ON RISK, RATIONALITY AND EXCESSIVE SPECULATION IN THE DEUTSCHMARK‐US DOLLAR EXCHANGE MARKET: SOME EVIDENCE USING SURVEY DATA†
| Author | R. MacDonald,T. S. Torrance |
| DOI | http://doi.org/10.1111/j.1468-0084.1988.mp50002001.x |
| Date | 01 May 1988 |
| Published date | 01 May 1988 |
OXFORD BULLETIN
of
ECONOMICS and STATISTICS
OXFORD BULLETIN OF ECONOMICS AND STATISTICS, 50,2(1988>
0305-9049 $3.00
ON RISK, RATIONALITY AND EXCESSIVE
SPECULATION IN THE DEUTSCHMARK-US
DOLLAR EXCHANGE MARKET: SOME
EVIDENCE USING SURVEY DATAt
R. MacDonald and T S. Torrance
One of the most widely tested hypotheses in the international economics
literature is the efficient market hypothesis (EMH). The efficient market
hypothesis is normally taken to be the joint hypothesis that agents form their
expectations rationally and are risk neutral. Researchers who have imple-
mented the EMH for the forward market for foreign exchange1 have gener-
ally rejected the joint hypothesis and concluded that risk premia (of either a
constant or time varying variety) account for its failure;2 researchers have
been loath to jettison the rational expectations leg of the joint hyphothesis. If
risk is the reason for the failure, then this has important theoretical and policy
implications. Thus the existence of such risk implies that bonds denominated
in different currencies will be regarded as imperfect substitutes by inter-
national portfolio holders and, therefore, the portfolio balance model3 of the
determination of the exchange rate will be the correct one from a theoretical
perspective. If bonds are regarded as imperfect substitutes by investors then
this has the implication that sterilized foreign exchange market intervention
can affect the exchange rate. However, the vast majority of previous tests of
lWe are indebted to the editors for their helpful comments on an earlier version of this paper.
The usual disclaimer applies. The software package used was Regression Analysis of Time
Series. MacDonald acknowledges financial assistance from the Scottish Economic Society.
See, inter alia, Bilson (1981), Fama (1984), Hansen and Hodrick (1981), Hodrick and
Srivistava(1984) and Domowitz and Hakkio (1985).
2See, for example, Fama(1984) and Hansen and Hodrick (1983).
For a discussion of the portfolio balance model see Frankel (1983).
107
Volume 50 May1988 No.2
108 BULLETIN
the joint hypothesis have in a sense prejudged the issue in that they have
assumed that foreign exchange market participants do not make systematic
forecasting errors, and thus any systematic errors observed in the data must
be reflecting risk. But before we can conclude that systematic errors in the
data are due to risk factors, we should ideally test that the systematic forecast
errors are zero. The recent availability of survey exchange rate data in the US
and UK allows researchers to do just that. In this paper we examine some of
the properties of the survey data produced by Money Market Services
(MMS) UK for the Deutschmark-US Dollar (DM/S) exchange rate.
The outline of the remainder of this paper is as follows. In Section I a
framework for testing for the efficiency of the forward foreign exchange
market is presented along with a brief resume of the extant empirical
evidence. In Section lIthe data used in the study are described and some
descriptive statistics presented. Some regression results using our survey data
are reported in Section III. The paper closes with a concluding section.
I. SOME ANALYSIS
At the heart of most recent theoretical models of exchange rate determina-
tion (see e.g. Dornbusch 1976 and Buiter and Miller 1981) is the uncovered
interest parity (UIP) condition:
S,e+,_ S,= As1= i+1- i,+1 (1)
Where s donates the logarithm of the spot exchange rate (home currency per
unit of foreign currency), s represents the expectation of the exchange rate,
formed in period t, for period t+ j, A denotes a first difference operator, i1+1
denotes the interest rate on a bond purchased in period twhich matures in
t+ j and an asterisk denotes a foreign magnitude. Condition 1 will hold if
investors regard the domestic and foreign assets underlying j and as perfect
substitutes. As we mentioned earlier, if condition 1 holds central banks can-
not use sterilized intervention to affect the exchange rate.4 Further, if 1 holds
then investors will be indifferent between holding open positions in foreign
assets and selling the assets forward. Equation 1 therefore implies that the
expected change in the exchange rate should be equivalent to the forward
premium on foreign exchange. ftJ - As±1 (2)
where f' is the forward rate for delivery in period t+j and f' - s has the
interpretation of the forward premium, fp'. If, however, agents are risk
averse they will have to be compensated by a risk premium in order to
4i speaking if UIP holds only non-sterilized intervention - intervention that affects
the money base - can affect the exchange rate. However, sterilized intervention may work with
UIP if it can somehow alter market participants expectations about future monetary policy -
see Genberg (198 1(and Kearney and MacDonald (1986).
To continue reading
Request your trial