558 Part 6 Working Capital Management, Forecasting, and Multinational Financial Management
of these studies suggest that the United States does not have the lowest level of
country risk. This is particularly signi! cant because even though people in the
United States often assume that its bonds have no country risk, others do not agree.
Foreign investors are concerned about how changes in U.S. policies (e.g., tax or
Federal Reserve policies) might affect their investments. To the extent that these
perceptions about U.S. country risk in" uence investors’ willingness to hold U.S.
securities, they will have an effect on U.S. interest rates.SELF^ TEST List some key di# erences in capital budgeting as applied to foreign versus
domestic operations.
What are the relevant cash! ows for an international investment—the cash
! ows produced by the subsidiary in the country in which it operates or the
cash! ows in dollars that it sends to its parent company? Explain.
Why might the cost of capital for a foreign project di# er from that of an
equivalent domestic project? Could it be lower? Explain.
What adjustments might be made to the domestic cost of capital for a for-
eign investment due to exchange rate risk, political risk, and country risk?17-12 INTERNATIONAL CAPITAL STRUCTURES
Capital structures vary across countries. For example, the Organization for Eco-
nomic Cooperation and Development (OECD) recently reported that, on average,
Japanese! rms have 85% debt to total assets (in book value terms), German! rms
have 64%, and U.S.! rms have 55%. One problem, however, when interpreting
these numbers is that different countries often use different accounting conven-
tions with regard to (1) reporting assets on a historical-cost versus a replacement-
cost basis, (2) treating leased assets, (3) reporting pension plan liabilities, and
(4) capitalizing versus expensing R&D costs. These differences make it dif! cult to
compare capital structures.
A study by Raghuram Rajan and Luigi Zingales of the University of Chicago
attempted to control for different accounting practices. In their study, Rajan and
Zingales used a database that covers fewer! rms than the OECD but one that pro-
vides a more complete breakdown of balance sheet data. They concluded that dif-
ferences in accounting practices can explain much but not all of the cross-country
variations.
Rajan and Zingales’s results are summarized in Table 17-5. There are a number
of different ways to measure capital structure. One way is the average ratio of total
liabilities to total assets—this is similar to the measure used by the OECD, and it is
reported in Column 1. Based on this measure, German and Japanese! rms appear
to be more highly levered than U.S.! rms. However, if you look at Column 2,
where capital structure is measured by interest-bearing debt to total assets, it
appears that German! rms use less leverage than U.S. and Japanese! rms. What
explains this difference? Rajan and Zingales argue that much of this difference is
explained by the way German! rms account for pension liabilities. German! rms
generally include all pension liabilities (and their offsetting assets) on the balance
sheet, whereas! rms in other countries (including the United States) generally “net
out” pension assets and liabilities on their balance sheets. To see the importance of
this difference, consider a! rm with $10 million in liabilities (not including pension
liabilities) and $20 million in assets (not including pension assets). Assume that the
! rm has $10 million in pension liabilities that are fully funded by $10 million in