinguish further among strategies by mode of growth (internally or by acquisition, with the latter mode attracting many Indian firms recently), areas of specialization, geographic focus, and so on. One could also look at cultural or geographic bases of arbitrage by considering India’s traditional medicine systems—Ayurveda, Siddha, and Unani—and its biodiversity. But the variety of possible arbitrage strategies should already be clear.

The responses from Big Pharma have also been varied. Novartis, the fifth-largest pharmaceutical company in the world, is one example. Novartis bought out Hexal of Germany for $8.3 billion in 2005 to cement its position as one of the two largest generic-drug manufacturers in the world, and has tried to bundle generic and branded medicines to offer health-care providers “one-stop shopping.”³⁹ As far as India is concerned, Novartis participates in the market there, as the fifth-largest foreign player. It also undertakes clinical trials and software development in India and, in early 2006, opened a global R&D center for over-the-counter medicines near Mumbai. But Novartis’s big resource commitment has been to China, which a number of Big Pharma companies view as having even more potential than India. In late 2006, it announced an investment of $100 million in an R&D facility in Shanghai that would initially focus on cancers caused by infections—a significant proportion of Chinese cancer cases. And Novartis has also been active on the legal front: in January 2007, it filed a challenge in an Indian court to a decision not to grant it a patent on a modified form of its leukemia drug, Glivec—prompting an official of Médecins Sans Frontières to comment, “Novartis is trying to shut down the pharmacy of the developing world.”⁴⁰ Arbitrage clearly poses a number of choices for Big Pharma, not just for second-tier players.

Analyzing Arbitrage

Given the variety of arbitrage strategies, there is no one way to analyze them. But, again, the ADDING Value scorecard can be used to structure the analysis and helps suggest some specific dos and don’ts. The key point to remember is that arbitrage can affect all the components of the ADDING Value scorecard, not just the first D of decreasing costs.

Adding Volume

Arbitrage can affect volume in a number of ways. Sometimes, arbitrage opportunities can open up entirely new businesses—for example, the cut-flower business in the Northern hemisphere in the winter. At other times, the volume increments come from the fact that, in the absence of arbitrage, one might have to turn business away—a particularly sore point for many executives in the high-tech sector, who complain about the difficulty of finding the right kind of technical talent in many developed countries.

A somewhat different source of volume growth involves securing market access. Reconsider Novartis’s decision to establish a major R&D lab in Shanghai. While cost arbitrage is presumably one motive for the move, analysts have stressed another: improving relationships with governmental authorities who will decide what kind of medicines to buy for their citizens.⁴¹ So, in assessing such a move, it is important to factor in the positive effects on future volumes instead of simply adopting a perspective focused purely on costs. Otherwise, one would reject some moves that made sense.

All these mechanisms presumably combine to account for the finding, in a recent survey, that growth is the second most frequently cited reason for offshoring, after cost reduction.⁴² It is not necessary to assess their relative importance to conclude that the idea of arbitrage adding volume has to be taken seriously.

Of course, along with the possible volume-expanding effects of arbitrage, it is important to verify the expansibility of the basis of arbitrage that underpins it. GEN3 Partners, based in Cambridge, Massachusetts, provides a good example of such constraints. The core business of GEN3 Partners is innovation consulting for major U.S. companies using Russian specialists trained in TRIZ, a rigorous Soviet-era methodology for inventive problem solving.⁴³ At the end of 2005, GEN3 had about one hundred personnel in Russia, of whom about half were PhDs or Doctors of Science who also met the company’s requirements of at least five years of practical experience. It was possible to imagine expanding this pool to several hundred, perhaps, but not beyond that number. Given those volume constraints, GEN3 had to pursue a much-higher-end business model than did, say, Indian software services firms with access to hundreds of thousands of new technical personnel every year. GEN3 was already operating at a revenue-per-employee level of $100,000 and was targeting $200,000-plus, versus $50,000 to $70,000 for Indian software services firms, even though Russian salaries per employee were still somewhat lower.

Decreasing Costs

While cost reductions are the most frequently cited reasons for arbitrage-related moves, analyses of them are often flawed in conceptually simple but practically dangerous ways. One frequent problem is a focus on snapshots of relative costs. As the GEN3 example suggests, this can be very misleading: in that case, the cost of personnel with the requisite training in TRIZ was bound to escalate sharply if GEN3—and other consulting firms tapping into the same resource base—proved successful at monetizing the resource. Other frequent, related problems include a failure to adjust for likely shifts in exchange rates (the Chinese yuan, for instance, is probably significantly undervalued relative to Western currencies, suggesting that forward-looking assessments based on today’s exchange rates will overstate the likely Chinese cost advantage) and for differentials in productivity (in many Chinese and Indian cases, still a fraction of Western levels). Even apparently authoritative estimates can be subject to such problems. For example, the Organisation for Economic Co-operation and Development (OECD) recently came out with a report that had China’s R&D level surpassing Japan’s. This “finding” was based on the idea that because China’s scientists and engineers cost one-quarter of Japan’s at official exchange rates, why not just multiply China’s R&D expenditures times four!⁴⁴

Nevertheless, while acknowledging some reasons why labor arbitrage may be less advantageous than it appears at first blush, it is also important to recognize some positive factors that often get overlooked. Consider the kind of calculation that I tend to frown on when I encounter it in the classroom: “The costs of Indian software personnel are (say) one-third those of U.S. software personnel but are escalating at 15 percent per year, so the Indian cost advantage will get wiped out within eight years.” As an argument against arbitrage, this misses out on several key points:

• The opportunity costs of turning business away (see “Adding Volume”)

• The likelihood of even greater pressures on cost and availability in developed countries—which leads to predictions that the total surplus generated by India and other countries that are currently low-cost suppliers of IT workers will increase rather than decrease over the next few years (figure 6-2)

• The possibility of faster productivity improvements and cost containment in India—as achieved in the 1990s through a shift from on-site to offshore development

• Quality differentials that, in this case, may actually favor Indian competitors, as discussed in the next subsection

The broader point to be made here is that it is important to go beyond naive labor cost comparisons to look a bit more deeply at the kinds of negatives and positives outlined above. Not considering them amounts to an assumption that they will exactly cancel out, which will happen only by accident.

Yet another problem has to do with the common focus on labor costs (and productivity). But as chapter 3 emphasized, we need to look at costs comprehensively. Thus, work by the Boston Consulting Group (BCG) has highlighted the capital that can be saved by building plants in rapidly developing economies such as China.⁴⁵ BCG estimates that for discrete manufacturing processes, it is possible to reduce typical capital investments by 10 to 30 percent below Western levels by using local equipment suppliers; by 20 to 40 percent by also targeting process improvements such as replacing capital with labor and rethinking make-or-buy decisions; and by 30 to 60 percent by totally overhauling the operating model by revamping entire production chains, redesigning products for local manufacturability, and moving to continuous utilization from a five-days-per-week model. The benefits include not just higher returns on investment, but lower fixed costs and breakeven points and reduced exit barriers in case operations have to be discontinued—that is, lower risks as well as costs.

The ultimate objective of the analysis should be, of course, to build up a comprehensive picture of costs rather than to focus on a single cost element, whether labor, capital, or something else. This, along with the differentiation-related factors discussed next, is the key determinant of whether a product or service is likely to be offshored. Cost-related flags of “offshoreability” include not only the factors already discussed, but also high value-to-bulk ratios, short supply chains, and broad availability of required inputs and skills. But to understand the degree of incentive to move products or services offshore rather than treating offshoreability as “on” or “off”—it is usually better to look comprehensively at costs instead of relying on flags of this sort.

To take a concrete example, why have Indian software services companies been able to grow much more quickly and profitably than Indian pharmaceutical companies? Part of the answer lies in the greater labor-intensity in software, with employee-related costs accounting for one-half or more of revenues. A more comprehensive answer is that Indian software firms’ total costs (per employee) are still one-third or less of the costs of the Western competitors that the Indian firms are trying to take business away from, whereas Indian pharmaceutical firms’ costs are probably two-thirds or more of the costs of Western generic companies. An even more comprehensive answer would also take into account the administrative barriers to economic arbitrage in pharmaceuticals.⁴⁶

Differentiating

The impact of arbitrage on differentiation or willingness-to-pay has attracted much less attention than its impact on costs, but can be at least as important. Cultural arbitrage, for instance, often involves raising willingness-to-pay based on country-of-origin effects. Of course, as the discussion in chapter 3 should have warned us, such effects can, depending on the situation, have negative rather than positive effects.

Examples of economic arbitrage reinforce the importance of actually analyzing the implications of arbitrage for differentiation. While economic arbitrage often reduces willingness-to-pay as well as costs, there are important exceptions to this rule. Software services seem to be one such exception. The Indian software firms are lower priced as well as lower cost than their Western competitors, but this seems to reflect reputational stickiness rather than quality differences. In fact, there is evidence that some of the larger Indian firms—such as TCS, the largest firm, which has also taken the lead at compiling such data—actually offer both higher quality and lower costs for software maintenance, in particular, than some of their better-known Western counterparts.⁴⁷ Corroboration comes from the fact that India still accounts for one-half of the software development centers certified to operate at the highest levels of process compliance even though it accounts for only about one-tenth of the total global IT workforce. This example and TCS’s inauguration, in the first quarter of 2007, of a marketing campaign focused on communicating its dual competitive advantage, underlines the importance of:

• Not treating price as a proxy for quality or willingness-to-pay in the long run.

• Really digging deep into buyer economics. Thus, TCS’s campaign stresses the implications of poor software quality for buyers’ total costs of quality, including rework—which is estimated to be one-half or more of the total IT spending at a typical large organization.

• Actively communicating such benefits instead of simply assuming that buyers will figure them out for themselves.

As in the case of costs, it is possible to flag differentiation-related correlates of offshoreability: a product or service is less likely to move offshore when considerable customization is required, demand is very changeable, local presence or service requirements are high, and purchasing decision makers are public rather than private. But, once again, it is usually better to try to build up a comprehensive, ideally quantitative picture of relative viability—of the comparative width of the wedge between willingness-to-pay and costs—instead of simply relying on such flags.

Improving Industry Attractiveness or Bargaining Power

In addition to possibly lowering costs or raising willingness-to-pay, arbitrage may improve industry attractiveness or one’s own bargaining power within it. Thus, IBM has increased its head count in India from nine thousand to fifty thousand in less than three years not only to improve its own economics, but also to put pressure on its Indian rivals by attacking what is still probably their single most important source of advantage.

Again, it would be rash to assume that arbitrage always has a particular type of effect in this regard. Thus, while global companies have implanted many R&D centers in China and India—with a focus on electronics and telecoms in China, and software and engineering in India—the protection of intellectual property rights remains a major concern. One study of global companies with R&D centers in China indicates that they have found several ways to address this issue. One particularly important method is to split R&D efforts across a company’s global network in such a way that the value of the projects undertaken in China is highly contingent on projects being pursued elsewhere in the global network or, more broadly, on firm-specific expertise.⁴⁸ Also note that this strategy is unavailable to local firms, probably contributing to their lower expenditures on, and returns from, R&D.

Such splits are not a perfect solution: think of Cisco’s allegations, since settled, that Huawei Technologies Co., Ltd., of China poached its switching technology across multiple geographies. But the concept does suggest two important lessons about strategy. First, in the study cited above, the firms that did particularly well with their Chinese R&D centers appear to have been ones with stronger internal linkages, reminding us that competitive advantage can be developed in dealing with institutional failures as well as on more traditional bases. Second, it should be clear that the external environment need not be taken as given: elements of it can and, ideally, should be influenced through firm strategy.

Normalizing Risk

Arbitrage is subject to an extensive array of risks, both market and nonmarket. With regard to the former, think of all the (greater) hazards to which supply chains that cross borders are subject: exposure to unknown and potentially less reliable suppliers and to exchange-rate fluctuations; the possibility of infrastructural and other bottlenecks at borders; the compounding of risks associated with supply chains that are very thinly sliced across multiple countries, as in the earlier example of the down jacket from Li & Fung, and so on.

Li & Fung’s network also affords numerous insights, however, in how to cope with such risks. After the 9/11 terrorist attacks on the United States, it reportedly took Li & Fung less than three weeks to relocate time-sensitive activities from partners in Pakistan to partners in other countries deemed more politically secure. Li & Fung obviously practices such dynamic arbitrage over longer time frames as well, in response to changes such as major shifts in exchange rates. Such moves occur within the context of a broader strategy in which capacity and materials subject to long lead times or stable demand are locked in well ahead of time, but decisions concerning attributes that are highly sensitive to market fluctuations are deferred for as long as possible.

A particular type of risk associated with arbitrage has to do with its political sensitivity—particularly evident around but not confined to labor arbitrage. Note that such risk isn’t confined to external constituencies: as IBM, for example, moves offshore, its top managers must be careful not only in their public communications, but also in internal ones. Successful arbitrageurs offer several lessons in this regard. First, be discreet: emphasize viability and growth as objectives rather than (just) cost reductions, and be cautious about capitalizing on health, safety, and environmental standards that are looser than at home. Second, think through a range of mechanisms—lobbying, working with natural allies, including companies who are otherwise competitors, investing in job creation, and so forth—for expanding your freedom of action. A final suggestion is to favor strategies that exhibit some degree of robustness to changes in the political climate.

The political risk associated with arbitrage strategies should be balanced with recognition of the possible political risks of counterstrategies. Think of the example mentioned in the last section: the lawsuit in an Indian court by Novartis. Risks, like the other components of the ADDING Value scorecard, should be looked at in comparative perspective, across alternatives.

Generating Knowledge—and Other Resources and Capabilities

The final component of the ADDING Value scorecard, which I’ll mention briefly, is generating knowledge—and other resources and capabilities. Again, arbitrage strategies can have either positive or negative implications. On the positive side, IBM and Accenture’s efforts to expand in India probably will help bolster their long-run capabilities, even if the short-run impact on operating economics is negative because of lower price realizations, setup costs, and the internal disarray associated with very rapid expansion. On the negative side, a major investment bank that has outsourced many analytical functions to India is just starting to realize that the move will significantly deplete its pool of senior analysts in a few years—unless it significantly changes its hiring and promotion policies.

This discussion of the importance of analyzing arbitrage and, especially, applying the ADDING Value scorecard, should have clarified why arbitrage is often thought about far too narrowly. Individual CAGE bases of arbitrage can be used to target different components of the ADDING Value scorecard; in addition, complex cases involving multiple bases of arbitrage further expand the possibilities, as illustrated by the discussion in the previous section of pharmaceuticals.

Managing Arbitrage

This chapter, like chapters 4 and 5, was meant to stretch one’s thinking about how to deal with differences—in this instance, by expanding on the potential for exploiting them through arbitrage strategies. A number of challenges arise in managing arbitrage strategies, though. Some, such as the risks that arbitrage engenders, particularly political risks, have already been discussed. But what deserve additional attention are the sustainability of arbitrage strategies and how they are influenced by firm-level resources, particularly management capabilities, as opposed to market-level differences in prices, costs, et cetera.

Note, first of all, that while sustaining competitive advantage from arbitrage can be a worthwhile objective, it is not necessary for arbitrage to make sense. Reconsider the example of Wal-Mart. Even if arbitrage didn’t offer Wal-Mart a sustainable advantage, it would probably be worth undertaking, if only to avoid ending up at a cost disadvantage relative to other competitors—a serious obstacle to Wal-Mart’s pursuit of its low-cost strategy.

The Lego versus MEGA Brands example is even clearer. Lego chose to deal with arbitrage opportunities by essentially outsourcing them to a contract manufacturer, Flextronics. That is, Lego moved to neutralize arbitrage as a source of competitive advantage or disadvantage and to focus on advantage from its intangible assets: its brand name, its relationships, and its innovation capabilities. In contrast, Wal-Mart, in keeping with its traditional emphasis on back-end efficiencies, has opted to try to build a competitive advantage around arbitrage by developing distinctive capabilities to manage what is probably the world’s biggest cross-border sourcing operation. This contrast highlights a second theme that is worth underscoring: building a sustainable competitive advantage through arbitrage generally requires a commitment to building firm-specific capabilities—a commitment that takes years if not decades to implement. Zero firm-specific capabilities, in contrast, may allow you to achieve parity, but not much more.

The same theme stands out from some of the other examples considered in this chapter. Embraer certainly benefits from cheap Brazilian labor, but its ability to turn that labor—theoretically available to all—into a source of competitive advantage depends on capabilities that it has developed to run a world-class aerospace operation amid the turbulence of contemporary Brazil. And Ranbaxy’s innovation-driven vision of success in pharmaceuticals was already in place a decade ago, when I first wrote a case on the company and invited its president to come talk to my Harvard MBA class.

For a somewhat more detailed example of what is involved in converting such visions into reality, let’s look once again at Tata Consultancy Services (TCS), the largest Indian software services firm. TCS arbitrages in the reverse direction from Wal-Mart: it buys cheap at home (in the form of software developers) and sells dear overseas. But this model, which TCS pioneered, is no longer unique: it is also used by virtually everyone else involved in software exports from India, ranging from Infosys, the second-largest firm (whose founder was quoted in the epigraph to this chapter), to the smallest body-shopping operation. The result of all this has been a large run-up in the cost of Indian software developers. How has TCS managed to sustain superior performance despite the sharp escalation of labor costs?

Figure 6-3 tracks TCS’s revenues per employee, costs per employee, and net profit per employee since the end of the 1980s. TCS’s costs per employee have more than tripled since then. To sustain its performance, TCS has had to increase the revenue per employee even faster: revenue has quadrupled rather than tripled over this period, and as a result, profit per employee has jumped severalfold! The revenue-per-employee performance is even more impressive when one accounts for a significant shift in business from on-site development over this period to more efficient offshore development out of India today: this reduces revenue per employee by 30 to 40 percent, although it does boost absolute profit per employee.

What lies behind these numbers? In addition to moving work offshore, TCS has, over time, migrated to working on projects that are larger and more complex and that offer higher realizations per employee—a shift in which it seems to lead the Indian software sector. And with revenues per employee that are still significantly less than those of Accenture and IBM Global Services—as well as quality advantages along some dimensions, as discussed—there is clearly headroom to do even better. The target for the medium run is to boost revenue per employee by another 25 to 30 percent. But, of course, the achievement of these targets, not to mention past accomplishments, has required a level of firm capability that goes well beyond the insight that Indian software developers are relatively inexpensive.

Similar capability requirements are discernible on the demand as well as the supply side of this business. To successfully offshore some or all of its software service requirements, a client company must have some ability to specify those requirements—greater than if the work were simply being done in-house, at corporate headquarters—and to track outcomes. In recognition of this fact, as some of the larger software outsourcing deals negotiated in the mid-1990s come up for renewal, sophisticated companies seem to be increasingly chunking them up and splitting them across multiple vendors instead of single-sourcing them from them a mega-vendor. But, as in any group, it is still possible to find companies well behind the curve. My most cautionary example is the large European bank that has hired Indian software companies to complete more than one thousand of its projects, but still does not track performance across projects or vendors.

If this last example sounds too extreme to be true or widely relevant, note that only 1 percent of the respondents to a recent survey by Duke University’s Arie Lewin say that their companies have corporatewide strategies for offshoring.⁴⁹ Yet, in the absence of such a strategy, one might see offshoring efforts run amok or, probably more likely, an insufficient amount of offshoring due to internal barriers. The solution to this state of affairs is partly strategic and partly organizational, involving mechanisms such as the creation of internal champions for arbitrage, incentives for project managers as well as top managers, and pull-through commitments by top managers.

To summarize this section, a sustained commitment is required to develop and deploy capabilities to arbitrage effectively—and not all commitments are possible at any one organization at any given point in time. Put differently, firms may have to make trade-offs between arbitrage and the other elements of their strategy. Some mixing and matching may be possible across the AAAs of adaptation, aggregation, and arbitrage—recall that TCS has managed to overlay a degree of regional aggregation on its core strategy of arbitrage. However, pursuing all three strategies, or even two of them, full-tilt may lead to inconsistencies.

For a cautionary example of a company that failed, for a while, to grasp such trade-offs, consider Acer of Taiwan, one of the world’s largest computer manufacturers. Acer entered into the contract manufacturing of personal computers early and made good money with that arbitrage play. But in the early 1990s, the company began to push Acer as a global brand (and a basis for aggregation) across countries, particularly developed ones. This two-track approach turned out to be problematic. While the branded own-product business did grow to significant volumes, it continued to generate losses. Meanwhile, customers for its contract manufacturing arm worried about the spillover of business secrets to, and cross-subsidization of, Acer’s offerings under its own brand. Matters came to a head in 2000, when IBM cancelled a major order, reducing its share of Acer’s total contract manufacturing revenues from 53 percent in the first quarter of 2000 to only 26 percent in the second quarter of 2001. Eventually, Acer made some hard choices. Its contract manufacturing continued to focus on customers in advanced countries and was gradually spun off into a separate company called Wistron. Meanwhile, Acer refocused its own-brand sales on the East Asian region, particularly Greater China. While the revised strategy faces its own challenges, it certainly seems to have worked better than the old one.

In my MBA classroom presentations, this is usually about the time that I show a PowerPoint slide of a celebrated and controversial beast: the jackalope (figure 6-4). I won’t go into the debates surrounding this animal, which may or may not inhabit the American West, lure cowboys to their doom with uncanny imitations of human song, or produce a milk that is a powerful aphrodisiac.⁵⁰ I invoke the jackalope here simply to make the point that asking an animal—or an organization—to be more than one thing at one point in time can lead to some very awkward results. Some degree of internal consistency is a basic requirement for a good strategy—or organization. A jackrabbit with a full rack of horns probably can’t hold its head up, let alone run down the road.

For now, I’ll let my jackalope metaphor serve as a placeholder for a more serious discussion of one of the most challenging questions in contemporary global strategy: To what extent is it possible to mix and match across the AAA strategies of adaptation, aggregation and arbitrage? Chapter 7 deals with that question at length.

Conclusions

The box “Global Generalizations” summarizes the specific conclusions from this chapter. More broadly, arbitrage further expands our strategy toolkit for dealing with the differences across countries. As the Acer example emphasizes, however, decisions about arbitrage cannot be made independently of decisions about the other elements of a firm’s strategy. The next chapter digs deeper into this point.