International Finance: Putting Theory Into Practice

706 CHAPTER 19. SETTING THE COST OF INTERNATIONAL CAPITAL

There seems to be two ways we could go about this, similar to what we did earlier
for risk-free cash flows. As we saw, a risk-free claim oninr1 can bePV’ed ininr
terms first, by discounting theinrcash flow (unity) at theinrrisk-free rate and
this value is then translated intoaudat the going spot rate. Alternatively, we can
translate the future cash flow intoaudusing the expected future spot rate, and
then discount at anaudrate that takes into account the risk. Both are linked via
the forward rate as the risk-adjusted expectation andCIP:

Et(S ̃T)

1 +rt,T+RPt,T

=

Ft,T

1 +rt,T

, (F=CEQ)

=

1

1 +r∗t,T

St. (CIP) (19.1)

Similarly then, in case of a riskyfccash flow, we could first translate thefutureinr
cash flows intoaudusing the expected future spot rate, and thenPVthese using
anauddiscount rate, sete.g. on the basis of the standard Capital Asset Pricing
Model (CAPM), the way Australians would value a domestic Australian project.
Alternatively, we could argue that the Australian ownership hardly matters, and
simply conduct the entire cost-benefit analysis ininr, the way an Indian owner
would do: takeinrcash flows, and discount at the rupee rate of return. Having
found the value ininr, we then translate thepresentvalue intoaud. And if that
second solution really works, exchange-rate forecasts and currency risk can be totally
eliminated from the analysis, it would seem.

In this chapter we show that the above analysis is quite incomplete. The main
lessons to be remembered from this chapter are the following:

• Translation ofFC cash flows requires more than just an expected
  exchange rateSuppose we follow the first route and translate our Rupee
  cash flows,C ̃T∗, intoaud. What we need are expectedaudcashflows; but the
  expectation of a product, E(C ̃T∗S ̃T) involves not just the expectations ofC ̃T∗
  andS ̃T, but also the covariance between the two.
  This, at first sight, makes the first route even more difficult. All the more
  reason to go for the alternative one, then? Unfortunately, this alternative
  would not always work:


• Host-currency v home-currency valuationValuation in Rupees, the way
  an Indian investor would do it—using the Rupee risk-free rate and a premium
  for market risk measured in Rupees—should produce the same result, after
  translation, as valuation`a l’Australienne only if the Indian and Australian
  capital markets are well integrated. Indeed, if investors from each country can
  freely invest in each other’s market (and possibly in other markets), arbitrage
  flows would occur if the value to Australians were different from the value to
  Indian investors (after translation into a common currency).
  In the case of India integration of the capital market into the mainstream work
  market is doubtful, for the time being. But even if it were true, the Indian-