123

www.hbr.orgNew tools based on modern communications technology, social networking, and game theory can help you prevent foreign governments from regulating away your profits.

The Hidden Risks in Emerging Markets

by Witold J. Henisz and Bennet A. Zelner

Included with this full-text Harvard Business Review article:

1 Article Summary

Idea in Briefthe core idea

2 The Hidden Risks in Emerging Markets

Reprint R1004HThis article is provided compliments of Verizon Business. The Hidden Risks in Emerging Markets

Idea in Brief

Governments in developing countries no longer seize foreign investments. Instead they find ways to divert the profits through regulation or selective lack of regulation. The costs are estimated to be equivalent to at least a 33% increase in tax.

Most of the traditional ways to limit risk to your companys foreign investments are not very effective anymore. You now have to learn how to play politics directly.

As companies such as the Italian oil giant Eni have demonstrated, managing political risk involves balancing operational effi-ciency with political capital, mastering the art of political spin, and hitting the pressure points of local decision makers. That means mastering new analytic skills and tools.

This article is provided compliments of Verizon Business.

page 1 New tools based on modern communications technology, social networking, and game theory can help you prevent foreign governments from regulating away your profits.

The Hidden Risks in Emerging Markets

by Witold J. Henisz and Bennet A. Zelner

When a firm with a value-generating techno-policy response to the crisis give way to de-

logical or managerial capability investsbates about who should pay and how much.

abroad, its shareholders and the host countrysPoliticians will struggle to balance popular de-

citizens both stand to benefit. But no mattermands to punish those perceived as responsi-

how good the apparent fit between what for-ble against fears of stymied innovation and the

eign companies offer and what host countriesflight of human and financial capital. Broader

need, success is far from assured. Elections anddomestic economic concernsfor example,

other political events, economic crises, andprotectionist sentiment in response to the re-

changing societal attitudes can disrupt thealignment of economic power in favor of

best-laid plans in both emerging and advancedemerging nations such as China and Indiaeconomies. The interplay of these forcesandwill inevitably affect the debate as well. The

the implications for the political choices thatmultinational firms best able to anticipate and

multinational firms makewill become espe-manage the related risks and opportunities

cially prominent as national governmentswill have the strongest competitive edge.

chart an uncertain course toward stabilizationHistorically, the biggest risks faced by for-

following the global financial meltdown.eign investors were in developing countries

Issues such as taxation of executive compen-with immature or volatile political systems.

sation, the proper scope of financial regula-The chief concern was "expropriation risk," the

tion, and international M&A have come to thepossibility that host governments would seize

harvard business review april 2010

foreground in the wake of the crisis, and starkforeign-owned assets. Today, this risk has

international differences in opinions and poli-largely disappeared. Stronger international

cies on these matters are already evident. Thelaw and the symbiotic nature of growth in

differences will only become more pronouncedemerging and developed economies reduced

as discussions about the appropriate near-termasset seizures to nearly zero during the 1980s.

page 2

This article is provided compliments of Verizon Business. The Hidden Risks in Emerging Markets

However, as interest in emerging markets haslandscape and for modeling political decision

soared, host countries have learned, accordingmaking. As with the management of any risk

to George Chifor at the University of Windsoror uncertainty, political mastery can become a

in Canada, "that more value can be extractedsource of competitive advantage in addition to

from foreign enterprises through the morea means of avoiding losses.

subtle instrument of regulatory control rather

than outright seizures." The risk that a govern-Its Hard to Hedge Policy Risk

ment will discriminatorily change the laws,Firms engaged in international business often

regulations, or contracts governing an invest-use some combination of legal contracts, in-

mentor will fail to enforce themin a waysurance, and trade in financial instruments to

that reduces an investors financial returns isprotect the income streams from their invest-

what we call "policy risk."ments against currency or price swings. These

Although the data on policy risk are lessapproaches, however, offer little protection

clear-cut than the hard numbers on direct sei-against policy risk.

zures, press mentions of policy risk(usingFor starters, legal contracts are useful only if

terms such as "political risk," "political uncer-they are enforced, and shifting laws and regu-

tainty," "policy risk," "policy uncertainty," "regu-lations can render them void. In the 1990s

latory risk," and "regulatory uncertainty") indi-many Southeast Asian governments wooing

cate that it has risen dramatically as seizureprivate power investors offered contracts that

risk has fallen. (See the exhibit "The Changinginsulated the investors from risks related to

Face of Risk in Emerging Markets.") Press men-lower-than-expected demand, fuel supplies, ex-

tions of actual seizures have also increasedchange rates, currency conversions, regula-

somewhat since 2001, but that does not reflecttions, and political force majeure. The Asian fi-

a broad-based resurgence in seizures.nancial crisis in 1997 brought those investorsOther recent data are consistent with thefavorable treatment into sharp relief as cur-

finding that policy risk has increased greatly. Arency values, share prices, and electricity de-

2001 PriceWaterhouseCoopers study con-mand all plummeted. Political officials had to

cluded that an opaque policy-making environ-choose between honoring the contracts, at the

ment is equivalent to at least a 33% increase inrisk of compromising their own popular sup-

taxation. A World Bank study in 2004 revealedport, and renegotiating them in order to main-

that 15% to 30% of the contracts covering $371tain that support. In the end, many career-

billion of private infrastructure investment inminded public officials in Southeast Asia chose

the 1990s were subject to government-initiatedto renegotiate or cancel scores of contracts.

renegotiations or disputes. And a 2009 surveyEven when contracts can be legally enforced,

by the Multilateral Investment Guaranteeexperience shows that inventive politicians can

Agency and the Economist Intelligence Unitcircumvent them, through a wide variety of

found that multinational enterprises consid-means other than changing laws. For example,

ered breach of contract, restrictions on thein 1998, when U.S.-based AES Corporationtransfer and convertibility of profits, civil dis-then the worlds largest independent power

turbance, government failure to honor guaran-companyacquired the Georgian electricity

tees, and regulatory restrictions all to be moredistribution company Telasi, high-priced law-

significant risks than the potential seizure ofyers constructed an ironclad set of guarantees

assets.that allowed AES-Telasi to pass the costs of pol-

Unfortunately, the traditional financial andicy and other risks on to Georgian consumers.

contractual mechanisms that firms use to as-One analyst remarked to us, "If you believed

sess and mitigate business risks have limitedthe contract, AES was guaranteed a 20% return

value. Therefore, investors must develop pro-on its investment." The Georgian government

Witold J. Henisz (henisz@whartonactive political-management strategies thatactually never interfered formally with AES-

.upenn.edu) is a professor at the Whar-lessen government officials incentives to di-Telasis ability to pass costs on to consumers.

ton School at the University of Pennsyl-vania in Philadelphia. Bennet A. Zelner ([email protected]) is a professor at Duke Universitys Fuqua School of

Business in Durham, North Carolina.


vert investors returns. In this article, we ex-However, the venture was doomed by public

plore the experiences of multinational inves-officials inactionfor instance, their failure to

tors as they confront these issues in a variety ofterminate supply to nonpaying industrial con-

industries and countries, and we offer best-sumers, to supply fuel to AES-Telasi, and to

practice guidelines for assessing the politicalkeep the governments own account currentpage 3


and by the governments demand for tax pay-ments on electricity for which the company had never been paid. The result was that AESs "guaranteed" 20% return became a share-holder loss of $300 million.

Insurance offers limited protection against policy risk because a firms exposure is largely determined by its own ability to manage the policy-making process. In the words of one in-surer: "I prefer to focus on what my assured [customer] can bring to a risk. My reasoning is that if you back the right assured, you can usu-ally keep problems from occurring in the first placeand if they do happen, you have an ex-cellent chance of mitigating your loss." Yet it is very difficult for insurers to know who the "right assured" is, and the firms with the great-est risk exposure are often those most likely to seek insurance in the first place. As a result, un-derwriters price their products extremely high, offer very short-term coverage, or dont offer any coverage at all.

Financial hedges have limited value for simi-

The Changing Face of Risk in Emerging Markets

Overt seizures of foreign assets by host countries in emerging markets essentially evaporated by 1980. However, other political risks to those assets (for example, from potential regulatory action) have risen dramatically since then.

PRESS

MENTIONS

SEIZUREOF OTHER

ACTIONSPOLITICAL

RISKS

Source: Seizure data (left side) from M.S. Minor, "The Demise of Expropriation as an Instrument of LDC Policy, 1980-1992," Journal of International Business Studies 25 (1994): 177-88.


lar reasons. Instruments for hedging against risks in specific emerging marketssuch as ex-change-rate, market, and credit risksare ubiquitous because multiple parties are willing to participate. The project- and firm-specific nature of policy risk, however, renders conven-tional hedging strategies infeasible.

Some of the more-inventive instruments are based on the average risk premium associated with existing companies in a given country but they give false comfort. Because the base-line risk premiums are those of firms that are actively participating in a given market (and that often have their risk-mitigation strategies in place), new entrants are likely to face far greater exposure. In fact, foreign investors who focus on constructing financial hedges at the expense of developing their own risk-mitiga-tion strategies may increase their exposure. It is therefore not surprising that, despite the ability to calculate residual risk premiums, no financial institutions have used such premiums to price an instrument that pays out money when a policy risk is realized.

The New Risk-Management Playbook

Given the difficulty of constructing hedges against policy risk through contracts, insur-ance, or financial risk-management tools, for-eign investors must accept the responsibility for directly managing the risk themselves. For many companies, that means rewriting the playbook. Instead of looking for immediate ways to improve operations, managers have to move beyond the quick cost-benefit analyses that they usually undertake and think more about how they can frame and shape public debate. And they must learn how to apply po-litical pressure, either individually or as part of a coalition.

Investing in goodwill. In the developed world, managers spend a great deal of time and energy on improving efficiency. When companies move into less developed markets, they often expect huge, instant efficiency gains from exploiting the technologies, busi-ness models, and practices that they have man-aged to hone in their home markets. Unfortu-nately, the political costs of such practices may outweigh those gains.

Consider the 1997 Christmas blackout in large parts of Brazil, including Rio de Janeiro. The then recently privatized electric utility

page 4


Political mastery can become a source of

competitive advantage and a means of avoiding losses.


Light (in which AES held a 13.75% stake) faced record-high outdoor temperatures that week, and it was already struggling with poorly maintained equipment that had deteriorated before privatization. However, the press and the public focused on the 40% reduction in personnel, combined with the utility com-panys record profits, to paint a picture of an exploitative foreign investor. The negative sen-timent toward foreign firms in general and AES in particular contributed to the awarding of a 900-megawatt energy-supply contract the following spring to a joint venture led by Bra-zilian firms (Votorantim Group, Bradesco Group, and Camargo Corra) rather than con-sortia led by AES and British Gas.

A smarter approach was used by Italian state-owned oil company Eni. After the 1998 devaluation of the real, when many companies put their investment plans on hold or even ex-ited Brazil, Enis then-CEO, Franco Bernabe, visited Rio de Janeiro to announce a $500 mil-lion investment. He proclaimed: "Now is the time to show that Petrobras [the state-owned oil company] has long-term friends." Eni and Petrobras have collaborated closely ever since.

Framing the debate. When companies enter new countries, they often engage in extensive PR campaigns that amount to little more than advertisements for the brand and specific com-mercial ventures. Instead, firms need to mas-ter the art of political spin. Presenting a ven-ture as "fair," "equitable," or "growth enhancing" is often a simpler and more pow-erful means of securing political support than providing a cost-benefit analysis. The precise meaning attributed to such labels varies de-pending on a firms market position. New en-trants garner support for policies that favor them over incumbents by citing the abuse of monopoly power. Conversely, dominant firms appeal to "fairness" by arguing that smaller entrants cannot survive without the govern-ments helping hand.

This type of debate played out in the South Korean wireless market. LG Telecomthe third entrant, behind the much larger SK Tele-com and Korea Telecommade repeated calls for "asymmetric" government regulation of the market leaders in order to "level the playing field." As the Korea Times reported, "The defin-ing question is whether the government will back new entrants in the name of encouraging fair competition, or limit the pool to experi-

enced players." LG ultimately prevailed: In May 2001 the South Korean government an-nounced that it would "guarantee a market share of at least 20% for a third major telecom operator through asymmetric regulation on Korea Telecom and SK Telecom."

Finding political pressure points. The net-work of relationships in a society greatly influ-ences policy outcomes, especially in countries with weak legal systems. To turn these net-works to their advantage, international inves-tors must identify and engage local politicians power bases. Once again, Eni has shown the way, this time in Kazakhstan. Through its sub-sidiary Agip KCO, Eni has adopted a business model that responds to the former Soviet re-publics economic and social needs. The com-pany favors Kazakh over non-Kazakh suppli-ers, and it conducts knowledge-transfer, training, and development seminars for them. At least 60% of local employees are Kazakh cit-izens. The company also funds the construc-tion of various public works, including the na-tional library, the prime ministers residence, schools, computer labs, and multifamily hous-ing units for the poor. As a result, many Ka-zakh officials now have a stake in Enis success.

For the vast majority of organizations which do not possess enough leverage to influ-ence the full range of relevant actors on their owna crucial component of an effective strategy is to assemble a coalition of interests. In the South Korean wireless battle, LG Tele-com benefited from the influence of upstream suppliers. The major Korean carriers wanted to shift to the globally favored WCDMA standard for the newest generation of cellular service, but domestic champion Samsung had devel-oped a global leadership position in the com-peting CDMA2000 technology. Under pres-sure from Samsung, the government insisted that one of the new 3G licenses be awarded to LG Telecom in return for its promise to adopt CDMA2000.

An international investors home govern-ment can also be a powerful channel of influ-ence. Observers in central Europe have noted the lobbying success of the German and French governments on behalf of national champions in countries seeking EU member-ship. However, the use of "foreign influence" may create a perception of meddling, can stoke nationalism, and is generally less likely to have a lasting impact. Theres also the risk that your

page 5


home government will sacrifice your needs in order to gain traction on another issue.

Taking these pages out of the political play-book requires building the sorts of capabilities in intelligence gathering and analysis that are familiar to politicians, spies, and journalists. Managers must begin by understanding the at-titudes, opinions, and positions of relevant ac-tors toward their firm, the industry in which the firm operates, and any specific actions that the firm might take to influence outcomes on the playing field.

Tapping the Right Flow of Data

Traditionally, managers who have undertaken political analyses in a host country have di-

Why Country Risk Ratings Dont Work

rectly consulted employees, local business partners, and supply chain partners. The infor-mation-gathering process varies in intensity and structure, ranging from surveying radio and newspaper stories to conversing with lo-cals to using computerized contact-manage-ment systems. Some firms rely almost exclu-sively on informal chats, whereas others favor more-formal Delphi (iterative expert survey) methods. (Also see the sidebar "Why Country Risk Ratings Dont Work.")

Although these sources provide valuable conventional input, they can require more time and money than such small, subjective, potentially biased snapshots might merit. Moreover, given the availability of multiple real-time indicators and metrics in functional areas such as finance, marketing, and human resources, CEOs and boards of directors in-creasingly demand similar real-time data on the preferences of key players. This human in-

When it comes to assessing levels of pol-short on analysis, as an example fromtelligence can be effectively and continuously

icy risk, managers are far too quick toChile and Indonesia clearly shows. Inincorporated into enterprise risk-management

rely on the subjective ratings of country1997, one risk index ascribed an identicalmodels and frameworks.

"experts." One popular index focuses onscore to those two countries. The mea-To broaden their perspectives, more and

asset-seizure and contract-repudiationsure took no account of the significant in-more companies are reaching out to nonbusi-

risks. Ratings are incorporated, in thestitutional differences between them.ness organizations that can help them antici-

form of country risk premiums, into theFaced with violent citizen demands to re-pate and preempt consumer concerns about

discount rates used to evaluate invest-distribute investor returns in the wake ofenvironmental, health, and safety issues. For

ment opportunities. This approach ap-the 1997 Asian financial crisis, Indonesiasexample, after a bruising experience over the

pears to have the formal rigor of finan-longtime military dictator, General Su-disposal of its Brent Spar oil-drilling platform

cial risk management, but it is actuallyharto, renegotiated contracts with for-in 1995, Royal Dutch Shell now routinely in-

inadequate.eign investors that were unaffiliated withcludes Greenpeace in substantive environmen-

To begin with, such ratings usually failhis family or close friends. After he wastal discussions. Some companies also consult

to account for the fact that the levels ofousted in a coup, the previously favoredprofessional experts, ranging from well-posi-

policy risk vary among different investorscompanies experienced a backlash as thetioned ex-government officials operating on re-

in a country, some of whom may adaptsuccessor government renegotiated theirtainer; to the stringers who write for the Econ-

their business practices to local normscontracts.omist Intelligence Unit, Stratfor, and Oxford

and lobby key policy makers better thanChile, in contrast, had a democraticAnalytica; to global political consultancies,

others do. Also, policy-risk exposure is tomultiparty system and possessed a well-such as Political Risk Services or Eurasia

some extent contingent on the relativerespected independent judiciarya fur-Group. Although employees, suppliers, and ac-

importance of the proposed investmentther check against arbitrary policytivists may have access to better information,

to the two parties (how easy is it for thechange. Pressures in Chile to enhance eq-they lack the specialized training that these ad-

firm to walk away, and how badly doesuity and social cohesion culminated invisers bring to the table.

the local government want the deal?). Fi-the 2000 election of socialist RicardoOf increasing importance is the vast amount

nally, country risk ratings are usually ret-Lagos as president. He shifted some dis-of information emanating from third-party

rospective, reflecting past policy out-cretionary spending toward social pro-sourcesprimarily the mainstream news me-

comes. To assess the correlation withgrams but also respected the rule of lawdia, but also bloggers and other observerscurrent policy risk, an analyst needs toand existing commercial contracts. Un-that routinely monitor the policy-making pro-

determine how similar the past and present policy-shaping factors actually are.

Even as purely country-level measures, most political risk scorecards are woefully


derlying risks in Chile and Indonesiacess in various countries. The large volume and

were therefore very different, but therelatively unfocused nature of the material

country-level ratings didnt reflect thosemake it hard to synthesize, digest, and act

distinct realities.upon effectively, even if a company has sub-

stantial resources for this activity. However,

page 6


with information-extraction software, its now possible to identify the relevant political and social actors on a given issue and their inten-sity of interest in it.

One approach, known as data mining, relies on the coincident location of words to derive information about key players preferences. For example, the occurrence of "Russia," "AES-Telasi," and "protest" in the same sentence im-plies a negative sentiment in the relationship between Russia and the electricity investor AES-Telasi. Another tool, called natural lan-guage parsing (NLP) software, facilitates more-refined sentence-level inferences by syntacti-cally distinguishing among subjects, verbs, and objects, thereby identifying the orientation of actions or preferences. Consider this possible sentence: "The Union of Consumers of Georgia is outraged by the AES-Telasi American com-pany proposal to increase the tariff on electric energy." NLP software would recognize the precise grammatical relationship among "Union of Consumers of Georgia," "is outraged by," and "AES-Telasi...proposal"pointing to a strong negative sentiment toward the U.S. company. NLP software can also gauge the in-tensity of sentiment. If the verb phrase in the

Are the Locals Hostile to You?

Information-extraction software can capture changes in attitudes toward a business venture by syntactically analyzing the content of media reports about it. This exam-ple compares the total number of sentences in articles about Gabriel Resources plan to develop the Rosia Montana gold mine in Romania with the percentage of sen-tences indicating NGO opposition to the plan.

TOTAL NUMBER

sentence had been "objects to" instead of "is outraged by," the software would have recog-nized that the sentiment of the Union of Con-sumers of Georgia toward AES was negative, but less so.

Similarly, information-extraction tools can readily and objectively highlight shifts in an ac-tors preferences over time. For example, a coa-lition of local and international activists sharply contested the plan by Canadian min-ing company Gabriel Resources to develop the Rosia Montana gold mine in Romania. The ex-hibit "Are the Locals Hostile to You?" plots the frequency in the worldwide media of sen-tences mentioning statements or actions against the mine by nongovernmental organi-zations through 2007, relative to the total number of sentences in articles about the mine during the same period. The data show that NGOs were relatively indifferent to the issue until mid-2002, when negative reports in-creased sharply.

The "Tummy Test" and Other Models

With data about political actors and their level of interest in hand, managers must then syn-thesize that information into a model of the policy-making process. At the informal end of the synthesis spectrum is the "tummy test," in which a decision maker who has spoken with or been briefed about relevant sources draws upon his or her own knowledge of similar cases to make an educated guess about the likely policy outcome. The accuracy of this technique clearly varies enormously according to the skill set of the decision maker and the relevance of his or her past experience to the current situation. To improve the accuracy of

10%

9%

% OF TOTAL

8%SENTENCES THAT

INDICATE NGO

OPPOSITION 7%

6%

5%

4%

3%

2%

1%


OF SENTENCES800

PUBLISHED

ABOUT THE MINE700

600

500

400

300

200

100

such judgments, managers can also involve specialized consultancies that draw upon a more diverse set of experiences from multiple firms and industries in the target country or a comparable one.

A sophisticated extension of the tummy test is the "war room," in which managers come to-gether for a one-off meeting or a series of brainstorming sessions. Sessions may be sched-uled regularly or triggered by a shock or event that requires a strategic response. "Influence maps" are used to depict each politically rele-vant actor as a bubble arrayed in space accord-ing to the players position on a given issue, with the size of the bubble proportional to the

page 7


Instead of engaging in PR campaigns that amount to little more

than advertisements for the brand, companies need to master the art of political spin.

players power. Linkages across actors or clus-ters of actors can be indicated by either loca-tion or connecting lines. Although no formal analytic tools are used, the maps can help guide discussion of action scenarios: What hap-pens if we target actor X? What if we break the link between X and Y? What if we try to reduce Zs power? The insights produced by this ap-proach are, of course, only as good as the infor-mation brought into the room and the quality of the team assembled.

The most formal tool for modeling the poli-cymaking process is the dynamic expected util-ity model, which is based on game theory. It as-sumes that, in each of several time periods, every actor (an individual or an organization) with a vested interest in an issue has a choice of three possible alternatives: proposing a pol-icy, opposing a proposed policy, or doing noth-ing. Each actor chooses the alternative that maximizes his, her, or its expected utility in each period. The selection depends on the di-rection and intensity of the actors preferences, the salience of the issue, the cost of proposing or opposing a policy, and similar information about other actors. The combined actions of all the actors result in a likely policy outcome. The sensitivity of the outcome to various as-sumptions and parameters can then be calcu-lated, helping to identify which actors are so pivotal that a change in their preferences, power, or salience would have a large impact on policy.

Models like this are widely used by the intel-ligence community and by specialist consult-ing groups such as Mesquita & Roundell, Sen-tia Group, the Probity Group, and Commetrix. A growing number of multinational corpora-tions are also adopting these tools. A large Brit-ish company, for example, used such a model to decide how to influence the climate change

debate in the European Union. Analysts first identified which actors were most commonly cited in the press and whom these actors refer-enced in their speeches and writings. The ana-lysts then constructed a network of key "influ-encers" and modeled various points of entry into this system to identify the target areas and the messages that would maximize their effect on the climate change debate.

Although the integration of automated data collection, dynamic expected utility modeling, and influence-map visualizations remains in its infancy, the potential applications are broader than the management of policy risk alone. Marketing research, financial analysis, opera-tions, and human resources all could benefit from a richer analysis of the best ways to affect stakeholders opinions.

Of course, the risks of investment may simply be too great to justify entry into certain politi-cal zones. But in many cases investors who ex-plicitly recognize the dynamism of the envi-ronment and implement appropriate strategies to address it will find the risks quite manageable. By combining data-mining and modeling technologies with traditional ap-proaches, as weve described, they can start the journey forward, moving from "tummy tests" toward an analytically oriented, defensi-ble system for managing policy risk that will greatly expand their investment options. At its heart, this system will always retain elements of tacit knowledge and experience, and not all managers and firms will be able to master its intricacies. But those that do will find it a pow-erful source of competitive advantage.

Reprint R1004HTo order, call 800-988-0886 or 617-783-7500 or go to www.hbr.org.


This article is provided compliments of Verizon Business.

page 8 COPYRIGHT 2010 HARVARD BUSINESS SCHOOL PUBLISHING CORPORATION. ALL RIGHTS RESERVED.

COPYRIGHT 2010 HARVARD BUSINESS SCHOOL PUBLISHING CORPORATION. ALL RIGHTS RESERVED.

123

Documents