CHAPTER 3

Hurdle Rates for Overseas Operations

Many companies consist of different business operations in the form of divisions and subsidiaries. For a company’s individual business operations, the cost of capital is a concept that is central to valuation, investment (and divestment) decisions, measures of economic profit, and performance appraisal. Finance theory says that an operation has a cost of capital based on that specific operation’s risk.

An operation’s cost of capital is typically the basis of the operation’s hurdle rate. In principle, if the operation’s actual rate of return is not expected to be at least the operation’s hurdle rate, the operation is not adding shareholder value to the parent company. Alternatively, if the present value of the operation’s expected cash flows, discounted using the hurdle rate, is not higher than the operation’s invested capital, the operation is not adding shareholder value.

In global finance, the idea of operation-specific hurdle rates based on risk means that operations in different countries should have different hurdle rates. As a case in point, Dan Cohrs, the vice president and treasurer at the telecommunications company GTE Corporation in 1996, said that GTE did in fact set different hurdle rates for the company’s operations and projects in different countries.¹ This approach does not mean that one hurdle rate in US dollars should simply be converted into different but equivalent rates in different currencies, but that the operations in different countries have different hurdle rates from the perspective of the parent company’s home currency, based on the operations’ different risks.

Although no single best practice has been established to address the complicated issue of operation-specific hurdle rates, the aim of this chapter is to show some basics of how a firm might estimate an overseas operation’s hurdle rate, from the home currency’s perspective. For an operation in a developed country, the cost of capital and the hurdle are the same thing. For an operation in an emerging market country, the operation’s hurdle rate should reflect a consideration for political risk in addition to the cost of capital.

Operation-Specific Risk and Cost of Capital

We should think of an operation’s cost of capital as the expected rate of return that would be required on investing in the operation by the aggregate financial market, as compensation for risk, if the operation were an independently traded, all-equity company with no off-balance sheet risk management positions. Therefore, an individual operation’s cost of capital differs from the parent firm’s overall cost of capital if the operation’s risk is different from the parent firm’s overall risk. If an operation has higher risk than the parent’s overall risk, the operation’s hurdle rate should be higher than the parent’s overall cost of capital. If the parent’s overall cost of capital were used as basis of the operation’s hurdle rate, the manager may make investments that do not offer enough expected reward to compensate for the risk. If an operation has lower risk than the parent’s overall risk, the operation’s hurdle rate should be lower than the parent’s overall cost of capital. A manager who does not recognize this may miss out on value-adding investment opportunities by setting the operation’s hurdle rate too high.²

One reason an overseas operation’s risk is likely to differ from the parent company’s overall risk is that the operation and the rest of the company are likely to have different systematic relationships with the global economy. For example, the sales volume or the operating costs of a subsidiary in Switzerland may have a different systematic relationship with the global economy than does the rest of the parent company.

Explicitly or implicitly, market investors require a rate of return on their investments based on risk. As we saw in Chapter 1, the market’s required rate of return on a company’s shares is the company’s cost of equity. A firm’s cost of equity is one component of its weighted average cost of capital, or WACC, which you probably recall from prior finance courses. A firm’s WACC is relatively straightforward to calculate, and WACCs are used for many purposes in corporate finance; thus, many companies compute a WACC. Finance theory tells us that a firm’s WACC reflects the risk of the enterprise, including the impact of any debt tax shield values and financial risk management positions.

The WACC is a useful cost of capital concept for an overall company, but impractical for individual operations, which do not have their own capital structure and market data to compute a WACC. Moreover, we do not know how to adjust a parent’s overall WACC to find a cost of capital for an operation whose risk level varies from the overall risk level of the parent. Because of the problems of applying the WACC idea to individual operations, we instead apply risk–return theory directly to individual operations.³

Business Betas and Proxy Approach

The direct approach to the cost of capital for individual operations uses the idea of business beta, which is the beta the operation would have in the absence of debt, cash and marketable securities, and off-balance sheet financial risk management strategies. Sometimes a business beta is called an all-equity beta. By beta, we will always mean global beta without saying “global” each time. With a division’s estimated business beta, we will use a GCAPM or ICAPM risk–return model to help estimate an operation’s cost of capital.

Estimating the business beta for an individual operation has its own problems, because a business beta is not directly observable, and there are typically no historical return data to use in statistical estimation for individual operations. So, we need to be somewhat creative when estimating operation-specific business betas. We rely on the proxy firm approach to estimating the business beta for an overseas division. The method is not perfect, but is relatively simple and should be helpful.

In the proxy firm approach, you identify a firm (or firms) with traded equity and in a similar business as the operation. Sometimes, the parent firm itself is a suitable proxy firm if the firm is focused in a relatively homogeneous industry. For example, if a U.S. multinational telecommunications company wants to find the cost of capital for one of its overseas telecommunications subsidiaries, the parent is a reasonable choice for a proxy firm. In more diversified situations, a better proxy choice would be a different firm with a relatively homogenous business like the operation’s.

The idea is to estimate the business beta of the proxy firm, but this task is not as straightforward as estimating an equity beta with historical data, because there are typically no market observations of the firm’s business value. Instead, we must unlever the proxy firm’s equity beta estimate to get an estimate of the proxy firm’s business beta. To properly unlever an equity beta, we in principle should consider the systematic risk of the firm’s net debt and off-balance sheet financial risk management positions. For practical purposes, however, we ignore these items, and in many cases, we will still get a reasonable business beta estimate using the beta unlevering formula in equation (3.1):

Beta Unlevering Formula

In equation (3.1), is the business beta, is the equity beta, ND^C is net debt, and is the business value. All variables are measured from the perspective of currency C. Typically, you can use enterprise value for , ignoring that there may be some tax shield value embedded in the enterprise value.⁴

As a (hypothetical) example, assume that a U.S. parent with diversified global operations thinks that the U.S. firm Anderson-Wheeler Company is a reasonable proxy firm for one of its overseas operations. Anderson-Wheeler’s estimated equity beta (in US dollars) is 1.20, equity market cap is $60 million, net debt is $20 million ($30 million of debt and $10 million of cash and marketable securities.) Moreover, assume that Anderson-Wheeler has no off-balance sheet risk management positions with systematic risk, and its net debt is denominated entirely in US dollars and has no systematic risk. Ignoring debt tax shield value, the estimated business value is the enterprise value, $60 million + 20 million = $80 million. The ratio of net debt to business value, , is $20 million/$80 million = 0.25. Using equation (3.1), Anderson-Wheeler’s estimated business beta (in US dollars) is 1.20[1 − 0.25] = 0.90.

Regardless of the overseas operation’s country, the GCAPM may serve as the risk–return model if the parent company is based in a currency area where the GCAPM gives an acceptable cost of capital approximation to the ICAPM, as in the United States. For parent companies in countries where the GCAPM does not give an acceptable approximation to the ICAPM, we use the ICAPM approach of Chapter 2, which for equation (2.2) requires two risk parameters: (1) beta; and (2) FX exposure to a foreign currency index. If you have estimated a proxy firm’s equity beta and FX equity exposure to the foreign currency index, you will need to unlever both risk parameters. To unlever an FX equity exposure estimate, , you can use an equation is analogous to equation (3.1), but specified in terms of FX business and equity exposures to a foreign currency index: .

Home Country Proxy Firm

If a parent company identifies a suitable proxy firm (or firms) in the home country, one approach to estimating an overseas operation’s business beta is the Lessard country beta method. This method starts with the proxy’s business beta estimate, , where H denotes the parent firm’s home currency. The next step is to estimate the country beta of the overseas country’s equity market index. Country Y’s country beta from the perspective of currency H, , is the systematic risk of an index fund of country Y’s stocks from the currency-H perspective, and is the beta of regressing country Y’s equity index returns against the global market index returns, expressing both index returns in currency H.

Exhibit 3.1 shows some country beta estimates from both the US dollar and euro perspectives (based on the 1999 to 2016 data described in Chapter 2). Exhibit 3.1 also shows the analogous estimates for country FX exposure to the foreign currency index, , which from the euro perspective are useful in the following examples where we apply the ICAPM. Even though the examples use the GCAPM when the home currency is the US dollar, the country FX exposure estimates from the US dollar perspective are shown anyway. (Note that the country index risk coefficients in Exhibit 2.3 are from the local currency perspective, whereas those in Exhibit 3.1 are in US dollars or euros.)⁵

The last step is to get the home-currency business beta estimate for overseas operation i, , by multiplying the home country proxy firm’s business beta times the ratio of the overseas country beta to the home country beta, , as shown in equation (3.2). Since this method is adapted from the one pioneered by MIT finance professor Donald Lessard, we call equation (3.2) the Lessard country beta method.⁶

Country Beta Method (Lessard)

For example, say a U.S. multinational wants to estimate the cost of capital (in US dollars) for a subsidiary in Sweden. A home country proxy firm has a business beta estimate of 0.50. Assume that in US dollars, the Sweden country beta is 1.45 and the U.S. country beta is 0.94, as in Exhibit 3.1. Using equation (3.2), the business beta estimate in US dollars for the Swedish subsidiary is 0.50[1.45/0.94] = 0.77. Since Sweden’s country beta is higher than the home country (U.S.) beta, it makes sense that the Swedish subsidiary’s business beta estimate is higher than the home country (U.S.) proxy’s business beta estimate. This method is ad hoc, but is practical.

Exhibit 3.1 Country Beta and FX Exposure Estimates

In US Dollars and Euros
Estimation Period: 1999–2015
United States (dollar)	0.94	0.89	1.02	0.67
Eurozone (euro)	1.26	1.90	1.06	−0.16
Japan (yen)	0.74	1.06	0.81	0.78
China (yuan)	1.19	1.65	1.16	0.25
Britain (pound)	0.96	1.46	0.88	0.29
Canada (dollar)	1.12	1.56	1.09	0.38
Australia (dollar)	1.16	1.95	0.99	0.01
Taiwan (dollar)	1.08	1.12	1.13	0.57
Switzerland (franc)	0.82	1.54	0.67	0.17
India (rupee)	1.18	1.68	1.12	0.28
Korea (won)	1.43	1.76	1.45	0.47
Brazil (real)	1.68	2.50	1.51	−0.06
Mexico (peso)	1.21	1.32	1.27	0.52
Sweden (krona)	1.44	1.73	1.35	0.07
Hong Kong (dollar)	1.05	1.40	1.06	0.52
Norway (krone)	1.43	2.31	1.20	−0.21
Denmark Krone)	1.04	1.67	0.92	0.13
New Zealand (dollar)	0.95	1.69	0.82	0.13
Singapore (dollar)	1.14	1.58	1.13	0.43
South Africa (rand)	1.20	2.03	1.06	0.11
Thailand (baht)	1.21	1.95	1.16	0.37
Germany (euro)	1.38	1.85	1.22	−0.08
France (euro)	1.21	1.85	1.02	−0.14
Italy (euro)	1.19	2.07	0.90	−0.37
Netherlands (euro)	1.20	1.69	1.06	0.01
Belgium (euro)	1.13	1.94	0.91	−0.18
Ireland (euro)	1.07	1.43	0.99	0.15
Spain (euro)	1.23	2.24	0.92	−0.42
Austria (euro)	1.30	2.62	0.92	−0.66
Finland (euro)	1.50	1.69	1.44	0.15
Portugal (euro)	1.00	2.10	0.68	−0.39

For parent companies in countries where it’s better to use the ICAPM of Chapter 2, we should adjust a home country proxy firm’s FX business exposure to the foreign currency index, , to get operation i’s FX business exposure estimate from the home currency perspective, . But a multiplicative approach like equation (3.2) does not work because FX exposures can be either positive or negative. So, the adjustment is to simply add the difference between the FX exposure estimates for the overseas and home country equity market indexes, , as shown in equation (3.3):

Country FX Exposure Method

For example, say an Italian parent firm uses the ICAPM in equation (2.2) to estimate the cost of capital for a Swedish operation in the parent’s home currency, euros. After unlevering the equity beta and FX equity exposure estimates, the home country proxy firm’s estimated business beta, , is 1.20, and FX business exposure to the foreign currency index, , is 0.60. From the euro perspective (per Exhibit 3.1), Sweden’s country beta is 1.35 and country FX exposure is 0.07, and Italy’s country beta is 0.90 and country FX exposure is −0.37. Using equation (3.2), the Swedish operation’s business beta estimate (in euros), , is 1.20[1.35/0.90] = 1.80; using equation (3.3), the operation’s FX business exposure estimate from the euro perspective, , is 0.60 + [0.07 − (−0.37)] = 1.04.

Once you have an overseas operation’s business risk estimate(s) in the home currency, you can use a risk–return equation to calculate the direct home-currency estimate of the overseas operation’s cost of capital, . For example, let an overseas operation’s business beta estimate in US dollars be 0.90. To get a cost of capital in US dollars, we use the GCAPM. If the risk-free rate in US dollars is 3% the global risk premium in US dollars is 6%, the division’s estimate is equal to = 0.03 + 0.90[0.06] = 0.084, or 8.40%.

The Italian parent above uses the ICAPM in equation (2.2) to estimate the Swedish division’s cost of capital in euros. Assume that the risk-free rate in euros is 2.5%, and = 5.54% and = −0.92%, per Exhibit 2.2. Therefore, the Swedish operation’s cost of capital in euros is = 0.025 + 1.80[0.0554] + 1.04[−0.0092] = 0.115, or 11.5%.

Political Risk and Emerging Market Operations

For operations in many emerging market countries, analysts like to consider political risk in addition to the standard systematic economic/ financial risk in risk–return models. Political risk is a catch-all term used to describe the additional risks posed in terms of illiquidity, civil disruptions, corruption, political intervention, expropriation, imposition of controls on funds repatriation, irresponsible economic management by the country’s policymakers, and the like.

There is some disagreement about whether political risk should be an adjustment to an operation’s expected cash flows or to the hurdle rate. Since political risk is not a specified systematic risk factor in asset pricing models, some argue that any adjustment for political risk should be made to expected cash flows. But this advice has proven difficult to follow, and so the trend in practice is to include an adjustment for political risk in an emerging market project’s hurdle rate.

In principle, a given country’s political risk premium does not have a currency denomination, and is thus the same number from any currency perspective. Still, political risk premiums are difficult to measure. Many managers and analysts estimate country Y’s political risk premium, PRP_Y, by the yield on the country’s sovereign credit default swap (CDS), which is an “insurance policy” against the default of a sovereign bond issued by a sovereign country. In general, a CDS is insurance against the default of a bond. As long as the bond does not default, the insurance buyer makes payments to the insurance seller based on the CDS rate. If the bond does default, the insurance buyer delivers the bond to the insurance seller, while the seller pays the buyer the face value of the bond. Figure 3.1 shows the basic idea.

Exhibit 3.2 shows sovereign CDS yields for some emerging market countries for May 31, 2013. The quotes in Exhibit 3.2 are in basis points, so the CDS yield for Argentina is 3,144 basis points, or 31.44%. For Chile, the CDS yield is only 76 basis points, or 0.76%. Of course, CDS yields are market prices that change constantly with conditions; those in Exhibit 3.2 are only for illustration purposes.

Figure 3.1 Credit default swaps

Many analysts now prefer to base a PRP_Y estimate on a market-driven CDS yield instead of the earlier choice, the sovereign yield spread, which is the difference between the yields on sovereign debt issued in US dollars and a long-term U.S. government bond. Technically, however, CDS yields and sovereign yield spreads both measure sovereign risk, which is the risk that the country’s government will not service its debt obligations properly. In addition to political risk, sovereign risk entails some macroeconomic and financial risk that is captured in standard measures of systematic risk. So, estimating a political risk premium by either type of sovereign risk premium, SRP_Y, tends to involve some double-counting of systematic risk.

Exhibit 3.2 Sovereign Credit Default Swap Yields and Political Risk Premium Estimates

Selected Emerging Market Countries
May 31, 2013
	CDSY	PRPY/SRPY	PRPY
Argentina	3,144	0.31	9.75%
Brazil	146	0.79	1.15%
Bulgaria	120	0.40	0.48%
Chile	76	0.73	0.55%
China	84	1.00	0.84%
Colombia	104	0.87	0.90%
Croatia	310	0.56	1.74%
Egypt	623	0.79	4.92%
Hungary	292	0.32	0.93%
Indonesia	162	0.78	1.26%
Mexico	102	0.70	0.71%
Panama	103	0.82	0.84%
Peru	109	0.86	0.94%
Philippines	97	0.85	0.82%
Poland	76	0.86	0.65%
Russia	155	1.00	1.55%
South Africa	191	0.79	1.51%
Turkey	131	0.79	1.03%
Ukraine	601	0.52	3.13%
Venezuela	840	0.60	5.04%

Source for CDS yields: Bloomberg quotes on Deutsche Bank site (with permission).
Source for Political Risk to Sovereign Risk Ratio Estimates: Adapted from Bekaert et al. (2016).

Some researchers have tried to correct for this problem by estimating the percentage of sovereign risk that is represented by political risk, PRP_Y/SRP_Y. Estimates of these percentages are shown for some countries in Exhibit 3.2. For Chile, for example, the CDS yield is 0.76% and political risk is estimated to be 73% of sovereign risk, so the political risk premium estimate would be 0.73(0.76%) = 0.55%. On average, the political risk is 62% of sovereign risk for the 20 countries in Exhibit 3.2. So, if you have a CDS yield estimate for a country, but no estimate of PRP_Y/SRP_Y, consider using PRP_Y/SRP_Y = 0.62.⁷

For an emerging market country with political risk, individual assets and operations pose different degrees of political risk exposure. The political risk exposure of operation i, denoted Φ_i, is the operation’s degree of political risk relative to the overall country’s political risk. Some operations may be relatively free of political risk. An example might be a tomato plant in a stable food industry. At the other end of the spectrum are emerging market industries that are highly susceptible to political intervention, like the power and oil industries, or have relatively high potential for corruption.

One simple approach to the problem is to group industries into low, medium, and high political risk exposure categories. The low exposure category includes consumer discretionary and staples; the medium exposure category includes health care, industrials, information technology, materials, telecommunications, and utilities; and the high exposure category includes energy and financials. For operations in the medium exposure category, or if the manager has no opinion about the operation’s political risk exposure, the manager should assume that Φ_i is equal to the average political risk exposure for the country, which is 1. For operations with high political risk exposure for the country, Φ_i should be higher than 1, say 1.50. For investments judged to have low political risk exposure, let Φ_i = 0.50.⁸

In some cases, managers may want to make other adjustments in political risk exposure estimates. One example is if a foreign operation has issued its own debt to local investors. This debt may reduce the operation’s political risk exposure. Another factor may be the extent to which the operation brings foreign currency into the emerging country. This idea is explored in the box titled “Brazilian Firms Embraer and Embratel.” One method in that box is to estimate Φ_i by the ratio of the operation’s local revenues to the average local revenues of firms domiciled in the country. For example, Embraer derives only about 3% of its revenues locally, while Embratel derives 95% of its revenues locally. Since the average Brazilian firm generates 77% of its revenues locally, the Φ_i for Embraer would be 0.03/0.77 = 0.04, and the Φ_i for Embratel would be 0.95/0.77 = 1.23.⁹

In this text, the hurdle rate for emerging market asset i (in the parent’s home currency), , is the asset’s cost of capital, , plus the adjustment political risk. The adjustment for political risk is technically not part of the cost of capital but is added to the cost of capital to get the hurdle rate because that’s a more practical way of dealing with political risk than adjusting expected cash flows.

You can use either the GCAPM or the ICAPM to capture the cost of capital. Using US dollars as the home currency and the GCAPM as the risk–return model, the cost of capital, , is equal to , and the hurdle rate is given in equation (3.4):¹⁰

Emerging Market Hurdle Rate Model

For example, assume that a U.S. multinational wants to use the GCAPM and to estimate the hurdle rate in US dollars for its subsidiary in Hungary. Assume: (a) Hungary’s political risk premium, PRP_Y, is 0.93%, per Exhibit 3.2; (b) the subsidiary has low political risk exposure, Φ_i, of 0.50; (c) the subsidiary’s business beta in US dollars is 0.90; (d) the risk-free rate in US dollars is 3%; and (e) the global risk premium in US dollars is 6%. The Hungarian subsidiary’s hurdle rate estimate in US dollars, applying equation (3.4), is = 0.03 + 0.90[0.06] + 0.50[0.0093] = 0.0887, or 8.87%.

Estimating hurdle rates for emerging market operations is clearly a challenging task. The previous suggestions may provide some guidance, but there is plenty of room for subjectivity and judgment. One unanswered question is: How does the level of local investment affect the political risk exposure of an emerging market operation? If a U.S. company produces in the United States and exports the products, does the company’s foreign operation have higher or lower political risk exposure than if the foreign operation has a local plant to produce for that market? If the operation has a local plant, there is more to lose in case of political disruption, but the operation might also be seen in a more favorable light by the host government because of the jobs being provided. And what about the political risk for an overseas production operation that ships the products to markets in other countries?

A final note before moving on: Because of sovereign risk, the yield on a country’s sovereign bond denominated in local currency is not the currency’s risk-free rate. Instead a reasonable estimate of the currency’s risk-free rate is the sovereign yield in local currency minus the CDS yield. The reason is that an owner of a country’s local-currency sovereign bond who purchases a CDS has engineered a risk-free asset in the local currency. For example, if the yield on Mexican sovereign debt is Mexican pesos is 5.02% and the CDS yield is 1.02%, the Mexican peso risk-free rate is 4%.

Summary Action Points

Glossary

Business Beta: (a) The beta an all-equity operation would have if it uses no financial risk management strategies; (b) the beta of the operation’s business value.

Cost of Capital for a Business Operation: The rate that discounts the operation’s expected future business cash flow steam to the intrinsic business value.

Country Beta: The beta of a country’s national equity index versus the global market index.

Credit Default Swap (CDS): A market-traded “insurance policy” against the default of a bond. If the bond does not default, the buyer makes payments based on the CDS yield. If the bond defaults, the buyer delivers the bond to the seller, and the seller pays the buyer the bond’s face value.

Hurdle Rate: The rate of return that needs to be expected for an investment to add value. For developed market assets, the hurdle rate is in principle the asset’s cost of capital. For emerging market investments, the hurdle rate is the asset’s cost of capital plus and adjustment for political risk.

Political Risk: A catch-all term used to describe the additional risks posed by emerging market investments in terms of illiquidity, civil disruptions, corruption, political intervention, expropriation, and the like.

Political Risk Exposure: A foreign asset’s political risk relative to the foreign country’s overall political risk.

Political Risk Premium: An adjustment for the cost of capital to get the hurdle rate for an emerging market operation that has the average political risk in the emerging market country.

Sovereign Risk: The risk that a country’s government will not repay its obligations.

Unlever: To find a business variable (such as business beta) by removing the effect of financial strategies on an equity variable (such as equity beta).

Problems

For 10 to 12: Lincoln Resources Co. is a U.S. multinational company with a business value of $40 billion. Lincoln has actual net debt of $10 billion, all of which is denominated in US dollars. Lincoln has no off-balance sheet risk management positions with systematic risk. Lincoln’s estimated global equity beta in US dollars is 1.20. Lincoln’s management wants to set the hurdle rate for its operation in Kazakhstan. For the operation’s business beta, Lincoln will use the Lessard country beta method with Lincoln’s own business beta serving as the U.S. proxy’s business beta. The operation is in an industry with a relatively low political risk exposure, so Lincoln assumes the operation’s political risk exposure is 0.75. The Kazakhstan currency is the tenge . Assume: (1) In US dollars, the global beta of the Kazakhstan stock market is 1.70, and the global beta of the U.S. stock index is 0.94. (2) The Kazakhstan political risk premium is 2.50%. (3) The GCAPM is the model of risk and return in US dollars. (4) The US dollar risk-free rate is 3%. (5) The global risk premium is 6% in US dollars.

Answers to Problems

Discussion Questions