Note on Interest Parity, Eurocurrencies and Capital Controls

a ne e

° February, 1976

‘NOTE ON INTEREST PARITY, EUROCURRENCIES AND CAPITAL CONTROLS

Michael P. ‘Dooley

a NOTE: International Finance Discussion Papers are preliminary materials circulated to stimulate discussion and critical conment. References in publications to International Finance Dis~. cussion Papers (other than an acknowledgement by a writer that he has had access to unpublished material) should be cleared with the author or authors.

r) . . : saeteneterevagpeptence sss apg. yar pgcslei'n pte. yanignynaapineicsspegenstmapttaengpLEngn tap eprtie teste

cr ee EN: cer HONS RESIDE RRA a: gg atin en Aintree at RNA REO CEE YATE RR: ne LURIRGO. Hig AUURRRL PEHLE SE Cen oe ATROPINE Ww ees Ae

( \.

—~

’

- . wo = . .

Note on Interest Parity, Eurocurrencies and Capital Controls

. by ‘Michael P. Dooley* 7. - 7 . |

In a recent article Robert Aliber [19737 suggested a reinter-

pretation of the interest parity theory. Aliber's reinterpretation

rests on the distinction between political nish and exchange risk as

_ determinants of forward exchange rates.~” Se Several ‘ studies (liber, 1973; ' pooley, 1974; Marston, 1974) have shown that. interest rates paid on Eurocurrency deposits, which are identical in terms of political risk, are offset by forward exchange prenia or discounts so that covered differ-_ entials among LEurocurrency deposits are essentially zero. The objective

of this paper is to analyze a corollary of this. ‘Finding. That is, if

interest rates paid on Eurocurrency deposits are , different from deposit -o s

rates paid by banks located in the country of issue, then covered.

| differentials do exist between deposit rates in national money markets. These covered differentials are of interest because under certain assump-

tions they provide a measure of private speculative expectations. In this

| paper the. well known argument that speculation influences forward exchange 7 -_ .

premia is recast in terms that are consistent both with Aliber's reinter-

pretation of the interest parity theory and with current institutional

arrangements in international credit and foreign exchange markets. We

x, The views expressed herein are solely. those of the author and do not

necessarily represent the views of the Federal Reserve System. I am indebted to Peter Clark and Frank McCormick for helpful suggestions.

1/ Political risk is defined by Aliber/1973/ as "the probability that the authority of the state will be interposed between investors in one country and investment opportunities in other countries. "

ON a) . . . 7 . af

then test the usefulness of this formulation in evaluating private

speculative expectations for the German mark relative to the U.S.

dollar during the time period January 1970-December 1975 a I. Covered Interest Rate Differentials as Indicators of Private - ‘Speculative Expectations. — .:

The hypothesized relationship between private speculative

-expectations and covered interest rate differentials is well known

and is only briefly recapitulated nere.=! Those who wish to speculate

on exchange rate changes | can avoid the political risk. of holdi ing claims or liabilities of residents of a’ particular country and can - specialize in bearing the risk of exchange rate changes by dealing in the forward

exchange market. The other speculator on a net basis, the central bank,

typically chooses to take positions only in the spot foreign exchange

~ market, possibly because it is less concerned about political risks: The

‘spot and forward exchange markets are linked by covered interest arbitrage.

If private speculators have expectations concerning future spot exchange

‘.yates different from those held by central banks, the spot and forward

rate will be bid toward different expected values. The,resulting tendency

for spot and forward exchange rates te be bid apart induces arbitragers to

3/

take increasingly large arbitrage positions. The effect of these changes. on the stock of arbitrage positions

‘on forward premia depends on the elasticity of the arbitrage schedule. :

, 0 e

2/ See for example: Tsaing /T959/, Stein /i9627, Kesselman /1971/.

3/ This analysis is equally valid under floating exchange rates if central banks continue to intervene primarily in the spot market. The stock of arbitrage positions still depends on the divergence of central bank and

y vivate expectations.

‘ee

aimee gener ERM

If the arbitrage schedule is not perfectly elastic with respect to

covered interest differentials because of political. risk associated with securities issued in a given country then changes in the stock of

arbitrage positions are associated with changes in the equi sbrton

"levels of covered interest ‘differentials. It. would’ seem that, if.

covered interest differentials are observed, it follows that the | arbitrage schedule is not perfectly. elastic, and that the nature of. private expectations can be inferred from the divergences of the £ forward exchange rate from interest parity. Since private expectations cannot

be directly measured it is important to know whether or not easily

measurable departures from interest parity.are reliable indicators of

.market expectations, or, more precisely, of the differences in expectations

- between private and official market participants.

preg cent apna teat, TE seeentase coptep scammer simee tage 90 AER ang nem S “et SOA an! geben.“ typelgeyn cae vantotevingeerranys ticagein alyelene ae pangs: copes n pete.cip: ears amen pads, oe oe

II. Arbitrage -Among National Money Markets and Within ‘the Eurocurrency ’ Markets. .

The starting point for an. empirical study of arbitrage is the

identification of the yield on similar financial assets denominated in

various currencies. We will concentrate on interbank time deposits and interbank loans at "name" commercial banks. These particular financial assets. are useful for a study of the arbitrage function since the "name" commercial banks which issue such deposits are nearly homogeneous, except for country of residence, and because such interbank deposits and

loans are known to be actively traded internationally.

“Lae “In the case study presented in the next ‘section we consider the relationship among Eurodol lar deposits in Zurich, Euromark deposits in Zurich and interbank loans in Frankfurt. We know that forward Cx-

change | ‘premia conform ‘closely to the Eurocurrency deposit yield differ-

—-

entials and much less closely § to yield differentials among financial

assets “such as | Treasury bills or bank deposits hich are denominated in

the currency of the country in which che issuer is ‘located.

The absence of covered differentials among Euro-currency

Toe eb Se ‘

deposit rates: is not difficult to understand. Itc does not seem likely

that a s depositor + would hold a Eurodol lar deposit ata Swiss bank when

_he could hold a Euromark < deposit at that bank, sell | the marks forward

for dollars (perhaps to the same bank) , and receive a “higher yield on

what is essentially a. dol1ar-denoninated Position. It seems clear that:

‘this “would cause at least ‘the forward rates quoted by a given bank ‘to ; 5/ :.. be consistent with Eurocurrency rates quoted by that bank.

zs -"" “The fact that interest arbitrage is nearly perfect among Eurocurrency deposits does not imply that interest arbitrage among national financial markets is nearly perfect. As shown in our case study below - ; there have been times when there have been significant differences between yields on deposits at banks located within the country of issue as

_compared to deposits denominated in the same currency at Evrobanks. These

4S/ The linkages between the Eurodollar market and U.S. money markets are not crucial for this exercise.

3/ A full specification of a stock equilibrium model is not necessary for the very simple hypothesis presented in this paper. The description of equilibrium conditions is not meant to imply causal relationships.

. e

EE TE EE I, TE Te LTE OEE LS GI CELE LEE: COLE CLL LO TE TICLE LLY ACCC CL OT EY EE ARES CR RRR nt sma ye one eames Oa meme aem ci on

Me see!

differentials are equivalent £8 to departures from from interest parity since they directly measure different yields on assets with the same exchange rate risk but different political risks. Why, for example, have depositors accepted a significantly lower yicld on Euromark deposits as compared with mark denominated deposits cf German banks? oo 7 One possible answer can be couched in terms of the traditional " model of interest arbitrage. That is, speculative bids for forward marks, coupled by central bank sales of spot marks, tended to open a covered differential in favor of mark-denominated deposits at German banks. Arbitragers purchased marks spot (from the Bundesbank) , invested the mark balances in German bank deposits, and sold the marks. forward (to private speculators). “As the stock of these arbitrage positions ‘grew abi tragers mignt have become increasingly uneasy since a larger a share of their portfolios was subject ‘to pol litical ‘risk peculiar to 7 élaims on German residents. An arbitrager could diversify his political risk by purchasing wark-denoninated claims on non-German banks, i. Cos Euromark deposits. ‘But while this serves to diversify ‘the arbitrager's political risk Eurobanks face the same risks that arbitragers are trying to avoid. Upon issuing a Euromark deposit the Eurobank must cover its mark liability to the depositor. To avoid taking an uncovered position in marks, the Eurobanks, as a group, then mst do the same thing that “ee

arbitragers are attempting to avoid, namely purchase claims on German

is

° residents in order to cover their Euromark deposit liabilities. At some

on

point the Eurobanks themselves might have become uneasy about the political

or Bewee we ae

- . - 6 ~

risk accompanying their stock of claims on German residents. Eurobanks can discourage arbitragers from further shifting of this political risk to , Eurobanks by offering a lower yield on Euromark deposits. The difference between Euromark rates and the interest rates available on claims on German residents then might have been due to the “reluctance of arbitragers, including Eurobanks,, to acquire a larger stock of covered claims on German residents. An alternative, and simpler, explanation of the differences in yield between Euromark deposits and similar ceposits in German | resident banks is that legal restrictions designed to inhibit capital inflows placed a tariff on the export of financial claims on German. ‘restdents.~’ This control program drove a variable wedge between the. price (interest rate). at which such claims were traded inside. and ‘out~ side Germany: and was itself ¢ a response to private speculative demand for mark denominated claims. ‘Once ‘ ‘in place, exchange controls did not : affect the political risk faced by arbitragers but rather the certain “effective yield on arbitrage positions. Interest rates available to German residents then were not effective rates for arbitragers, and their use in _ computing an interest differential introduces errors in

variables.

6/ The importance of controls has been suggested but nottested in almost all treatments of the interest parity theory. See Officer and Willet £1970, p. 25...

3

eee ot we

sr

ase

“o,°

“ls

we

a“ as

Weid0ag

yoru

’

99

Wud

+ 1909 WO3 -

"=. WOR MENIWO

ry "s

With this in mind, we can reformulate the hypothesis about | -

‘the relationship between speculation and covered differentials. First, the absence’ of covered differentials within Eurocurrency markets does not imply that arbitrage among national money markets is perfectly elastic. ‘Second, the proper measure of covered differentials is the difference between Eurocurrency deposit interest rates and interest rates © ‘available on similar deposits at banks located in the country of issue. Finally, we can ask whether or not interest differentials between

| ‘ Eurocurrency deposits and domestic deposits denominated in the same

currency are caused by changes in the stock of ciaims subject to polit-

_ ical risk or to legal restrictions on capital movements.

. ITI. ‘Case Study of German mark and U.S. dollar January 1970-December 1975

In this section an attempt is nade to empirically assess the Amportance of various legal restrictions imposed by German authorities .

ro oe between January 1970 and December 1974. ‘In the upper panel of Chart Ir | | the end of period interest differential between German interbank a deposit rates and Euromark deposit rates and the covered differential between Eurodollar and Euro DM depesits for the time period January 1970 through January 1974 are presented. In the lower panel of Chart I an attempt has been made to chart the legal restrictions faced by non-resident arbitragers in acquiring claims on German residents. “The effect of this program was to driye a wedge between interest rates paid to residents

and non-residents and, thus, directly to influence the effective yield

Ow remem ee oe

tae - . eet war eevtaemanpen spat rnibtiAiin + tetas ype yee, OAR Tag UNM eee aE «tee ape egenry: ramet’: cerpwag: saprorges TE emgphe eyagie se nip. ee! byte Sees meTne Orme ~%, nama NplgNee “Toe Neha. 1: Gangyaaig Aa Rty gtiang Wee iste Mah ometecarety® -if;~mitSNEI.¢ sume pata <qmmagea ememtigtain aipsaign'yr+ ipo. Pcire.cunpzasmpantgms wadinetgmae.csagas das waa; ma tae yams 6 eet atte epnsest.sen van

-8= .¢

that non-resident arbitragers could earn on additional claims on German ‘residents. It is impossible to exactly quantify the effectiveness of this program but two qualitative measures of effectiveness | of controls are suggested here. First, at various ‘times, a given type of control was made more stringent. For example, the differential re-

serve requirements on bank liabilities to non-residents, as well as

- reserve free bases of such liabilities, were adjusted several times ‘during the time period. Second, controls on different financial |

' Intermediation channels were imposéd. For example, when controls on

German banks became, less effective, short-term borrowing by German non-

" - financial corporations was , penalized by the Bardepot which placed re-

serve requirements on nonfinancial corporations’ _Uabilittes to non-

residents. As the effectiveness of this program eroded, non-residents

were prohibited from acquizing, fixed interest securities. Finally, in.

a February 1973 acquisition by non-residents of virtually all claims” on

German residents was prohibited. The capital controls considered here applied only ito changes

in claims on German residents during the time period studied. In this

* ease the contrel program was effective only when desired positions ex-

ceeded some initial base level plus changes not covered by the control program. After substantial capital inflows ‘through the 3rd quarter of 1973, in October and November 1973 capital outflows from Germany totaled

about DM 3.3 billion and DM 3.8 billion respectively. Therefore, even

- though the control program was not formally removed until early 7 tn 1974 it’ apparently became ineffective during the last quarter of 1973. A detailed explanation of the controls imposed during the period is given in the ‘notes to Chart I. es thé relationship between these measures and the observed interest differentials is certainly striking, if not conclusive. This relationship suggests that the observed covered interest diffe ‘entials were not produced primarily by movements along a less than perfectly clastic arbterage schedule but rather werd. die in large’ part to ‘shifts in | the schedule caused by the ‘program designed to

change the ‘effective interest rate that could be earned on 1 addition

a clains on German residents. ‘Capital control measures “thenselves may

ene

rte

al

have dees correlated with the intensity, of private speculative activity,

and one 2 might suspect ‘that the correlation; if any, between private

speculation. and covered interest differentials. at ottier times for o

oe

currencies is due to this same factor.

Unfortunately, since it is not possible to precisely quan

the control program, it is simply not possible to di fferentiate the

ther

tify.

effect of private speculation per se and the effect of the Gemnan con~

trol program on forward exchange premia during this time period. W cannot, therefore, reject the hypothesis that the covered different observed were due to movements along a less than perfectly elastic cose ac , arbitrage schedule, But we can warn that any inferences drawn from covered interest differentials for the speci fication and estimation

theoretical models, are valid only if the dual nature of covered

Toa achapabalanic Ste aad ene aaa RRR Ri meemcaaieiaes ne oh-adedenetanamnanetameantenan ueenidhamincetadanmmemmmininmmmanaeemnatntEeiesaummeenenaees anamennmnnenenclernttaeeteemamentiee eee cE

e

ials

of

> eR REE ~ “enemas

interest differentials is considered. : wee

In summary, there seems to be no escape from the conclu- ; ; . . . > sion that detailed examination of capital control programs is a ‘0 necessary, if unattractive, aspect of our understanding of interest.

rate and exchange rate determination. We have shown that control’

programs do not simply introduce random errors into calculations of interest parity but are instead correlated with an important variable in all models of international financial markets, i.e., differences -

in exchange rate expectations between private and official market

‘participants.

wale epsee: ees

° ATONE te ARIA ERAN RR A YAR UE PORE A eee OCT CAA EAA RY ET PE EN Ee ee Reh Min mmm,

‘Chart I -- Notes and Sources Sms

Upper ‘Panel

DAINT = 3-month Frankfort ‘interbank loan rate at or near . "end of month WFM ° mL: ..

“EDM = ‘Semonth Euro DM Deposit, at or near end of wonth . WT, SBC oo

es -

_. , ,EDOL = 3-month Eurodollar deposit;. at or near end of.

Hd B n

3-month sonar’ prentun, at or near end of month, -

wes

"pourees LET = = London Financial Times oo. -. SBC = Swiss Bank Corporation, internal records | WEM = World Financial Markets, } Morgan Guarantee Trust Company New York

Lower’ Panel -% On April 1, 1970 the Bundesbank reintroduced a spectal reserve ratio: on the growth of banks’ liabilities to non-resident. With the exception of a four month period, September through December 1971, when liabilities of both: residents and non-residents carried equal special reserve ratios, some differential reserve. ratios were imposed on bank liabilities to nonresidents throughout the time period. This program served two purposes. . First, it induced German banks to pay lower deposit rates to non-residents as compared to residents. This effect of the program probably was less important after May 1971 when payment of interest on deposits held by

non-residents was made subject to prior approval by the Bundesbank

srs shdiepaitr aaa acibdeetheaierdaatineman nak atnanintcl aereeataensenaeaietteaenaaat ies adatadetemmea neti sacaca otiieeme tasapiaatetie cic ate koran eee ore eee eR ee

4 . - although they still served to (da scourage implicit payments for deposits. ' Secondly, the reserve requirement absorbed reserves and thereby | - "sterilized" the increase in the monetary base resulting from bank reee ported capital inflows. The percentages shown in Chart I are the . sO approximate percentage differences between average plus marginal reserve ‘ratios on liabilities to residents and non-residents. a a . TI. On May 10, 1971 interest ‘Payments on non-resident. bank deposits eXx- . ceeding DM 50, 000 were > made subject to” prick approval by the Bundesbank. iit. On March 1, 1972 the Federal Coveriment introduced a “eash deposit requirement (ardepot) of 40% on most types of credits of non-residents _ to Gexman non-barks in excess of DM 2 million j per individual, | The cash deposit was a non-interest bearing deposit at the. Bundesbank. The op | . cash deposit was increased to 50% on july 29, 1972 and the exempt amount ~. | was reduced to DM 0.5 million. On January 30, 1974 the cash deposit * requirement was reduced to 20% and the exemption raised to DM 2 mitiion. ‘On September 11, 1974. the cash requirement was eliminated. | IV. On June 29, 1972 the Federal Government decreed that the ‘Purchase ’ of fixed-interest securities by non-residents was subject: to prior authorization. Fixed-interest securities included all maturities of * bonds including, for example, all bank bonds, mortgage bonds, communal bonds, industrial bonds, and public authority bonds. The authorization . ' requirement was, in practice, equivalent to prohibition of such purchases. The authorization requirement for all but short term securities (less

than four years to maturity) was terminated on January 30, 1974.

SON TN ES AT REET ARE RR RR TOT TE LTC CT CEE! RT TREE hE RE I SA Eee eR tee ee i ennenyeepepr eg [ce

< ° ° 2 - 2° 1W- . oa V. On February 5, 1973 the Federal Government made acquisition of domestic shares and mutual funds by. non-residents and the raising of © loans abroad by residents, including trade credits, subject to" _ authorization. This measure was designed to prohibit essentially . all. capital transactions with non-residents. All of these additional measures were terminated om January 30; 1974. 9 --- ttt a _ Source: Various issues of Monthly Report of the’ Deutsche Bundesbank. el .deadechersd as dn - > : te - , . Versdion Lo DMg ae beastie aa das. . . . qeeat : aD lta . so. ofse ws ON . ° c

C Lo - © 14-

a References

as Aliber, Robert Z. “The Interest Rate Parity Theorem: A Reinterpreta- ‘thons" J.P.E. 81, No. 6 (1973) : 1451- 59. oe -

Dooley, Michael P. "A Model of Arbitrage and. Short-term, Capital Flows."international Finance Discussion Papers, No. 40, Board of Governors of

the Federal Reserve System, 1974.

Herring, Richard J. and Marston, Richard Cc. "he Forward Market and . Interest Rates in the Eurocurrency and National Money Markets." Rodney | L. White Center for Financial Research: Working Paper, N . 17- 76, a = | University of Pennsylvania, 1974.

Kesselman, Jonathan "The Role of Speculation in Forward-Rate Determination: ‘The Canadian Flexible Dollar 1953-1960." Canadian Journal of

Economics 4, No. 3 (1971) : 279- 299.

" Officer, Lawrence H., and Willet, Thomas D. "The Covered-Arbitrage \ Schedule: A Critical Survey of Recent Developments." Journal of Money,

Credit and Banking 2, No. 2 (May,. 1970), 247-257.

Sohmen, Egon "The Theory of Forward Exchange" Studies in International

‘| Finance. - No. 17, Princeton University, August, 1966.

eo4 €

Stein, Jerome L. "The Nature and Efficiency of the Foreign Exchange Market.”

Essays in International Finance. No. 40, Princeton University » October, 1962.