CASE STUDY

New Plant for Houston Marine Electronics

International Cost of Capital and Capital Budgeting

Sitting behind his desk, Ben Nunnally told Reid Click, “We need to do an analysis of HME’s request for more plant capacity.” Their meeting was taking place in May 2013 in Nunnally’s office at Adventure & Recreation Technologies, Inc. (ART), in Boise, Idaho. Nunnally had been ART’s Chief Financial Officer (CFO) and Treasurer since before the company went public in the 1980s, and Click had been with the company for five years as an Assistant Treasurer.

ART is the parent company of three businesses: (1) Divemaster (scuba equipment); (2) Houston Marine Electronics (HME) (sonar equipment); and (3) Watercraft (canoes and kayaks). ART’s strategy across its three business segments is to use sophisticated research and cutting-edge technology to create the world’s best-known brands of outdoor recreational products.

The previous week, HME’s Chief Executive Officer (CEO), William “King” Kennedy, and CFO, Stafford Johnson, traveled to the ART headquarters in Boise and made a presentation to Nunnally and Click. The presentation outlined the increasing orders from Australia for HME’s sonar equipment and the shortage of production capacity to keep up with the growing demand. HME’s three plants in the United States were already producing at full capacity. HME had recently signed a two-year contract with the Australian boat builder Gold Coast Ships, and HME’s operations managers were already straining to stay on production schedule. In the meantime, HME’s head of sales, Steve Magee, reported that many other Australian and New Zealand boat builders had expressed a keen interest in putting HME’s sonar equipment in their boats. Simply put, Kennedy and Johnson were letting the ART corporate officers know that HME needed additional production capacity dedicated to the Australian market. The HME executives also knew that ART had accumulated about $60 million in cash, some of which could be used for the capital expenditure.

Nunnally continued talking to Click, “Magee says that HME’s sonar units will penetrate the Australian market quickly. He strongly suggests stabilizing the product price in Australian dollars, which means we’ll be looking at some FX risk because of uncertainty in the future $/A$ FX rate. With the pricing strategy Magee projects, we should get Australian revenues of A$10 million per year.”

Nunnally continued, “It seems clear that HME needs additional plant capacity for the Australian business. I expect that the numbers will support the strategy of overseas sales expansion as a sound business move. Still, I am asking you to prepare a formal capital budgeting analysis. One thing on the table is the plant location. I want you to consider two location options: (a) Australia; and (b) the United States.”

Click returned to his office and immediately began to think about the project. He understood the inputs needed for a net present value (NPV) analysis are: (a) the hurdle rate, based on the cost of capital; (b) an estimate of the expected operating cash flow stream; and (c) the investment outlay. Each of these inputs depended on the location of the plant. The international nature of the capital budgeting problem required that he also do some FX rate research.

Click had recently helped estimate ART’s overall cost of capital, successfully arguing that the modern integration of global financial markets implies that ART’s equity beta should be estimated relative to a global equity index (measured in US dollars) and not a U.S. stock index. Using historical monthly equity rates of return for 2007 to 2011, Click had estimated ART’s global equity beta is 1.28, compared to a local equity beta of 1.43 relative to the S&P 500 index. ART’s relatively high beta estimates are consistent with the notion that its high-end recreational products are “luxuries” that have a relatively high income elasticity of demand. (At the other end of the spectrum, companies that sell “necessities” such as food and basic apparel, with low income elasticity of demand, tend to have low betas, all else the same.)

During 2007 to 2011, ART had gradually paid off all its $60 million in debt and accumulated $60 million in cash and marketable securities. Click reasoned that because ART’s average net debt was 0 during the 2007 to 2011 period, he was comfortable with an estimate of 1.28 for ART’s overall operating beta. (Q1)

Australian Production Option

Click decided to first estimate the project’s NPV assuming production in Australia. Using the typical HME experience of operating costs of 75% of revenues, Click estimated annual operating costs of A$7.5 million. Therefore, he decided to base the NPV analysis on a perpetual expected operating cash flow stream of A$2.5 million per year.

Click knew that he could not just apply ART’s overall operating beta estimate of 1.28 for an operation in Australia. He needed a cost of capital, and thus an operating beta, specific to the operation. For the Australian production option, Click decided to use the Lessard country beta method, with ART as the home country proxy firm. With country beta estimates (in US dollars, versus the global equity index) of 1.16 for Australia and 0.94 for the United States, Click estimated a project operating beta of 1.58, given production in Australia. (Q2)

To find a cost of capital estimate in US dollars, Click used the global CAPM (GCAPM), assuming: (a) a global risk premium of 6%; and (b) a US dollar risk-free rate of 3%, based on the 30-year Treasury yield. He decided not to adjust for political risk because of Australia’s developed country status. Click estimated the operation’s cost of capital (and hurdle rate) is 12.5%, in US dollars, for the Australian production option. (Q3)

Click knew that he needed to convert the 12.5% cost of capital estimate from US dollars to Australian dollars. For the conversion, he would need estimates of (1) the project’s FX business exposure to the Australian dollar, from the US dollar perspective; (2) the volatility of $/A$ FX rate changes; and (3) the expected rate of intrinsic FX change of the US dollar versus the Australian dollar.

(1) Using a “what if” analysis, he estimated an FX business exposure to the Australian dollar of 1.40, inclusive of an economic “demand effect” of changes in the $/A$ FX rate. (2) He found an estimate of 12.7% for the volatility of the $/A$ FX rate. (3) Finally, based on a currency risk premium estimate of −1.38% for the US dollar versus the Australian dollar, an Australian dollar risk-free rate of 6.84% and a US dollar risk-free rate of 3%, Click estimated an expected rate of intrinsic FX change of 2.46% per year for the US dollar versus the Australian dollar. (Q4)

Putting all the inputs together, Click estimated a project hurdle rate in Australian dollars of 13%. (Q5)

A commercial real estate contact in Australia suggested to Click that a plant facility to make the sonar devices in Australia would likely require an outlay of A$18 million. Based on the assumption of a perpetual annual expected operating cash flow of A$2.5 million, Click estimated a project NPV in Australian dollars of A$1.23 million. (Q6)

Although the project’s NPV in Australian dollars was positive, Click believed that the more important NPV in cross-border valuation is from the perspective of ART’s home currency, the US dollar. At the time, the spot FX rate was 1.08 $/A$. Based on research supplied by ART’s bank and other sources, Click estimated the Australian dollar was overvalued by 8% versus the US dollar, implying an intrinsic time-0 spot FX rate of 1 $/A$.

Click assumed that Australian dollar would likely give up the 8% overvaluation gradually over the next four years and be correctly valued versus the US dollar from that time forward. He estimated an expected intrinsic rate of FX change of the Australian dollar versus the US dollar of −0.85% per year. At this rate of intrinsic FX change, the expected future intrinsic spot FX rates would be 0.992 $/A$ at time 1, 0.983 $/A$ at time 2, 0.975 $/A$ at time 3, and 0.966 $/A$ at time 4. (Q7)

To estimate FX forecasts consistent with the actual spot FX rate gradually dropping from 1.08 $/A$ at time 0 to 0.966 $/A$ as of time 4, Click’s actual spot FX rate forecasts were 1.05 $/A$ at time 1, 1.02 $/A$ at time 2, 0.99 $/A$ at time 3, and 0.996 $/A$ at time 4.

Click realized that if he converted the expected Australian dollar operating cash flow stream to US dollars using the actual spot FX rate forecasts, the result would be expected cash flows in US dollars that contained an FX windfall gain because of the Australian dollar’s overvaluation versus the US dollar, in addition to the business cash flows. He knew that at the intrinsic spot FX rate, the present value of the business cash flows had to be equivalent across the two currency perspectives. He would find the present value of the FX windfall portion of the cash flows separately, using as the discount rate the equilibrium required rate of return on a risk-free Australian dollar deposit, 5.99%. The present value of the expected FX windfalls in the future cash flows is $0.255m. (Q8)

Given the time-0 outlay of A$18m for the Australian plant, Click estimated an NPV in US dollars of $50,000. Click saw that the NPV in US dollars was relatively small, compared to the one in Australian dollars, because of the need to convert US dollars to Australian dollars for the plant outlay at time 0, at a spot FX rate that represented an overvalued Australian dollar versus the US dollar. (Q9)

United States Production Option

Click next wanted to find an NPV estimate in US dollars if production were in the United States. He understood that with U.S. production instead of Australian, there would be “some good news and some bad news.”

The “good news” had two parts: (1) the FX windfall would be higher, because only the revenues would be converted at the future FX rates that are projected to reflect an overvalued Australian dollar versus the US dollar; and (2) the US dollar outlay for the plant facility would not have to be converted to Australian dollars at the time-0 spot FX rate that reflected an overvalued Australian dollar versus the US dollar.

The “bad news” was that HME would have no operational hedging of the FX risk of the Australian revenue stream, and so the FX operating exposure to the Australian dollar would be higher. Given the Australian dollar’s estimated currency beta versus the US dollar of 0.53, a higher FX operating exposure implies a higher operating beta and thus a higher cost of capital.

Click understood that regardless of the plant location, some operating costs will be in Australia, in Australian dollars, including costs of selling, distribution, installation, service, and so forth. He estimated these costs to be 10% of revenues. Therefore, he expected an annual net amount in Australian dollars of A$9m. When Click converted this stream to US dollars, he found that an annual amount $9m was close enough to use in the analysis. (Q10)

The rest of the operating costs would be incurred in the United States, in US dollars. Of course, there would be some shipping and other exporting expenses, but he decided to use the standard estimate for total operating costs as 75% of revenues, which implied U.S. operating costs of 65% of revenues.

Click next estimated a project cost of capital in US dollars, assuming manufacturing in the United States. Click estimated an operating beta of 3.17 in US dollars, given U.S. production, which implied a cost of capital (and hurdle rate) of 22%, in US dollars. (Q11)

Click’s analysis showed that for an outlay of $18m for the U.S. facility, the U.S. production option’s NPV is −$5.7m. Click’s analysis thus showed that the “bad news” (about the lower operational hedging) outweighed the “good news” (about the higher FX windfalls), and the Australian production option was clearly better than the U.S one. (Q12)

Click was prepared to recommend the Australian production option, based on the higher NPV in US dollars. But he leaned back in his chair and began to think about other considerations in the decision. (Q13)

Questions