David J Sharp.

Performance measurement of foreign operations under floating exchange rates online

. (page 1 of 4)
Online LibraryDavid J SharpPerformance measurement of foreign operations under floating exchange rates → online text (page 1 of 4)
Font size
QR-code for this ebook




-7ss.*MST r,-^



David J Sharp

WP 1393 - 83







David J Sharp
WP 1393 - 83 DECEMBER 1982

Massachusetts Institute of Technology

I would like to record my thanks for the support and advice of my
advisers, Professors Donald Lessard and Richard Robinson; to Professor
Peter Brownell. the 'outside reader'; and for the helpful comments of
Professor H. David Sherman and of my fellow students of International
Management; Bruce Kogut's advice was especially helpful. I also record
my appreciation to Professor Vinod Bavishi, Director of the Center for
Transnational Accounting and Financial Research, University of
Connecticut, for generously providing data processing facilities and
making data available for thii. study.

Working papers are a series of manuscripts in their draft form. They
are not intended for circulation or distribution except as indicated by
the authors. For that reason, working papers may not be reproduced or
distributed without the written consent of the author.

I ....



'^ i^oj



a. To provide information that is generally compatible with
the economic effects of a rate change on an enterprise's cash
flows and equity

b. To present the consolidated financial statements of an
enterprise in conformity with US generally accepted accounting
principles. ..."

Statement of Financial Accounting Standards No. 52
Foreign Currency Translation, December 1981.



One of the problems facing the managers of corporations with overseas
subsidiaries which distinguish them from their uninational colleagues is
that a large proportion of corporate activity is undertaken within
different monetary jurisdictions; specifically, foreign subsidiaries
earn income denominated in different currencies. For the purpose of
preparing consolidated quarterly published accounting statements, as
required by the Securities and Exchange Commission (SEC) , rules have to
be devised to translate these foreign currency (FC) accounting
statements into local currency (LC) , i.e. US dollars.

In December 1981, the Financial Accounting Standards Board (FASB) issued
Statement of Financial Accounting Standards No. 52: Foreign Currency
Translation (FAS 52), which replaced an earlier statement FAS 8:
Accounting for the Translation of Foreign Currency Transactions and
Foreign Currency Financial Statements.

FAS 8 had been issued in 1976 in order to replace the variety of
previously acceptable methods of accounting for foreign income with one
uniform standard. The historic-rate translation method of FAS 8 treated
foreign income as if it had occurred in dollars. It also required that
all translation gains and losses be recorded in current income, thereby
producing net income figures which

i. were as uncontrollably volatile as exchange rates

ii. included large foreign currency gains and losses which bore no
relationship to so-called "real economic" effects of foreign
exchange movements.

As soon as the business community recognised the effects of FAS 8 on
reported earnings, pressure was brought to bear on FASB to revise the
standard; not so much, one suspects, because FAS 8 income was different
from "economic reality", but because managers had little control over
the huge effect of unpredictable exchange rates on reported earnings.
(To be sure, submissions from the business community were frequently
couched in terms of the lack of credibility which FAS 8 had brought upon
the accounting profession, and the unrealistic nature of net income
figures.) FAS 52 has therefore been welcomed by the business community
and the profession (1). We will see that FAS 8 generally produced
unbiased income numbers which had a high variance, while FAS 52 produces
relatively stable income figures, but at the expense of a bias created
by the interaction of historic cost accounting rules and differential
inflation between domestic and overseas currencies. The avowed intent
of FAS 52 to represent economic reality is likely to endow translated
dollar income figures with high credibility, and lead to their increased
use as a basis for management performance. For this reason, we evaluate
FAS 52 from a performance measurement and control perspective. It turns
out that the bias in net income is potentially large, .and since internal
control systems are already largely based on translation rules for
external accounting requirements, decisions concerning resource
allocation across national boundaries may be made incorrectly after the
adoption of FAS 52.

This paper is in four main sections. In the first, we review the
current understanding of the behavior of foreign exchange rates. We
show that exchange rates, inflation and interest rates are inseparably
linked in foreign exchange markets. In the light of this, the
literature on the accounting treatment of foreign exchange is reviewed,
and is found generally not to be consistent with the economic

The second section critically examines the practice of accounting for
foreign operations as required by FAS 52 and US generally accepted
accounting principles (GAAP). We do this by developing a normative
basis for performance measurement, against which FAS 52 is compared.
The long-run effects of FAS 52 are examined in the context of long-run
expectations in foreign exchange markets, and are found to cause a bias
to net income and return on investment (ROD prepared under the historic
cost (HC) convention of US GAAP and FAS 52. This bias was not present
under FAS 8. We show that an inflation-adjusted accounting standard
would remove this bias. The short run effects of FAS 52, while not
significantly different from those of FAS 8, are found to misrepresent
foreign exchange gains and losses. Again, we find that an appropriate
treatment is possible if accounting statements are inflation-adjusted.
The requirement of GAAP to state monetary assets and liabilities at
nominal value, not discounted net present value, is shown also to lead
to inconsistent treatment of otherwise identical transactions in
different currencies.

In the third section, we test whether control systems in use in US

multinationals are designed in such a way that they encourage
non-optimal decisions, and whether managers are aware of this. We find
that reward systems in a disturbingly high number of multinational
corporations potentially encourage non-optimal behavior, but there
appears to be some awareness of the problem.

The final section discusses the need for more objective reported
accounts, and the importance of the role of the accounting profession in
setting such standards.

FAS 8 was a historic-cost oriented standard; by translating at historic
exchange rates, foreign accounting statements were consistent with US
dollar statements produced under the historic cost convention. The
all-current translation method of FAS 52, however, is conceptually
inseparable from inflation-adjusted accounting, and cannot be expected
to produce sensible information otherwise.. FAS 52 has a few critics,
including three of the committee of seven who drafted it. (FAS 52,
p. 15). Their criticism (that FAS 52 is an unacceptably large departure
from the HC principles of GAAP) it seems to me, is misguided; it should
be directed instead at the much greater problem of the relevance of US
GAAP, and specifically historic cost accounting.



In this section, we review the literature on the behavior of exchange
rates in a floating-rate world such as has been the case since

Smithsonian parities were abandoned in 1973. V/e find the following
fundamental relationships:

i. In the long run, purchasing pov;er parity (a goods market model)
and the International Fisher Effect together suggest relationships
between differential inflation, diffrential interest rates, and
exchange rate movements.

ii. In the short run, exchange rates behave like financial assets
in an efficient market, with the result that expectations play the
dominant role in determining short-run exchange rate movements.

It would seem reasonable then that accounting rules for translating into
dollars economic activity which is measured in foreign currency should
be built on a foundation which recognises these fundamental

An exchange rate is simply the price of foreign currency in terms of
local currency. Like any other price, it is determined in a market
which may be more or less free of controls and imperfections. On the
one hand, the exchange rates of centrally planned economies are fixed
entirely by government fiat, while on the other, exchange rates between
the US dollar and, for example, major European currencies are determined
in large, well-organised markets free from controls, in which publicly
available information is rapidly disseminated, and which therefore
correspond closely to a semi-strong form efficient securities market
(2). We consider here only the efficient-market case, tnough we mention

briefly the important implications of inefficiency.

Purchasing Power Parity (PPP) has enjoyed a long and successful history
as the long-run determinant of exchange rates. Briefly, PPP states that
exchange rates tend to move in such a way that the relative purchasing
power of all currencies is unchanged; that is, nominal exchange rate
movements v;ill be in proportion to relative inflation.

The term was first coined by Cassel (1915), and the voluminous

subsequent literature is summarised in Balassa (1964) and Officer

(1976). They also point out that there are reasons why exchange rates

might not tend exactly towards PPP (owing to relative differences in

productivity in traded and non-traded sectors between the high-income

technologically-advanced and low-income countries, for example). Such

reservations notwithstanding, this paper takes Krugman's (1978) point of


"Few international economists would deny that purchasing power
parity holds in some sufficiently long run..."

The analysis that follows is based on the view that PPP represents an
'ideal' situation, and that deviations from PPP are temporary
aberrations which, although they are obviously short-term equilibrium
rates, nevertheless distort translated accounting statements, since
statements are translated at actual exchange rates, and not at some
hypothetical rate based on ex post differential inflation.

What are these short-term deviations? Krugman's next sentence:

"But most [international economistsj would probably be skeptical of

any assertion that PPP holds in the short run."

Since the ending of Smithsonian parities in February 1973, exchange
rates have been highly volatile, and more or less unpredictable. In
that respect, they behave, not as prices of goods and services, but as
prices of financial assets. The primary determinant of price movements
is therefore expectations, and as information arrives in essentially a
random manner, exchange rates should also move in a random manner.
Giddy and Dufey (1975) showed that the behavior of exchange rates
between a number of major currencies was not significantly different
from a random walk or martingale. Frenkel (1981) took this further
demonstrating the importance of information (i.e. news). He showed
that a strong relationship existed between unanticipated exchange rate
movements and unanticipated movements in interest rates ( a variable
which was expected to capture the effect of 'news' very quickly).

Dornbusch (1980) proposes an asset model of exchange rates which
elegantly explains that the high volatility of exchange rates arises
because changed expectations about future interest rates and inflation
cause the exchange rate immediately to overshoot the long-run
equilibrium rate, and only slowly to return to it.

However, there also exists a strong relationship betv;een the spot
exchange rate and the forward rate in the short run; tne difference
between these rates is determined by the powerful covered arbitrage
transaction between differential risk-free interest rates in different
currencies. The forward rate thus always represents that exchange rate

which can be guaranteed by hedging in forward markets. But is the
forward rate an unbiased predictor of the spot rate? Giddy and Dufey
above assumed that it was, and found that this submartingale model of
exchange rate behavior was also statistically significant, and in fact,
that the martingale and submartingale models produced virtually
identical results.

It is not clear that the forv/ard rate is an unbiased predictor of the
future spot rate, since holders of FC financial assets would require a
risk premium if the expected cash flows (i.e. exchange rates) were
correlated with security market movements (3). The literature on
theoretical reasons why such a risk premium should exist is thoroughly
reviewed by Henriksson and Lessard (1982). Empirically, Hansen and
Hodrick (1980) have demonstrated the existence of either market
inefficiency or a small risk premium, but in common with all studies of
this type, it is not possible to test separately the validity of the
model and market efficiency. A rather pragmatic test of market
efficiency is whether, ex post, so-called experts at forecasting were
able to outperform the market. Levich (1981, 1982), and Henriksson and
Lessard (1982) have both found that some forecasters can consistently
outperform the market over the period studied for some currencies, but
by and large, the major currency markets are indeed at least weak-form
efficient most of the time.

The link between the long-run and the short-run is made via the
International Fisher Effect (IFE), which states that the difference
between the current spot rate and the expected value of the future spot


rate equals the difference in interest rates. Giddy (1977) showed that
this relationship held quite well for one-year interest rates, but less
so for very short-term rates. Furthermore, the (domestic) Fisher Effect
states that the nominal rate of interest equals the real rate of
interest plus the rate of inflation. Fama (1975) showed that, at least
for the US dollar, short term interest rates incorporate inflation
expectations well. If capital markets are perfectly integrated,
expected real interest rates should be equal for all currencies.

To summarise so far, we see a relationship between inflation, exchange
rates and interest rates. In the long run, exchange rate movements tend
to follow FPP, but in the short run, they are highly volatile as
adjustments in expectations are incorporated simultaneously into
interest rates, and forv;ard and spot exchange rates. In the absence of
deviations from PPP, expected real interest rates in all currencies are
equal. Forward rates are probably the best estimates of the expected
future spot rates, albeit poor ones. Finally, covered interest
arbitrage ensures that the difference in nominal riskless interest rates
is exactly reflected in the difference between the forward and the spot

The literature on the application of this theory to accounting for
international operations is of two types. Those who write from a
financial background have concentrated on the relevance of hedging;
Logue and Oldfield (1977) for example clearly view hedging as a zero-NPV
transaction, and therefore irrelevant as far as the value of the firm is
concerned, though they also take the view that accounting performance


(at that tine under FAS 8) did not matter. Giddy (1976) clarifies the
distinction between hedging and speculation, but he does not discuss the
misleading accounting treatment of hedging. Other financial writers
discuss the appropriate measure of exposure; Dufey (1972) in a brief
descriptive but insightful article points out the importance of
cash-flow exposure but was unable to relate this to a specific standard
of accounting for exposure, since none existed at that time. Few
writers have specifically examined the impact of accounting practice on
foreign income measurement. Fredrikson's (1968) must surely rank as the
clearest analysis of the confusion created by accounting practice for
foreign exchange transaction losses, in spite of the fact that he too
had no accounting standard to which to refer; nor was he able to use
subsequent developments in the theory of accounting for inflation.
Aliber and Stickney (1973) present an empirical analysis of the validity
of PPP and the International Fisher Effect, but were not able to
incorporate their findings into specific implications for reported
accounting (again, a standard did not exist at the time).

Those writing from an accounting perspective on the other hand, have
frequently emphasised the effect of exchange rate movements without a
complete consideration of the behavior of those exchange rates.
Contrary to most of the polemic generated by FAS 8, Mueller (1976) took
the view that FAS 8 was no better or worse than any other standard, but
he also recognised that the real need was for an inflation-adjusted
standard. Shank (1976) pointed out that one result of FAS 3 would be
that reported earnings would be more volatile than had formerly been the
case (though Rodriguez (1977) was unable to verify this empirically).


Shank's view was confirmed when the parity of the dollar subsequently
fluctuated wildly against other (especially European) currencies. (See
for example Curran (1981) and Alleman (1932).) The inconsistent
accounting treatment of transaction gains and losses, and the similar
treatment of the results of the translation process (both of which are
discussed below) continue to divert many writers' attention from the
real issues; Prindl (1974), Evans, Folks and Jilling (1978),
Czechowicz, Choi and Bavishi (1982), The interrelationship of exchange
rates and inflation has been discussed in terms of the
"restate-translate vs. translate-restate controversy", but without
reference to exchange rate behavior, by Lorensen and Rosenfield (197^),
while Choi (1977) recognises that under an inflation-adjusted accounting
standard, the issue is trivial; which indeed it is, but only if PPP


This paper seeks to present a comprehensive critique of current

multinational accounting practice in the US. We now examine the
relevant accounting rules and principles .

1 . Background

Annual audited accounts are published by corporations in order to convey
information to shareholders; the main reasons for this are stewardship
and accountability. Audited statements are the means by which


shareholders are assured that directors (and by implication, all other
employees) are performing their fiduciary duties correctly. The
objective is to disclose specific information prescribed by a strictly
defined set of rules, US GAAP, which originally evolved in a one-country
context, but have had to be adapted to a multinational environment as
the need arose. Three well-known fundamental principles lie behind the
formal financial accounting rules: objectivity, consistency and
conservatism. There have always been problems of compromise between
these three and a fourth principle, that of relevance, but the advent of
multinational accounting has made the problem acute. In addition to
GAAP, FAS 52 is a specific statement (which purports to be consistent
with the more general guidelines of GAAP) of how to account for foreign
operations. Since 1971, the international economic environment has
become less stable, less predictable and arguably less well understood
while multinational corporations have spread their operations throughout
the world. Small wonder, then, that accounting for multinational
operations has experienced such growing pains.

A further problem confounding the measurement of results of foreign
operations is that the strategic view of overseas subsidiaries varies
from one corporation to another, and even between subsidiaries within
the same corporation. A subsidiary may be part of a tightly-controlled,
highly integrated worldwide system, or completely autonomous and
self-sufficient; at the same time, parent objectives might be to
maximise dividend repatriation, or to consolidate and build market
share, with no immediate expectation of dividends at all. These are
serious issues, and cannot lightly be dismissed, but they are also


applicable to large diversified domestic corporations. Our attention
here is focussed specifically on those problems which are currency-
related, and multinational in content.

A study of accounting-based control systems for use in US corporations
with overseas operations raises two problems which do not have exact
counterparts in the purely domestic case. The first problem, which is
the subject of detailed discussion in this paper, is that the reporting
system should correctly identify a cost item both by nature (an expense
labelled "foreign exchange loss" should exclusively be the result of
unforeseen exchange rate movements, for example) and by responsibility ,
which in its simplest form raises the centralisation / decentralisation
issue: if exposure management is to be a centralised treasury function
- the most common and most efficient organisation - then the speculative
(unforeseen) component of foreign exchange gain or loss has to be
attributed to the treasury function, whose responsibility it is to hedge
the company's overall foreign currency exposure against such unforeseen
losses. (It is true that the treasury may also be responsible for
taking speculative positions where they have better knowledge of, or
access to, imperfect or controlled markets.) The expected component of
exchange gain or loss, meanwhile, should be attributed to subsidiary
operating management, so that they are motivated to incorporate those
expectations into operating decisions. As will be seen, the
conventional accounting treatment of foreign exchange gains and losses
does not enable this important distinction between the expected and the
unexpected gain or loss to be made.


The second problem is more subtle, and is discussed only briefly here.
The fact that exchange rates and relative prices (among other variables)
do change means that the standard against which performance is judged
(i.e. a budget) should also be changed to reflect uncontrollable
changes in the business environment. An appropriate control system then
is one which measures actual against standard, given a particular
uncontrollable state of the world.

These two problems can be illustrated by an example. Suppose that the
currency of the foreign subsidiary depreciates by more than would be

required to maintain purchasing power parity. Such an exchange rate

movement is entirely plausible if, for example, a newly elected

government announced an expansionary monetary policy to reduce

unemployment and increase output. A result of problem 1 is that

exchange gains or losses may arise which are likely to be attributed

incorrectly to the subsidiary. The currency depreciation, outside the

control of the subsidiary manager, will certainly result in lower

translated dollar operating income, even if the subsidiary fully

incorporates local inflation into its prices. These are the issues we

discuss in this paper. A result of problem 2 is that since exchange

rates have moved in an unexpected way, the competitiveness of the

overseas (and, for that matter, domestic) operations has changed, so

that the assumptions upon which the budget was based have become invalid

(Lessard (1982)). Control systems then should not only separate

anticipated and surprise elements of exchange losses, and allocate them

correctly, but should also recognise that as a result of the

'over-depreciation' of foreign currency (in this exanple) , the


subsidiary has become more competitive in world markets, and the
overseas operating manager should be judged against a revised
standard . (U )

Even though published accounting information does not claim to represent
a true picture of the economic performance of an entity, nevertheless
the two are closely related. Furthermore, the accounting information is
prepared under a well-understood set of professional rules. It is only

1 3 4

Online LibraryDavid J SharpPerformance measurement of foreign operations under floating exchange rates → online text (page 1 of 4)