ifdp · June 30, 1974

Empirical Research on Financial Capital Flows

1 me ee em

+ pee

au iews

INTE

wyee SRA

Discussion

we tse

a“ wove

2 8.

Teese | (#723

LI ONAL FIRADCE DISCUSS 10H PAPERS . LIP IRICAL RESEARCH CH FINANCLAL CAPITAL FLOSS . . . . js f] . by ‘ . 4 oe Ralph C. Bryant . . e oe ‘ ; Faper No. 50, July 1, 1974 -- “a Of International Vinance . evs of the Federal Reserve Syston . ° “

the views System or cireculeat: comment

Sjon of the av

ens of this paper re

rnors ef the Federal a pamers in many ca ‘m to stimulate disc

a Ss 9 nm uA he 46)

present the

si eee sould not be interpreted as reflecting re

Reserve ses ere

2 ussion and neo cited or quoted without the parmis-

on . . shee . 00 cme Semews vow 0. ener nero: —_— . . . . . . e ° . . oe . ¢ )

sae ow ot

~ oo 8

This paper 15

a *¢erm "finen jai"

. £ows) is dixect investments. direct investments “ar

the Nufbauer peper and “The scope of

‘yell. There have been @ n . theonetical in neture, models of snternational tr

- Jiterature is discus have not attempted

gixrect bearing on cr piri paper will show, of it pre-dat published

- relatively underdeveloped

account in the balance of payments.

of real activity and prices on the ceux

payments have

. v/ Jt will : Trade anu University Tress volume"

appe ear as a J Financed’ \.

oe eee

reporting on cenpirical research on in capital {lows (or, used in the paper to comote capita

rd the multingetional co

this pei

Readers anterested primarily

sts references eisewnere im

weber of articles in xe

dealing with what

a a

to cover

. As a glance at the 1

es tha decade of the ithin the last five or six years.

state of econ

a much Longer and tm

chapter Yrenziezs

wo Wererenc es

are ta ether chaprers

wight be

: : - ‘ de and capital flovs.

sed in the Hells. sell peper 2%

this ground oxcept ins

research rex se.

ternational capital flows.

‘ rporation

per is vestricted in

of .

concerned with the growing body of literature

/ The

alternatively, portfolio capital °

1 flows other than

Cad

in the voscareh on shoule consult

wt o- Cir1s yoL nee

enother way @§

cent years, primarily

rere “ Cornnacu

macreecononLe

Same On chis

st of this Literature is of re

sec

1950's; the +

Eoononetric

in Peter B. Neren in to

ist of reference ty aL

This fact

cmetric research

rent account

the

sent vintegc. Little £ it eas

Sor sieb oe

underlines the

the capital

)

shed of the impact

. Gies

sn the telance of

ore sophisticated wistory oo not to

Trtemmahic oni!

Cambri ine "ehis

(ed. y. Tnter

Yas theomias ve other papers 10

re asad

se Bee toe?

aebh fe

wer

fete

. 4 . ’ e - 2 - . J . \

mention the great volume of empirical research carried out on

* ‘ . . domestic financial markets, domestic expenditures on real goods and

services, and the linkages between them. oo - -.* Z should caution the reader with one further introductory

coment, This paper is not, at least in some senscs of the vord,

- a survey of the empirical research on financial capital flows; I

have not tried in most cases actually to susmarjze the Literature, Yhe paper is most accurately described, not as a survey or a sumiary,

but as a selective assessment of empirical knowledge and of the state

of the art in research in this ficld,: - Before plunging into a discussion of the strengths and

weaknesses of the work done in this area, it seens useful to begin

by providing a bricf£ road map of the literature, .That is the purpose

of the following section.

. . . . ° . . . . bd . ca . ° - . : : e bd . : . ve . . . < eos . . . . . . : . . . . . e . . . . - ¢ ° é . . . . e . . . . . bd . ° . ° ° . . . r) . . - . ee « . . . ° oe . « - e . . . . . e . . . ° . . . . wet” . e . . . ° 7) . e s . . - . . . . . . . . . . °

te ’ ets oo t. A icf Men of the Literature eT , -

A_Yo. rap_of the ’

One method of mapping the Literature is to classify it by the main country of interest and the extent of coverage of that

‘country's capital account, The pulk of the research has been done

on the United States and on Canada. Bilateral capital flows. between

: the United States and Canada have 21so received considerable attention jn their orn right. Recently, morte sntensive empirical work has

begun on capital flows ina the balance of paymoants of European countries

end Japan. Empirical research on the Eurocurrency markets, although

‘showing promising, signs of growth, 35 very much sn its infancy.

o Lod . 4 so The first systematic econometric work on the v.c. capital . . aecount was begun in the: firs& past of the Gecade of the 1960's.

Tae growth of interest in capital flows seflectcd in part a rising ° concern about weakness in the U.S. balance of payments end instability

- in the snternationel monetary system. The growing snterest also

coincided with the development by aceadomic economists of a literature

on the so-called Nassi.gmnent” problen, sn much of this literature, °

ht was suggested whet monetary policy should be directed at the

: an - “balance of payments, while fiscal policy ‘should be aimed at domestic . 1 :

. policy objectives: - yarly publications py Bell (1962) and Kenen

“(1963); commissioned by the U.S. Government, drew attention to the

‘5 . .

- | _

"1/ See Mundell (LOGE, Chuts. 11, 16-18) for several of the important .- carly contrive sions and Whitman (1970) for a SUrVCY.

. .

area and stimulated further work -- See, for 3

and Kenen's (1973) subsequent research (carr

put not made available for general circulati

1965 article in the American Econe:i.c Review gnducing critical comnents (see Heckerman, 1

paffer, 1967; Stein, 1967) and focused more exchange-market aspects, . ¢

By the end of the decade, a large,

gone, employing inexcasingly sophisticated Tm techniques. Bra anson' s important Ph.D, disse 1968) was followed with further irticles ref Rranson and Hili, 1971, pp. 11-26; Dranson é€ collaboration of Miller and Whitman resulted .(1970a, 1970b, 1972). Bryant and Hendershot bilateral capital flows between the United S to illustrate the general probleas confronts

this areca, . ..

exe enpley Cohen (1963)

ied out in 1964-67,

on until 1973). Stein's

had a similar effect in

967; Nendershott, 1967;

attention on the

volume of work had been ethodology and statistical | rtation (published in inins nis work (1970; nd Wille: ., 1972). The

in three centributions

t (1979, 1972) focused on tates and Tapan in order

ng cmat cient researen in

; . o e e

Kwack (1971a; 1971b; Kwack and Schink, 1372} has estimated

equations for the U.S. capital account as pa to construct an econometric model for the UV

Avhole, Prachowny (1969) had a similar obj

yt ef ac. avbitious effort

. e oS, belence ef payments as

cetive. ttarston and

Fuack (1973) have » recently abridged Kwack's mode} zs pare of an effort

to {ncorperate it into the Wharton School mo

‘ ° de], cf the U.S. economy.

"a recent Princeton Ph.D, dissertation, Richard Herring (1973)

estimates an aggregate equation for net capital flows and compares

the results with the disat SHTETS ted re esults of Branson and Nill. (1971).

_ The substantive contributs.ons of these various studies are discussed

in subsequent sections below. . ew

. Early contributions to the Literature on the Canadian capital account sncluded those of ‘ Phombers (1960, 1564) “and pouric " (1964).

Black (1.968) stvd® ed pilaterel us. -Canezdian flows im the course of testing his extensiona of Tsiang's (1959) model of exchenge-market pehavior. Armdt's "say 9605" spotsed the partial-adjustment, distributed las model to y.§, -Canadian data, focusing on speculative behavior. Lee (1969) looked at U.S. residents' holdings of Canadian

he wee the fire explicitly to 2

wee °

poxtfolionadjuetment theory ~7 sec below -- £0 U.S. -Canedian flows.

: still another paper employ3ng the pilateral U. s,-Canadian ‘ deta was

done by Nawkins (1968), Caves and Reuber (971; 1972) have carried

out extensive research on the Canadian capital account, with special

moms eseee ert

a, enphasis en Canadien economic policy. virtually all these studies, in

_ one way OF another are concerned with exchangermarket behavior; this

" preoceupe ation can be traced, no doubt, to the interest in and concern “ebout this gubject” on the part of Canadian policymakers. .

.* for fuli-seate econometric models of the Canadian palance of

payments, one can turn to officer (1968), the TRACE model constructed *

University of Toronto (choudhry ct. ales 1972; Carr and Sawyct, 1973)

eee

as part of their model project

-countrics.

(1972a),

Ind

and the impres ssive efforts sponsored

for exanple, “Hellivell et. al. vo (1969, 19722) -have made ori modelling effort

Welliwell et,al.

xelevant.

that of Charles Frecan2

availa able gn published fora,

Econometric analysis of capit

of payment

‘exanple, Bo

Will (1971), and Nerring (1973) h

‘United King

The studies by Branson an

Receat cmpi

es is currently geing

atvrigh=

dea.

yical work

ginal contributions to the capita (sce Scetion IT below).

in the volume on Project 1L

mtn e-

by the Bank of Canada -- sec,

(1971) and Stewart (1972).

and Hellivell and Maxwell (1972b, 1972c) in

The chapter

. ny;

Helliwell

particular

L-account’ part. of this

done by

JNK (Ball, 1973): is also

on Cenadian capital flows includes

the empirical work itself is not yet

forvard

end Renton

sarch on the ce spital

*. S$neluce res

(1971) and the interesting P

(1972).

but see Freedman (1972).

al Flows in the U.K. balance

at’ the London Business school

tor the U.K. econemy;

sce,

Lor

Nodjera (1971), Branson and

ave also xeported equations for the

accounts

Additional work on German capital flews

Argy

. are discussed -- (of the Geiaan

. equations for

in Section LIL below.

eeenemy incorporated in Pr the Ceamman capital account (

‘Martionsen and Sin

wiownann, 1973).

and Zouri (1972) and Kour

apers of Porter (1972), K

na Will (1971) and Nerring (1973)

of several other Europcen includes Willms

ourd and Porter

i (1973). These Latter papers

oject LINK now jncludes

The University of Bonn model

sone

see Sandennann, 1972;

et ore one Ae

Kours ‘and porter (1972a) and brgy and hours

report equations for. the Netherlands and realy.

on the capital account i

carried out by, economist

n the Ttalian balance re)

s at the Banca a" rtalia,

others at the Istituto ai Scienze Econemiche at

Bologna. .

Porter (19728) y, and also sis the focus of Zecher's

° usaf

Austrelia is the fourth’ country stu

(1973) done at the University of Chicago. The

project of the heserve Bank of Australia has al.

flow in the Japanes

to Austo:alian capital flows. , * .

. pespite having worked

e balance of payments (bry

e .

myself on cur cagn

972) also

pepixical research

£ payments is being

“and by Basevi and

the: University of

econometric model

so devoted attention

4 { ° t .

e ane Yerdershett

Epes of capital

1970, 1972), 1 aa not sure how extensive the research has been on

investigated as part of

gn Tokyo, and at the Bank cf Japen. Amanc's

the Japanese capital account, To seme degree,

this area has been

the econenetric model projects at the In

_ of Economie research, Kyoto University, at the

*(1973b) is one jmportant source of information . L :

e «

Econcini¢e pionning

on this rescarch.

In generals as a selective scanning cf the volume .repe

on project LINX (Ball, 1973) will show,

models have concentrated on the current account in t.? palance ©

payncnts. Given the yelative lz

ack of developmen! of the monetary

stitute

Agency

reing

most national. econenctric

£ a

sectors of these models ana “Chie “STureity’ o # data for the capital account

that is both reliable and eccessible, the ‘onission of equations for

the capital account 3s not surprising. Several partied pants jn Project LINK hope to see the financial and capital~flows aspects of LINK. given greater attention in the future (see, oF exenple, Basevi and Waelbroeck, 1979). | Theoretical modelling and empirical study of the Eurocurrencty market's has only vecently begun to grcv repialy. This leg can he ‘attributed, at least in pert, to. the fact that approprj.ate and reliable

data on the assets and siobilities of surobanke have been difficult

to come by. Moreover, time series of sufficient length to allow

medixatons ceenencts york only becane available towards the end o£ the 3950's, lienderso t (1267b), Bw ack (1971¢); Axgy and Hod jexa (1975); Mis eso ena torsng (1973, Chapter 6) have all considered the determination of Eurocoilar suterest T2Les. Papers by Makin (1972)

and Marston (1974) attempt to estimate equations explaining dcposits

‘gn and porrowing froia banks in the Evrodollar market. I bave myself

some unconpleted research (Bryant, 1971) oa the denen by nenresidents .

of the United States for liguid dollar. assets. Several recent papers by two economists at the international Monetary Fund (Hewson and

Sakakibara, 1973a, 1973b, 1974) are janportant contributions, for their

more careful theoretical approach as well as for theix empirical efforts

- . Black (1971) studied the weekly. behavior of the Liabilitics

of U.S. banks to their forcign branches, & verd important element of

the behavior of the Furodollar markct during the period of his study; for two couments on Black's papers coo Valentini and Nunt (1972) and “Hassaro (1972). Mastrapasqua (1973) has also reported equations for the Jiabilitices of U.S. banks to their forcign branches. Unpublished / research bearing on this aspect of the rurodollar market includes the University of Maryland Ph.D. dissertation of Bradshaw (1973) and 2 paper by Ciccolo and Mekelvey (1971). . The distinction between Nchort-tenn and “Yong-tern" . - | financial capital flows is in sone respects not analytically interes sting “gs the di.scussion in Section II below makes clear, in principle both need to be considered together es paxt of an Sutegrated theory. See also Section

wTT!e) AL nmnenntan of the maawennetiats degree ana + was TZI's discussson SH wee appreprists SSRs wee <yPe of ‘¢iccEs>

orcover, existing tatistical a2to distinguish between thom (if they are separated at all) on quite arbitrary srounds . usually on the pasis. of whether the original maturity of the financial gnstrument is greater ox sO less than one yeaz. vor these reasons, I have not emphasized this distinetic: “in this paper. These especially interestce ja enpiri.cal research on “portfolio capital flows of a longer-tena nature should consult,

inter alia, spitaller's survey article (1971), Miller and Whitnan

(1970); Lee (1969), Boatwright and Renton (1972), and the relevant

parts of Branson and Wilt (1971), Kwack (1972), and Amano (1973b).

PO Oe ce

a” Another method of mapping this literature is Co classify - > the contributions by the theoretical approach or type of theoretical

specification employed by the researcher,

. . ee . bad

ta : Q

a ee ome com © . . ~

. . 8

. . . .

. . . ’ { rs e . .

° o ° . .

A majority of the carly contributions relied on what came + to be called a "flow theory". (See Section TI below for 2 discussion. ithis specification was used in part of: Bell's work (1962; contrast the “form of the equations in Appendices IL and I1J.), and is found for , example in Kenen (1962), Rhomberg (1964), Powric (1964), Stein (1965), Kenen (1973), pxndt (1968), Navkins: (1968), and Prachowny (1969).

More recently, the bulk of the xesearch has been based on

more or less sophisticated versions of a Nstock-adjustacnt' or portfolio-balance type of specification (again, see Section TI below). . Branson (1968), Grubel (1965), and Rondershott (19674) wore among the

first to criticice the theoretical shortcoaings of the earlier xesearch (discussed below}. For a representative eapniing of

contributions based on the portfolio-acjustment approach, see Amano

(2973a), Branson and Nill (1971), Bryant and Nendersnoct (1970), | Rwack (1971la), Lee (1969), and Miller and Whitman (1970a).

i Most of the studics cited here, with the iinpertant exceptions --o£ those dealing with the Canadian cepital account, heave mot showa an especially strong interest in exchs nge yates ana the Forci.gn- ~oxchange “markets as such, Of the research focused on this exca, black (1968,

_ 1973) and Nellive .2 et.al, (see the various ref xences cited above). stand out in importance, Additional studics that concentrate attention

.. on exchange o~narket aspects of capital flows ine)ude stuli (1968),

Kesselinan (1971), and Dooley (1974).

ry . ary a ad

{

. ow coo oo omeeree * .

ae aneamaeeeepnesiee at,

wee

yardly any enpitical-3 seesgarch notivated by

approach to the balance of payments

pepers such as Johnson's (1972), however, have pr

. Girton and Roper (1974). More work

xesearchers to Exy their hand. Examples

has yet be

en publis

a Mnonetarist"

shed. Theoretical

ompted several

are Zecher (1973) and

en a

in this vein will probably ‘be

available in the near future. On the distinguishing chazacteristtes

of the moneterist approach to the balance of payments,

- (1973). : Fg,

Department at the Intemational Monetary Fund

fouxi and Porter (1972a) ~~ is aifficult to characterize as either

“Umonetarist’ or Nportfolio-balance™ +

Finally, to conc luce this

reader to & recent survey by Wodjera

in its app

$ts authors, it constitutes a synthesis of bot

road MAP y

see Salop

wt Some empirical work recently coming ‘out of the Resea arch:

-- see for example

roach. in the mines of

, approaches.

J should alert the

(1973); which in a different

ysaanner is also a wap of the literature on sntexrnational capital

flows). _

js not practical to try to gunmarize in any

specifications used or the empiric:

general issues that arise in “connection with €

. . e.

se sear neme emp tnnees om oe we mms e ee pearung enemas

Although in what follows I discuss

a results of a nunber of the individual

my Limited objectives are to call attention to

. e

il results reported.

movements “(although Yodjera restricts his survey to short-tena capitel

selected aspects OF studies mentioned above, it

detail the theoretical .

Instead,

some important

he theoretical

+ rage.

e? ‘ . « ¢ ‘ . e . ° e ° ° . ‘ . . . . “ . °: °, . . ey ° . {-10 : = . ; ° . ° . ° . oe . ° . . e e ‘specifications used in these studics, to point to sone unresolved - Le questions of strategy for future research, and to present ay: own . assessment of how far the profession has gotten jn our empirical - work in this area, ‘The following sections of the paper take up these topics in turn. ‘ - ; . - . e ° e ee . ° e . bd . e & . e ° . . . + . : . : ‘ ° . ° - oe, : 2 : . . e owe one . a ° ° aes : “4 . : ° ry . . e . . ‘ e ; e . ° “ ~ . . - ; : . . . . e . ry ~~ - er een te en een game a _. o e : . .

el . . ‘ Mo LI-1 mo Te . ° 4 o Ny . . “eo, . " .- . ; an . . ITI. - General Tasues of Theoretical. Specification ., . . : "ops - Ag already noted, it is the so-called portfolio approach to capital Slows that underlies. the bulk of recent ‘Literature in . this area. the basic ideas underlys ng this theoretical approach go back at least | to the works of Mark owl tz (1.959) and Tobin (1965). . so At theix Lowest common denominator, all those studies that ; - base themselves on the portfolio approach , isnply an uncerlying specis .- 4 . - : fication of the following form: . ° . FY = £(SRyre eRe ea ; - ° ° oF : S$ . eoaeoeves .o weer O Xapkasceecd . ° _¢ ° : “ ° - . oe ° e , . e . ° e .- ° ° . ° ° : . ° : ° ° . : FF, §(SsRyreee ee RyeOpye een ny? sRerecees . . “.. % ee. . e : ’ e s° * . . . ; . . : The yariables Fy%yeeeYn® repres sent desired 9 iar te Lies of the financial oot instruments held in the decision-making unitts porttolie of assets and . Jiabilitics. Tnes2 . functional forms are, Sn the ease of assets, stock ; deinand functions; when Fy* is 4 Jiability, the equat icu is a stock se ‘supply function. Ideally, the systeal of equations spoutd refer either t eo e ° ° wy: * an sndividual decision-making unit or to an angee eRe of dacisic enemas} - 1! “- ynits that are reasonably honogencous gin characters.” ; . . \ ad —_— e ; 1 “* J] Sec section LI be sjJou for a discussion of tne appre pe se te desyee ANe type of ais agerenarion iu empirics) apa iea ties

pare ca et a cer erceeerrerren—imnrer eee on ye, - Pg ee_eEy <P Ree ose PP ENS. EE PE Re . naman wens LTTE serra on: “pam? . . me we PO I He STR GE ee Chenin de ‘ad peewee apumnce ay

Do “d

The symbol S represents the “scale variable indexing the

, i, $i.ze of the decision- -malting unit's portfolio. In the case of an in-

dividual houschold, for example, the scale variable vould typically

be taken as wealth (met worth). T£ the economic unit has no Liabilities--

eee

that is, if expected returns fron investing do not exceed the er <pected

costs of borrowing by enough to overcome the risk avers jon of the unit-total assets and nce yorth will be jdentic al. In the more general case where expected asset yields exceed expected Jiability costs by an amount

sufficient to make borrowing desirable, the unit will have 4 detenainate

scale if it has aversion ‘to risk and if its -warginal utility of wealth is

nonincre2zsing. In principle, with th 3 entire portfolio correctly

specifica, 4 balance-sheet constraint will hold such that 5 will be equal to the s sun of the Py” * (where liebilities are treated as negative assets).

In a true general equilibrium approach to the theory, the scale

_variable $ would be detemined endogenously and simultencously with all

the other com ponents of the portfolio. In practice, both theory and empixical applications have assumed that the scale variable can be taken

° °

as exogenously daternmined. For households, this assump ation is rations}

by arguing that income can be taken as predetermined (that is, based on pars~

decisions 3) and that the saving-conswiaption decision can be analyzed 2/ s$ndependently of decisions ‘about the composition of the balance shect.

. .

<<< . ; 1/ See Bryant and Hendershott (1970, pps 5-6) for a more detailed discvesi<

2/ See, for example, the discussion in Yobian (1961). ; °

ae.

eseeae

TI-3 : .

om ow

that profits can be taken

For fjsms, an analogous rationalization

as predetermined and that the dividend-retaintd earnings decision can

- be studi.ca independently of balance~ sheet decisions. .

The vector of returns in the above sct of equations, the Ry,

and the vector of associated risks, the G3, are in principle the

expected risks and returns associated with each of the assets in the

portfolio of the Gocigionemaring unit. Jn the case of Jiabilities,

these vectors are expected borrowing costs and the expected risks associated with these borrowing costs. Fach composite return or borrowing cost may have several components: for exemple, the noainal Snterest

and--when assets or Liabilities

yate, the expected capite

pa Cc? ro) he os) °

. +4 1-4 ° 1) n .

denominated in several different currencies are part of the portfolio-the expected change in scjavent exchange yates. In principic, eacn of

‘the return and risk variab?.

“+ e 177) 9)

_ppears in the equation for every asset and

Liability sn the portfolio. . oe

, The vector of variables Xy.%2s etc. represent all those other

. variables that are relevant to denand or supply: An example vould be the

volume of transactions of tne decision-making unit. Tf @ voriabie X54 influ: -the demand oF supply fox one particular asset or Jiability, it must (becaus

of the balance-sheet constry2int) jnfluence demand or supply for at least om

other asset or Liability. ..

Suppose that one or more of the T;* represent a claim on or

Liability to economic units outside the country in which the holder of the

portfolio is resident. ‘There is then the potential for snternational

accep ee ‘

sap —weeeog ee re ow ee ome

ese oan Ss a

* Security holdings

i to

a on . . - : LI-5 ; *e ‘OO Cha . ; Tab).c 1 es oe Hypothetical Ralance Sheet of a Japanese City Bank ° se . . (June 30, 1973; pi.dlions of yen) ; 4 . ° oo ote . Assets - Liabilities and Net Worth

eens eon

Liquid assets éenoninatec in yen Liquid assets held abroad denowinated in S$ aad other foreign currencies

and dome

Loans

c¢ customers

In 8 e

- to

forcs.en residents,

. in yen

Other asscts (in yen)

.s ° . . . . . : . . . . . . ee . . . . ° . id . - . s : . . ~ e 2G . a” * e e . ° . . . « ° e . e ee é e ° . . % . -4 . 7 : . . ° °

LAd °

discounts (in yen) Ld . sti . and aiseounts’ (in yen) LE

d minated enomin ed og

bilities (in yen) se residents

Deposit “iiabilities (in yen) a to forcign residents

LAE a,

Borrowing in yen from Bank

-of Japan and other Japanese Bj commercial banks . YorrowLag denominated in § B ; . froin U.S. Banks oy us

Borrowing denominated in $ from Eurodollar market

OA Other liabilities (in yen)

Capital Accounts . a

e . . . . e . e . . é . ” .

capital flows -- changes jin assets or liabilitics involving transactions

‘between res dents’ of two different countrics. et As an exemple, consider the simplified balance sheet shown t ae in Table 1. ne assets end liabilities of a Japanese city bank are

_broken down into tuelve main categories plus net worth. Five of these

.. -categories -- LAL, Lf, DE, Bus, and Be$, three of which are denominated

- * :

in foreign currencies -- represent cleins on or liabilities to nonresicents

of Japan. Changes in any of these five components of the bank's balance

shect give rise to an international cap

tal flow as recorded in Japan's : balance of payments. An analysis of any one of these types of capital

- €low along the lines of the portfolio appreach yeevires the specification

. . . of a consistent set of stock asset~- denand and Liehility-supply equaticns ‘€or each of the separately identified ‘Componente fs this illustration —- if . .

syelve) of the bank's portfolic. 2

Typically, ‘though not exclusively, 3c a: -ssiued that the Genanc /

yO supply functions for the decisium-making unit ave Linear homogeneous in the . scale variable: “ . ‘ . ; Lone ee & . - XL ) . . . . . . F = R eeee nN o eee Oe “27 wee eS - . . . . 1 Bi 3? 3 n’ 1’ n? S 3 3 2% e ° . - ° . . * . . $ . w . XY to . “¢ ° ‘ ‘ . . Fo = Bo (Ry seeee Ryyre ee Py org- rare DS ; . . . . . . - . y a ° . . . : . e ° . e . . . e -- -* . i F } X, Xa “et n = By (Ryser Ry Oper Tyre begets 3L o . . , . a the then sheet illustrated here is siuwiler vo Cie consolidated bal: . Oo . Io — att Japanese city banks studied by Brpayt asd teadershoce (1970, te Ms Tor another iMustration, woe the beliace sheets specil: . act 5 te A ape Aap ines eee A BENCH S Mewer country waeeh O97, pps S78".

Pan WC Rhee

mote oge bees ©

Poe

oxel

Shis assumption implies

variable S and in the no

‘TI-6

that an increas sc of Z Pex cent in the scale

n-return Kaistribution' | variables xX will,

ether things equal, raise the desixed stock demand or supply Fy* by .

Z per cent. This partic

seve oral practical advan kanes

searcher more ‘voadi ly to

empirical estimation. Lt a1s0 makes jt ca

of inerements in the sc

ilar speck f.eation of the equations has

eee te

among other things, it allows the re-

enforce the bala ance-shect constraint in

sy to sce that the impact

fale variable on the desired quantities is

dependent on the levels of the risk and retum veriables, and the

Jmpact of changes in the Level of the scale vari

s asset Aasands or 1

" shange, or first differe

~ ‘°

One can see from this

eorclusion that internat

return and ris

Sabili.ty supplies

nec, of the above

o% .?

able. Flow denands

% variables is depencent on the

or SUPP niices -- changes in

-- are the time ratc 0

equations: -

Last sct of equetions the now accepted

ienal cepstat flows, seen in a port folio-~

pbelance pers ective, have both a Noontinuing- -~flow" and a “stock- P P

. pdjactrent” component. Civen a onee-for

eall chenge in vetums or

risks, the oxisting-stock effect produces capital flows “that are also

once-for-all in nature ( a continuing~ flow effect

gn the scale variable is

a reallocation of

existing portfolios), while

persists sndefinitely as long as the chenge

not zero.

tI-7 0 a ce

awe bon

To digress for a moment, T should point out that much of the

‘early enpirical Literature on snremmational cepital Flows suffered from a lack of fomiliarity with contemporary trends sn monetary theory, end hence prominently featured a confused debate on "stocks versus fJows:' For

we Oe e

example, many studies regressed capital flows on levels of interest

° °

yates and other variables such as the Jevel of trade flows. The specifi-

cation of these equations took the basic fom: . .

. - ° . . . . . e :

where Ry, Rg; ete. were the Jevels of nominal snterest rates and X might

be the flow of smports or exports. This so-called “Elow theory" had no *. theoretical juscification when carefully examined. It implied that desired equilibrium stoexs of assets or liabilities depend on the sum

(integral) of the current ane ail past values of the relevant returmm .

variables, and thus that elasticities with respect to these returns ere

snfanite. Similar difficulties arose from the assumed dependence of capital flows (thet is, changes in stocks) on the flow, rather than cheages in

. e . .

the flow, of imports of experts.

. ems

this incorrect specification were reported by

L/ Examples of equations with thi Kenen (3.962; sce also .1¢72), Powric (1564), Rhombers (1964), Stein (1967), ; Laffer (1967), Arede (1968), Hawkins (1968)., and -Prachowny (1959). Bel?

. (1962) was also unclear on this issue. A similar problema was prevelont 5.u

“4

i many of the theoreti1es.+ contributions to the snternal-extermal- imbalance Literature, waere capital flows were made a fuaction of the Level of intere. c

yates. Comments on this guoeck vs. flows controversy include Nendershott (>

ar Willett and Forte (1057), aul Transoa (1970).

An alternetive

Literature ai (3), but wes were regre

nges

% exemple, ché (4)

d not suffer still -incosplcte.

ssed on chenges sn the level

bYs = ¢( ADs

om 1ceee

{1-8 ”

es Ustock theory" chployed in the earlier

1 defects of

capital £lows

From the serious theoretica

In this Latter case,

5 of interest retes and, for

in trade flows:

e 4) bXy> « e¢ .)

SNS Beng > *

This formulation: can be sntexpretcd as approzims ting the existingstock responses, put it disregards the continuing~

* Oa theoretical &%

is clearly prefer

There i

“make to the prece

Responses to changes in the v

axe typically not york requires the

long-run des

rent 4

holdings.

supply in the short ren -=

nd legsed valu

flow effectse

ounds, therefore, the specification tn (2) above

ayle to either (3) ox (1). \

t ther one must)

s one further jSrooctaat acdit

ton

ding, gener o approach.

al description of the partic

ariebles determining cesixred demand/supply-

Tnerefere & CC:

joe

irmediatc. aplete theerecical freme-

and

specification ce)

£8 relationship sctveen snorte . : .

grea holcéings of the Financial InstrEnenes, where acnend/

F.*

$077 is expressed 65

a4 function of cure

es of all che determi min desired

nents of Jong-

Pa) ated cautions of this so-called Netock" form. , Branson § carly work (1968 ) also employed specifications of this Je, EXPCs

2/ Branson (1952) was among the £irst to sndarporate an ayarencss wi of the importance of lagsed responses nto empirical research ca

capital £lows.

See also Hendersno€e (1eG7).

Such, in the briefest of tems, are the theoretical under~

* pinnings of the portfolio approach to cepital stows! Now faith-

fully is this framework jasplemented in actual eapiricel research? There are several jmportent ‘respects, Tam afraid, in winch all of us who have been working in this area tend to be

cheating when we claim to be applying 4 Markowits-Tooin type of

theory to Snternational capital flows. Tor cne thing, many stud

are exbarrassingly sloppy in defining waich dec3j.sion-meking units

—

are ectually being studied and wnat their aggregate balance sheets look iike. In principle, studics dealing with changes in assets

end Liabilities that are sntemmetional capital Flows are conccr-

. & av, ary . ot. & yom w Leave eee ae ee - mee et fe re erating on a SucCroE SKA ehe re fou Gone SATCECiiaee EEZLEZLLE OL

économic units,” - In many ceses, beceéuse authors have not even

thought their wey through the problems at a conceptusl level,

yeaders of the literature on capital flows encounter difficulty in

Geteraining Vho these transactors are, let alone what other assets

‘or Jiabilitics they hold in their portfolics. There are egregious

cay) Yor more @etaiicd discussions of this theorckical framework

and its applicebility to enpirical research, see for example Bryant and Hendarshott (197C),° Branson end Wil1 (1971), and Millex and Wiritican (10708).

2] Yo continue the exeaple given sn Table 1 above, Bryant and Nendersholt (1.970, 1972) were conc ntrating on changes in the variable Bus, aggregated over all Japsnese banks.

. . . . . .

examples of t

Wi11 (1971) ©

emphasis on careful delineation of the decision-making units and

their balance For

nusber of the

IEse onisss.o

there is typi variables in

‘awong then,

may argue the In other inst

to collect al

ay .° . - {I-10

4 his sloppiness in much of the carl

before the perspective of the poxr’Ecisp mpproech became widespread),

- And even in the better recent studies -- for example, Branson and

yr Miller end Whiten (1972) -- there is Jittle or no

sheets. another thing

variables that the thcory

ns sometimes occur for valid reasons, To iJ lustrete

ar eo

g, many studics simply omit a laxge

argues should be relevent.

cally a scvere problem of including all the return

any given expirical regression because of collinearity

: : 5 . . < - . +t --t ° TE the eollincarity is services encut! the wesGare!

eo were ee

t he is

hs % ov) Ve bas} 2 rt Fe ct ct 1s

cane Uottys

ng sene of the variables.

ances, however, the researcher may not even attempt

1 the return variables that are relevant and may not

even bother to argue for their exclusion on grounds such as multi-

collinearity,

deci.sion

e/ “Interest to be str

_. research “interest

(of course, if no explicit sttempt is made to specify

wat ‘ Des retese iy van .

° al

aa be “> Vat

Ons a ecst aue

13, and this

so Section IIT below, where there is a discussion of the a prespective researcher must meke about the appropriate

eresation of aguregative data, én end the United States, for example, tend

difficulty plegues all the

re on Canadisn-U,S. capital flows. Furodollar and U.5. s

a yates move clo

$s still more severe en nag interest

“country; separatc within th

as is clear froa (1) above,

role in the portfelio decis

e country.

aly together. The preblon of collinearity

rates within a particular

y jJiterature (written

many of these play an jmportent,

sions of trunsactors yesj.dent

e aor"

ee ed

“ee ee 8

". trend (19702, ‘pp. 162-183). They further pocit & relationship

_the appropriate aggregate balance sheet, then ene docs not even

know in theory which return variables are xelevant. ) Even in ceases where a careful researcher excludes certain variables because of

precticel considerations, one is bound to be lcft weasy. We know

from studies such as that of Brainard and Tobin (1963) that it is

possible to get into serious trouble tu finance

tal model buildcong if one forgets tco easily ebout balance~sheet constraints and omitted variables. sO , an

° .

“The literature pays little more than lip service te the

- theory in another respect, Risk variables moy be mentioned in

introductory naragrephs, or in & section d@oscribing the theoretical

frenework which supposedly unde xpins the empirical research, T ne

actual enpirical york, however, seldom includes weria les purporting

-to represent these risks. The lip service varies From study to

study, but the most that any heve done is to gsttempt on an ed hoc: ¥3 ) } ed hoc

basis to create a few proxy variables to represent these risk effects.

“As an example of such én att emp, censicer the Miller end

“Mitman pepers (1970a, 1 972). They argue that the riskiness o

domestic assets in the United States is inversely correlated with

fluctuations of aggregate economic activity ebout its‘leng-tena

See Grubel (10SE) Ler. ene of the aa lier disevssions of risk aversion @s an explanation for the internatsonas diversification

of portfolios.

: . .° . . . . : . . . . . .

between the capital control programs of the U.S. Govermment and the riskiness to U.S, asset hoiners” efforei.gn Jending Finally, they

argue that:

manifestation of imperfect infomsation, the risk-estimate associated with ‘ an ecset should dinini as information concerning’ 7” ° a the probable retura on “the asset increases, For ’ his reason, we hypothesize that there has been a secular dowward trend in Cy (the risk of investing - ‘abresd), ing From the jnereese in knowlecsge - and communications, which symbolize ss the gradual movement toward integration of intern national shorttema capital markets since World var IL (1972, p. 26%).

ee 0

ee e SINCE risk

i rs

RK ee . wel

“Tn actual practice, what are Hiller and \Vhnitns: 's empirical proxics?

“They arc deviations of U.S. GNP (not seasonally - aajunced) from its

trend regress:

a S$) K We =, a ~ Q

on line over the esti mation pevileds a duzm

os Nee eee y

c . oe + com. - «ee

equal to zere until the first quarter of J

365 and unity thereeftar;

and @ simple time trend, The reader can judge ‘for hinsel£ whether

these variables seem convincing prosies, fox the risks discussed in

portfolio theory 2/ a _ i oo BT

The poz rtfolia- ~choice theory vorked out by Markowitz and

Tobin also argues that covariances emong returns are iuportant,

The existing empirical litexeture on capita al flows docs not even

bother to pay lip service to these covariances, They are typically

not: mentioned at all in the paragraphs of a study describing its

. eee : , y Willer and Whitman pay more attent son to this problem than most ; _ yesearchers, and therefore deserve less criticism than the averege . study. . . aren . . oo % " ¢ oa : ‘ e . . . he, ;

“theoretical -sancont,3/ (Tne same eriticisn also applies to the

“ exposition above. )

. °

mo ’ Z noted carlier that the returns and risks that are

relevent in the theoretical specifications arc expected “effective

. returns and risks. Host of the existing empixice 1 ‘vesearch sidesteps

the question of exp ectaticn -5 and exnectations-formation altogether. 1 o

It sisaply © es cbsexved values of the variables, making no effort x to collect date on expectations or to formula te proxy variables . : ON .

thought to be correlated with the theoretically appropriate expected values. An outstanding exception is Dlack’s application (1973, pp. 22-27,

39- 52) of the rational expecta ations hypothesis to exchange ‘parket “Se How important are such civergences between the theory believed to.be relevant and the eguetiens that are ectually estimated

in practice? If we answer this question honestly, we have to admit

that we do not know, ; .

e . mad

Jn all empirical vork, it is true, there are serious a

problems ia obtaining ompiricel approximations for tive theoretical

: . a constructs thet the theory says is relevent. these problems of

. te ° .

obtaining suitable empirical approximations ‘may not be any more

difficult in the ar a of international capital Flows than in other arcas of empirical research, MNaxreover, one often hears an argument ~~ dntended to be comforting -- that empirical equations actually estimated

\ . <<< Sy, Waller end Viderxn show the covariance terms in their theoretical discussion, but co nee try to incorporate them in the empirical analysis.

KN .

ie + oS

me ee bas.

: . . . .

would not fit very well if eni.tted considerations were really important,

Jn other words, it 15 argued, one can velax about the divergences between

fit the data reasonably well.

still does not comfort me very much, a@ woman Wwno welhks up to a gtreet corne

snapping his fingers, and lookin

ay . .

The women asks the man: ‘Why are you standing on this street comer

theory and enpirical practice so long as one's estimated equations

° seen . .

. .

. . . e

este eH

\ This line of reasoning, while 1 believe it has some validity,

t reminds me of the story of

to find aien standing there

g all around in different dixections.

gsnapping your fingers?", The men replies, “I'm snapping my fingers

to keep the tigers awa © The woman with a. puzzled icek on her face . e - . 7

asks the wen: “You don't expect me to believe thal. siucy, Co you?"

And, the ian contjauing to snap his fi 3 vo

shoulder, says: ‘You don't see any ti

which are queatitatively wost imporxrtent’ is somenPath “umaio ‘4 x :

a &

oe very well if we hac failed to snclude those explaretery yeriables

man's contention that the tigers are staying away because he

gers While ac ivoks over his

ers hereabeuts, de you?"

J." The argument thet our estimated equations would not fit

Y a

sviapping his fingers. Tor a high valuc of the coerficient of determinatios

jn a reported regression is not sufficient evidence that the cepression

actually represents a cod approximation to Economic Solinvies, Few P

researchers are rigorous in specifying

. rigorously testing that one hypothesis

a hypothesis and then .t

alone, Yt ia sempoing ¢° rerun

a oe 1 2 wren.

equaticns to sec what would happen if this vari able were included

or that variable omitted, By the endef such a process -- which can

easily, often inadve ‘ctently, turn into @ fishing expedition -- one

ought to have greatly reduced confidence jn the final results. Strictly speaking, the treditional tests of statistical significance are not ~

yalid, Certainly, one cannot convincingly claim that any omitted considerations thet are theoretically relevant but not onbodied in L 1/

the empirical estimates really are unimportant,

. ‘ - In Bryant and Hendershott (1972, pp. 227-236), a number of ° so for the same set of data ° deliberately mis-specified equations hore reported for purposes of

comparison. Several of the mis- speci cd equations had standard \

errors Of

ewe wsl tlh, ha ernta Le-— Tore Meek cee Ue Aen dee

aatimate that compared fevorabl

correctly specificd equa ations, Yet the mis-specificd equations

implicd jmnplausible or even ludicrous econemic behavior. Tnese

yesults demonstrate all too clearly thot ‘a researcher may frequently

be unable to discriminate, in purely ‘statistical terms, between alternative imperfeck spec ci fications, and that fishing expeds. tions are virtually bound to be "successful" j£ one is satisfied’ with merely

finding some specifica ton that will give a good statistical fit for

the sample period, - ot . cer AA °

Mu For another view in a sinilar vein, sce Rhaabexrs (1972, pp. 314-318).

I 7 L1-16

By wey Ve another illustration of the gap between theory and

* empirical practice, consider a partic ‘in connection with the scale variable peadaches involved in obtaining adequa

-* capital. flows themselves oF for an ap

of adequate date forces rather drasti

compromise. ss to cait the scale varie

ular set

. There arc, of course,

ate data either for the international

propriate scale variable.

c compromises

eno

of problems that arise

mOUS

Lack

One possible

equations, This, for exemple, is the route chosen by Branson and

gn their study of capital movemcnts

p. 27££). Whatever the sapirical consequences of this

omission of the scale variable elearl

the scale variable, An wxeaple here

and Whitman of shorft-tera capital flo

represent the net yorth of economic units resident

States, but also a sea.e variable rep

averege of Foreign wNP'S to serve 2S

S.. "" comprowise is to make souc sather str

(1972, 1970b). Their theory calls not only for

therefore resort to an arbitrary assumption that the

y guts the theory. Another

vould be the studies by

aresenting the 2ggr

cele variable

in the United

a proxy. Niller and Whitman

ratio of the

a scale variable to

ong, arbite “ary agsunptions about

-¢o obtain in practice. It is difficult even to construct a weighted

ble altegether from one's estimated Hill anong OCD countrics (1971, assumption,

os in the U.S. balance of peyments

egate net worth

« of course impossib!

¢ foreign scale va riable to the donestic scale varsable is constant ovey

.- - have becn pragaatically rensonable,

the period of their study, All things

Lt is nevertheless imposs

cons sidered, this assumption may

I-17

justify on the basis of the theory and even

. . « . . oe

stretches the Smagination

a bit to assume jt was approximately operative in practice. ; lw

Even if data are available for the

assumption of linear homogeneity 77 see equa

* be justified cn the basis of the portfolio-b

Tests of this honogenei ty assumption, moreov

Branson and Will (1971) dic make an effort t

of non-honogeneity in their capivical specif

capital flows sn the U.S. balance of paynent

that in those few equations in their study wv significant departure fren homogeneity -- sc

scale variable, the

tion (2) above -~ cannot alance theory per Se. -

er, are seldoa attempted.

o alles for the: possibility seation for certain

s,’ 1 suspect, however,

here there 1S an apparent

e, for exemple, the

equation on pase 18 for U.S. short-tene Liabilities co foreigners 77

other factors are at work in the equations a and iendershott (1972, Ppp. 224-25) made «en © this assumption within the context of 2 non- In effect (using the notation above), they es

following foun:

(5) Foe gGes .s? ".

Instead of assuming that the parameter 9 was

gr . .

es Hi 3 nd the ompivical estimates

do not represent a good test of the homogencity anciceph ion. Pryant

xpliczi: attempt to test

linear .iny.rsion progres timates an equation of the

— eM .

eauct tic crty, ia other

y Yor a more extended discussion of this peint, sc0 2 yen und Nendershott (1970, p.7). See also Roman Wi. oS aisevesion ef

this point in his ecrments in this. volume: (y pers}.

ee -

words, they allowed the pareneher.£o.talke,on any positive value, In

Pret) .

fact, the estimate of @ was close towm.ty, lending some apparent

support to the use of the homogeneity as

test is hardly cenclusive evi.cence about

the assumption.

: . Another respect in which the specifications used in the

failure adequatcly to incorporate the ef Here egain, most studies pay Jip service rates, while doing little, if anything,

into account. The most careful and inte

sumption, But this isolated the gencral ‘applicability of

empirical literature fall short Gf what would be desivable is their

Feets of exchange-rate changes. to the importance of exchange to try to take these effects

resting work of which IL an

evare has been done by Black (1973) end by Nelliwell and his colleagues

(1972, 1972b, ro720) A A distinguishing feature of these studies is the

endogenous determination of spot and forvard exchange rates within 4

model thet specifies all the demends and

- ghis is a big improvement over specifica

exchange rates altogether, or inelude th

element in the return variables entered

equation, In view ef the international m

eriod, which syoduced such dramatic alt P > prc

. o

eed

“supplies for forcign exchange.

tions that cither ignore oa siiply as an exogenous

on the right-hand side of the

onetary upheavals of the 1971-73

erations in exchange regimes,

1/ For earlicr studies that recognized the importence of the problesa, s¢% . fe . . ‘ . Stein (1965) and Rack (1952), Readers especially interested in the

verte

relationships betveon capital flows and exchanse rates should alse bg wneta > . «se , 7consult Stoll (LeGe) and Kesslerman (uev.1). Dooley (74) has reeomily

- conapletad a paper which J have uot yer bed time to 5 luey carefully. |

oo ne oo. Wi-19 cet - a .

‘ . a : eo? . o

: ‘ 4 , : . van cxplicic consideration of exchange yates in research on international

capital flows has . become even more important. Indeed, now that a

.° majority of major exchange rates are £loating subject to managenent

_ via official intervention, researchers studying capital flows have no

xeasonable alternative: they wust, by a quantum jump , jnerease the

sophistication with which they handle this set of problems.

. To make matters stil more gieficult, basic questions of

sesearch strategy are involved for those who try to tackle the capital

account in the balance of payments as 2a whole in their empirical Work,

The traditional approach -~ see, for example, Kwack and Schink (1972 or

Branson and Nill (1971) -- hag been to estimate separate equations for . ‘ . . each component of the capital account other than the change in official

yeserves, with the exchange rate assumed to be ecterained exogenously.

The change in reserves (the official settlements balance) is then

derived fron the balance-of-payments identity. Even with an exchange

regime of the Bretton-Woads type; this approach had unsatisfactory 1/ ;

features, With widespread

flozting of exchange rates, the rationsle for this traditional epprcach has been undermined quite ccijously.

. .

LV Within the Yretton-Yoeds exchange merpins, spot exchange rates were endogenous voriaples; forverd exchange rates were net even coenstreine’ by mergins. Horcovers novements of beth spot and forvard retcs had . importent effects on exchange-rate expectations (see, for exemple.

White, 1963), which im turn vere influential eceterminants of capital

- flows (especially et times shen the ability or willingness of gowerne

-. ments to maintain che spot margins were suspect). Yo complete the circle, capital flous often triggered cheyees in the spot margins.

Noenee, especially over sence lenger time period, there were wEny

: elements of ti fielality in treating exchange yates as exoganens ly

determined eutates enels medel of the capital account,

ie ett ae

' e | ; ae toe ; {1-20 .. : =. a e , : . | . . ; r . ;

. . : eo . .

Yhe line of attack followed in. the papers of Nelliwell and

his various collaborators may turn out to be more promising. This

alternative approach jnvoives direct specification of ene or more equations deseriping the intervention pehayior of monetary authorities gn‘the exchange market, (Looked at from another perspective, these

° " equations can be construed 4s modeling the official demand Lor reserves.) : Such equations, teken together with curreni-accounk and certain

. s

capital-account net demands for forcign e :change y- Lin ultencously ° . “

determine the exchange rate and the change sn official reserves. The ©

. residual iten doxived from the balance-of-payments jdontity is tnen

that pact of net cepital flows not already “explicitly accounted Zor by

wo -

_possibilitics fox the modcling of many alternative exchange regis (including, of course, managed floating) ane its possibilities for the

. specification of official exc change-market intervention as an importent 2 | endogencus component of the model. . ° ” 2 « °

Kote, however, thet even this’ Latter epproagh yuns into great

: difficulty in circumstances of wicesproad | floating, whe Canadian - 7 models of Yellivell and his colleagues could reasonably use the : - Canadian dollar/U. 8% dollar cross rate as "the exc change rate" so lon . _ aS any" che ange in the Canadian dollax vis-acvis the tmited States also ° involved a similar change agaiust third currencies. With widespread

floating, no single cross rate “- even for a couner? such as Ganeada 77 i.

ae .-

awe ky.

can adequatcly Prony for “the exchange rate," One hes no choice but

to resort bo concepts such as the "effective exchange rate" and

attempt to approximate them empirically with various weighted averages.

Yet this path runs right through the middle of the familiar swamp

. of sndex-nuaber problicms. Difficulties of this sort are paxticularly

geverc when modeling the exchange regime for a model of the U.S.

balance of payments.

Hy main purpose sere has been to peing out scac respects in which the enpirical literature relies on cquation specifications that

tcnd to be theoretically invalid or jnappropriate, Still another . jaapor tant illustration is the failure of the literature to incorperate

the effects of governmental restrictions on international capital fLlevs. Kone of the stadics with which 1 am familiar, with the excepticn of

Mryant and Nendershott (1970, 1972), has even tried to cope with this

problem at the level of the theoretical’ Lramework. Instead, researchers

have relied on doing sonething ad hoc when it cones time actually to

yun a regression, Typically, simple off~-on dummy variables are

insertee into the empirical regressions in an ad hoc way with no attempt to blend the underlying theory of the capital flows with the

governacntal restrictions which stand jn the way of private individuals carxying out their portfolio decisions, Needless to say, the fora

(een a

1/ Yor a useful sumnary of the limited work that has been done in this - area, see Cheng (1973). See also Nwack (1973).

om TI-22

° Iyi es 5 of such dusuy variables, Waich is u

a particular (arbitrary) interpreta

eapitel controls. °

As on example,

sually highly axbitrary, forces

tion of the effectivencss of the

~ . .

consider the equation reported in Brenson

and VWi12 (671, pp. 13-15) for changes in short-term claims of U.S.

xesidcnts oa foreigners:

. . * -. .°

. bof = £(U,S, weeltn yeriable, U.S, interest rete,

U.K, intexvest rate,

. + sessonel dunnies

~ 392 Ist) ~ 607 DF1 - 213 DF2,

In thi. equation, ‘and Hill explszin es follevs: “Since the t

ar progrems is the stock of £

the ennual tightening er

continuing inflows, er redu

. St ; we have SGcsd ten Gummy ow:

U.S, merchandise exports)

- 5 . . 2 ere ell duamy variubies, Branson

£ the [U.S. capital contrel] tending claims, we would expect pregrens to be associated with tflovs, of capital. Thus to the analysis to reflect

2 the progress. The first, prl, is set to unity in 1G45T- WE6STUL oad nace elcevnere reflecting the fact that the lurgest chenge fn the presren o2curred with its initial Siaposition, A second Curmy,; pr2, is set to zero Cnrougn ° LOGSITY, end unity from 1SG51V on to reflect the continucd

tightening of che progrens.

but: cxwe’: quantification

Mnoral .uasicn’ and a

: Zs impossible," ' , ° “Yhe YET cucmy was addcd Lecouce of 2 data."

change from voluntary to yaande tory vv

‘hese variables may be exucc, of increases in pressure through

is-1,0 in 1963111, This dummy

large unexp)aincd (the IFT applice only to leng-temn cépstal) residual in that quarter, effectively exops that observation Drom the regression

and

o 7 re i

As Branson and Will acknowledge, since the “Interes st Equalization Tax

did not even apply to the capital flows in their equation, their ,

LET duasay variable serves only the purpose of throwing out an awiw ord

obsexvation; jt certainly does not indicate that the Interest

w Equalszatto on Tax reduced U.S, short-term claims on foreigners by

- $392 willion in the third quarter of 1% 63. The other two cumy

‘variables, if token at face value, suggest that the U.S. Volunt

Ow Yoreign Credit Restraint (YECR) progres reduced the flow of U.S. short-term lending to foreigners by $607 million (conapared with what

ft otherwise would have been) in ¢ach of the first three quarters of

1965, and thet in cach quarter thercaftex “xéduced the flow by ; ..

$213 willion, , - - . ‘Yhe arbitrary specification of the impacts of the VECR - } i ,

progran by Branson and Hill wey be contrasted with an equally arbitrary

. and incompatible -- specification in the Miller and Whitman studies

- of the same capitel flow, As noted above, Miller and Whitman employ

a dumny variable equal to zcro through the first quarter of 1965 and °

equal to unity in each quarter thereafter, Tney alse cupley a “partial-

adjustment" spec cification in an effort to capture lagged xsesponses,

7 which inter alia forees the impact of their VFCR dumuy variable into the sane geemetrically-decaying lag pattern forced onto all the other explanctory variables in theiy equation, Since Miller and Whitenan

stimate a low value (.25) of the "speed-of-adjustment" cocfficient in

meee es seers . eAdenemnendenenesaneeraeeenetie ee ee te ee Seeeeneemanenememdnnettietae te

eet me eee mpe ite eee

' ad . ae

‘ . a . . . . their preferred equation (1972, pp. 272-218), their results -~ if

taken at face valuc -- suggest thet the effects of the VFCR in

restraining U.S, short-term lending to farcigners wore still gracuslly “ building up by the first quarter of 1966 and were still having only

_ two- -thirds of their eventva J oimpact. Lo, - oy

To gut the point milely neither the estimates -- or, to } y; )

speak riore precisely, the assumptions -- of Brenson and Hill nox those

on fi sf

s Willer and VivLinavar is spixve fone dence as reliable indications of

the impacts of the VFCR progran.

. . . o

T£-governmaent controls effect the returns oy other variables in a way that can be specifi ied ore or less exactly, the preferable thing to do 3s ta citemp: to aaju rot the return vaxiebles directly to - take these effects inca account. Fox cyanple, in “nadine 9 purchases

of foreign securities by U.S. residents or long-tesn Lending by U.S.

: banks to foreigners in the 1 953-75 period. one can add the rate of the

SIntexcst Equalization Tax dixeetly on to the appropraate yveturn variable:

a ave

used in the regressien, Similar quontitative edjustments can be mede

aryes or yoserve, requirements arc

g vy ft) 3 0) pay roe] %S a i) £2 A v? de o (9 We 7s er ayy bar] io) ea —n 7 te)

3.£, fox

Zmposed on bank Miadilities, 7

But usually the problem.is more complicated, It is worst

_ of all when the sovernmental restraint or stimulus is carried oul just

by administrative guicenc e (co®nonly knowa by such euphemisms as jae

: : “boning, car-stroking, threatening Co use the big club in the closet,

1/ Yet note thet tnese danny xariebhesy an het h the Preason Til! and

+. statistical ests!

i TI-25

vtec.)

except in cases Wiere controls arc

ald. economic units, the researcher

relaxation of controls wild reduce

the response of desired quantities a

minants, For sone

can be xeprescatecs the

tS’ i

(6) . ot oF ,°

(7) Fe

TH(Cy Cos Cg r- 004)

vere Ps ss the observed

anadias Agate 8a Fe

alpha vould equa and would

ail,

quantity

which governnent, authoritics use to stimulate capi.

than

rake on values greete of capital controls requires the

could

etc, -- wich reflect chan

Ch, Coy

controis 2! this b,

and Veneer

1 For example, Bryene ~varizble to measure pp. 60-62 \, See Bryant compar ssons of

an off-on dcuiy varicble).

One suggestion made by Bryant and Nendershott (1970) is that,

1 unity vhen the centxyols are bo less than vnity wien

below the desired quantity.

researcher fo construce 3c8

of course, an extremely difficult

the effects of the U.5. and Nendershott (1972, alternative specifications (including the we of

lnown to be strictly binding, on can assume that the tightening or

or increase, but not dampen altogether,

to chan ges in their economic ceter-

jtel flow being controlled, this view following way: i:

% , .

guantity of asset denend or Licbility supply,

vw. Fe ge Liners t wf ane ae Pe ed pate Perro Bare RUPE OrET

fanned

Jomand,. ebormi ox not bindicg et the controls kee

In’ the case

jel £lows, alph . . A) ° unity. This trestuent of the effects

variables 7-7

jn the incousity of the different

task, But it

shott etkenpoad to construct 2

Vv BGR prGOS ,oean (1970, nn, 228-22 OD) Cor

has the merit of forcing the researcher to try to specify explicitly

the manner in which the controls have their impacts and how the

dntensity of the controls varies over the period of the study -

something that should in any cése be attempted. u Thi.s type of

ee . . . . °

specification also tends to force the res searcher to use nonlinear estimation techniques. While these techniques involve addcd expense

-“ t

end inconvenience, these costs ney walt Le vorth it if the alternative is nothing better than the use of ad hee Gumay variables that do

noths ag wore than zlter the constant term in empirical regressions,

The use of various types of capital controls has been much

more the rule then the exception throughout’ the postvar period, For,

a survey of counivies! pr seu Lor exeampie Hilis (1572 19755) EUV Fo countless p actices, See Lb XNaMpLle 11S \iv (ya }

or the annual repoxts of the Intemationel Monecary Fund en exchange

restrictions; see also Johnsen (1973). Moreover, if anything >. gevern- . .ments have been Snereasingly prone to utilize these controls in recent

years, ‘Thus, cespite the practice in the literature so far, it is

Sble to dowmplay the importance of those controls and

simply not pos

still do valid eapixical research on the capital accounts of west

countries’ balances of payments,

T have said little up to this point about estimation problems

per sé, and for 2 good reason: no estimation technique, however poverful or sophisticated, can preduce satisfactory empirical results if

‘the theoretical specification itself is inappropriate, Morcever,

er cc en

i/ Perhaps this point should be put even more strengly: oa researcher has no Des isest applying ceoneretric techniques to Che data vat il such an attempt has been wadc!

wos . 1-27 a ,

gnadequate theory or faulty empirical approximations of theoretical

construets ere the most frequent causes of estimation difficulties

- + Ronetheless, it 25

that the simultencous-equations

Sn the discussion ebove about encn

rical resesrch on c2

fica in «mp the sect of equations of return variables, the Rye’ one-way: the return variable R demands

determined by all the

. . be | Suppey

Geman Gr 2 More the market for j on of economic concentrete on the intl Luence of influence of Ps in de

For many purposes, “studies, it may be necessary to both Pr; and R, appear as

tive research, of course, such

to yield explicit equation specifications

tested,

OY a Arsen eigen: ~ PepamnnmEmein outa, auen,.

chal corxelation of the vesiduals). 1 ote

probably %

(1), “in which the verious Fy

In principle,

trensaccor

units

therefore, and especially

endogenous variables.

«a

wecful here to remind. the reader

problem -* elready referred to implicitly

ange rates --.is an Smportent complica:

flows Tinm back for a moment to

vob e@

p rat depend on the vector this causetion is not just

sone Financiel instrument j will be

particular economic unit

-~ and a fortiori if. one

the less velid will it be to

R. on F:

5 j and to ignore the sibnultencous

. : °

se termining Ree

in macroecencmic

specify & theoretical mode) in which

To be useful in quentite model must be well cnough articulated

that can be empirically

Let 17-20 ' , :

. oe .

. t Stein (1955) was one of the first to stress the need for a

ginultencous-equetions approach in the context of international capital

flows, For additional references, fcc Diack (1968), Miller and Whitmm

(1970p), Bryant and Nendershott (1970, Appendix 4), Kouri end Porter

eta de

(19722), Herxing (3.973), and Norman Miller's coments in this volure

(pp. below). For tvo examples of papers that derive reeuced-form equations from & more complete wodel, thus solving out am endogenous

veyisble and removing one source of sinultangous-equaticns bias, eee

“

Nouri and Portex (19722) and Black (1573).

. e . e - ° . 7 —— i . 1 e ° t ° « -- r i . ~ é . . . ‘4 ete : : . . . Doo. . + . 3 io . . . . . . ° ° * . - . . . . oe e ° . . : e e . e . ° ° . ° . . . « . . . . e - Phd -” . ° ° . . . - ° e . ° e e °

a et: etme area es ae eres Sone

a aS

re ‘4I1T-1

‘ J ILI, Sone Unresclved Questions of stratezy for Future Research —

‘ . : . ° . 6

As the preceeding aiscussion indicates, there js no shortage

: of jnadequacics in the existing literature to which current and future researchers, aspiring to remedy the deficiencies in the york of their . precursors, con address themselves. Quite the contrery: " there is 2 _« gurfeit of opportunities, and only a shaky foundation of existing,

Inowledse on which to buile,

To make matters still more difficult for the prospective

researcher, there are several major questions of strstegy to be grappled with at the outset o£ cne's work, At the present

time, we simply do not know cxough even-to give sound advice on the

‘on of these questions of research stratery

ary °

most appropriate resolut Ro wonder, then, that one cannot yet narrow dowi the choj.ce of reseerc!

-topics and methods of attack to @ small renge of possibilities and

confidently cleim it is within that particular range that future

research will have the highest pay-off. . . ne of the most important strategic questions facing the . prospective reseercher is the cegree to waich he or she sheuld pursee

a "structural," as contrasted with a reduced-fomn, approach. Tne

. yamifiications of this Snitial choice may be particularly important

$£-one aspires to carry out research relevant to governnat policy

* decisions, 4 . . ° 7 To see why this may be so, the reader should ask himself.

how far the existing empirical research in this area carrics us

towards the objective of improving the formulation of cconomic policy,

N In an open cconory, the effects of sscncgery policy on capital fLlovis

may be of such jwnportance that they alone msy determine whether or

not monetary policy can be successful in facilitating the attainment

.~ of desired levels of seal activity, employment, and prices, Yet the : bulk of the onpixical rescarch surveyed in this paper does not allow

one to trace the Jimpacts ox @ chenee in a policy instrument (e.g., an

open-market sale of securities by the central bank) all the way through the financial and real secters of the econany to the ultimate effects

on the palence of payments, domestic activity, end the price level. Typically, the estimated equatzons fox capital flows contain explanatory . interest. rates ov activity variables --

thet are rather far renoved in the chain of cuusation frem the policy Snstrunents themselves. Before one can predict the jmpact of changes

| Flows, therefore, it is necessary

jn policy instruxcnts on

apite

ty ray

yranges in the policy SInstuuments on the

to predict the impect of

te e e

explanatory varie les in the estimated equations, Since this latter

prediction problom is Very Giffioult (the relevent empirical research |

a .

being either non-existent ox also at a yelatively prinitive stage),

lthe estimated equations for capital flows by themselves are of limited

_ help to policymakers.

; An appreciation of this dilemuia has led some rescerchers

- to attempt to estinate ns -Adauced-form" equations, with policy oe

eSnetruncnts appearing, e:mlicitly in the estimated equations as

rer erapran me entree, oo eer ee eae arene np ame

ya ee . os ». +» LIAB. Co Mae YEQICSSOLS, . If successful, such an approach would obviate the . . "4 . . need For an elaborate systen of structuxal cquations or the full-

. a we 6peale cconometric models that vould otherwise be required.

For example, Snteresting work along these Lines hes been

- done recently at the International Monetary Fund by Pentti Kouri, Michael Porter, and Victor AYSY« They begin their analysis with

Lt -

Ie Ios Ba

a siinayp). ed pertfolio-cajusinent model focused on aggregative economic behevior in @ smell open econcnay. By solving out the domestic

1 an estimating equation which relates

—

snterest rate, they ere left wit capital flows to various monetary policy variables, the current account in the country's belence of payments, the change in donestic - $ncome, and the change in 4 foreign interest rete. Foreign voriebles, "\ an particelar foreign interest “ztes, are assumed to ba exogenous $n the analysis; that is, the country is assumed to be small enouch _ gn relation to the rest of the yorld 50 that monetary conditions and rcal activity in the rest of the world can be taken as jndepencent of activity ani menctery conditions -in the country being studied.” As em 4

noted above, this vesearch at the I nes produced some joterostings

empirical regressions for capital flows in the balances of payments

of Geraeny, Ital, the Netherlands, and Australia.

), Porter (1972), Argy and

8 «- . sag te oe coors emseree .~ oe & - * estes ef8 © er 6 . ec ee mH OP ee

Yrom the point of view of policy, however, it can be o argued that equations of the type estimated by Kourt and Porter --

even 3£ they turn out to be yeliable © relationships 77 may not get

us much further than the traditional equations gn the literature ‘ghat purport to be closer to Nepructural™ xelacionshtps. In order to use the ewpirical yesults of, Say, Branson and Nill, hve ck, or Willer

and Wnieeen © 6 forecast the gmpacks of monetary policy on capital~

Flow aggregetcs, one requires senarete predictions of domestic -

activity and domestic interest rates, 2s well predictions of foreign =

explanatory variables, The equations of Kouri and Porter require

cep parate predictions of foreign variables, of domestic activity, end

of the current account in the balance of payments + The practical

ditferxence petween, etic two app soaches sn their present gnearnations

thus boils cown to the question of whether oF not me is in a bettcr -position to predict the current account in the palance of payments

than domestic return variables. The record of explaining and forecast@

ing $a the current neconnt belance 25 not especially nOOd,

G2 2 2 = G2 fe) oo)

“ as ve know from some of the reseacch thet is surveyed elsceuhore in

2 a this volume by Magee, Stern, and Kiecin. Another draw ack ef the

Youri-Portcr approach that mitigates its practical usefulness in mary

; - oy circumstances is its heavy reliance on the small-country assumption. Ythout an explicit relaxation of this assumption, for example, the . approach could not be applied to the United States . . e

bd .

ate

. . o +8 of mean M0Oree

ay , cot me 7 : , " 71r+5 . . ae ?

Further york needs to be done along the Jincs pursucd by

-Youri and Porter before we cen more conkid Jentl udpe the relative y 3

merits of this approe rch and its practical usefulness to policymakers.

But this work, 45 well as recent enpirical. studies that take the Mmoneterist" approach to the balance of payments @s their starting ‘point (see Section I above), do serve to emphasi7e the need far prospective researchers to be more nclfeconscious | about their rescarch

objectives =~ and in particular the balance they want to strike

between on the ong hand analyzing behavioral, st sactural rele etionships;

and on the other hand finding 4 rveduced-form short eut to prediction of the effects of policy instruments on ultimate target variables.

- The choice of emphesis between a structural oc 4 yeduced-form

sppreaes would pose 3 yo problaz, of course, if one could be confident

of a yeduced- x01 eppros sch being succe essful and reliable. With such 1

a. happy state o-

affairs, thos¢ \who hea en intrinsic interest in the

ss wnderlying economic behavior would focus

- structural yelacionsh

their efforts on these relationships, while everyone else would °

clearly cheose the shone cut of going directly to the reduced forms

But jn practice, there js substantial uncertainty as to whether ‘yeldieble short cuts #¢ etually exist. Unfortunately, jn other words, 3. aS still essentially a matter of personal hunch and judgment -7 not “of verifiable fect -- whether reliable predictions of the effects

: t . "of policy instruments on target variables can be made without

first spelling out the underlying structural models,

Ny own pexsongl bias is to doubt the existence of both

free lunches end reliable short cuts, Yet I feel sure that a Full-

blowa "structuralist" appreach to the empirical study of international

capita) flows wil). take years, if not decades, to cone Lo xyeal fruition,

Japortant pelicy Gecisions will continue to be mede in the meantime “=

for the most part, in a glass darkly. Under the circumstances, it seens fully justifiable -- af least for a.further exploratory period

of several years -- to heve the available resources dispersed all along the spectrua fren the structural miccroeconanic extrene to the

: . . . \ . : others extreme of highty-sgsregeteve yecuced=fomn CGuacions « .

Regareloszs ef the balence one strikes between 2 structural

-and reduced-form approach, but especially if one is inclined to

emphasice structural) relationships, a researcher in this area must

resolve ancther major question of strategy at the outset of his work:

‘

what is the appropriate Gooree and type of dicegstegetion?™’/ As in

‘the preceding case, one's ensver to this questica will pertly depend

on the specific objectives of the research; if ome has an intrinsic

oe . 8 oo cer en aes aul Onna A i! ss . . . . , .

a/ Leamer and Stern (1972) discuss this question an their survey of the probiems in the theory and empirical estimation of ¢ Snternational capital movements, .

interest in the behavior of capital flous generated ‘by commercial

banks, fox example, if 15 obviously essential to disace ryegate the capital-flo-s data csufficientiy to eliminate changes in the assets ~

and’ liabilities of nonbanks. But suppose one's primary objective is to obtain, for some given country, predictions of the capital acccunt

4s a whole. At what point | go the additional costs associated with

further disaggregat ston come to outreigh the smeremontal banefits

+O .

(i£ such benefits actually mate erjalize) of more precise end relizple By 7 predictions? . . . .. . e . a

s : "Ls _ The question of the eppropriate dearee of disaggregation

of aggregative date turns out to be, in practice, snseparably related

to dccisions about the apprapyiate Lype of diseag : \

three mein concepiucl pessi bi watiow by “Erausactyy

1-8 tee ur. o 17) a v w x

v2 14 (2

: (type of decisson-Zaking unit}, by seographical region (point of origin

or destination of the capital flows), ox by instrument (chara acteristics

: - of the assets oF Liebilities being oxchenzed)s

A moment's vocoljection will make it evident that che a

theoretical frameuork spelicd out at the pesginning 0 o£ Sect} fon IT e@bov

“presupposes all three typcs of disaggregation, since St assumes the

. - existence of detailed data on the balance sheets and incone eccounts of

: gndividual econo sco units, In practice, yesearchers in this ares scldcs

can acquixe such data for indivicvel econenic units, And, to repaat, eve"

$f these data vere available, one's objective will often be to analy4¢ o

a oD Contrast, +07 onaiay! c, tive approaches of Kwaick (19712, L7T1b, We! " and Verrins (373), both of view wish to produce ¢s timares for me , an U.S, capital eeceunt af 2 Wiolc. Rerring empleys 4 single Pree

‘Foy the entire eapital aecemme (ined wsing U.S. Goverment epi ee mS and direes Invectasny ). yvack's model 3s the most diseanres ne bes of any that dewk vdiuh the entire Gays.tel account,

ah et ae ? western ye ame “ae

2 Se meen name NE eee ed

occas page em fo .

11-8 8 .

. if

—

predict capital flows at a more aggreuative level. ‘To make one!

research permane to policy decisions, it is certainly the case that

-conclysi.0ons must have a bearing on = aparegative data, lence the familiar problems of all empirical research: the microthcory of

gnidinvi.dual behavior has to be supplemented \ with wanerous additional

: . assumptions and questionabl.c approximetions before it is possible to cazxy euk any CEP oiricel testing at all. ce - When 6ata are agarogatad over a set of relatively L, gmogencous

economic units, the hope is that ean aggregate function woich takes

the sane general form es the individual funetions will not introduce,

an unacceptably high ecgree of aggres ation bias. The more hetexogencous

the individual econaaic units in the group and the more heterogeneous

the economic environments it sich they operate, however, the Less

‘confidently one can hold to such a hopce

The rationale Lor sone significant amount of disaggregation

of aggregative date “by type of decision-msking unit is strong. Comnerei.s

banks, nonbenk gsnancial institutions, nonfinancial corporations,

“~%

LACS

governmcat agencies, and housgholis, for exanple, all constraints and regulatory environacnrts.s Differences in theix asset

“preferences OF other aspects of their econoini.c behavior may be quite

. 4mportant. Moreover, soue disaggregation by type of transactor may 7 “be essential if one is to mateh data for capital flows with appropriats ° “gata for the yclevant economic detemnants (c.f, the scale vé ariable : S)e 7 : . ° ra ; 7 , : |

sree nen ee =

by geogrepeicat region.

. LI1-9 | por,

. > Some of the greatest disparities among, economs.c units

. . 1 ‘

and econoni.c environmaies ate those created by or embodied in national political boundaries, It ss cextainly the case that the nature and tining of ceonomi.c pelicies s of national governments | -~ ‘to take the

most pertinent exarip)c, “policies intended directly to influence

Snternetions) capita) Tiovs > differ markedly across countries.

£55

Even in inscences eheore it is icss cife cult to swallow one's doubts about heterogeneity y of environments acress countries, there are no easy, clearly appropriate weys to cons solidate econonic conditions

across a group of countries into a few sumaary variates 2/ These

consideratiens arous strongly for sone sien} £Licant desree of disegsrcg eaticr o> -¥ ca Cc ae

tes¢

. Some Gata on Sutermecicenel capital flows are collected by s « bd type of financial instrvment, with Little ox no cross-classification

“by type of transectoy Ct geographical region. Worcover, published

. : os ° rn | data for most CouUnlyies send to be in this format. Thus a researcnct s

‘ . a J Fox example, 29 the study of any sincn capital flow, what weights

should be usec to construct weigntc GeHaverxvaye variables for GHP,

- ox monetary ageregarcs, er short-term jpterest rates in the

countries of cne Europenn Econo: sie Community? We axe of course back again in tie swamp of jndex-number problems as soon 3S . we ask such quest 10nS.

ao *

nT GGFR. prnngerty : 480 eee renee, _

necd, based on his theoretical framevork, For sone Aiseggregation

along this third line may weld be satisfied. i. Dot, Even if adequate data vere availeble to allow a significant

dezree of disaggregation 2long all three lines sinultancously (typically

“they are not), researchers gn this axea would still be faced with an

enormously aifficult judgment. To put the matter bluntly, the

cality con£ilict

tte

requirements of theory end the dictates of pract

violently. And TI know of no easy cenexe Lizati ons with which to

help, or confort, those who must make the aykyard compromises +

.. The problea of the aveilability and reliability of data deserves 2 conscnt gn its own yvight. For here, toc,, prospective

hd “ge rare * wwe are ey ate “A e- 2. oy -< -- PEsed Lens yusu weESoIVe far thenseives a stryategs+ GUeGGeLCN how ad

much effort to devote to data collection per SE. Trere is of course the inevitable eifficulty that Meaatever We measuye is never quite

the thing that represents, oF closely corresponds tna, the theoretical

7 concept of ovr theories" (tachlup, 1972), Roneche:ces, meny researchess

fot)

rr Pee ts Ck a S set BL ea ed St for

may lean too havily on this fact of life 4s

avoiding the effort to acquire snetitutional knows..dge and for devotins

. e

As sugessced carlier, St is the Netructural™ pirast who most

agonizes over these inevitable comora:nises. “ire pure monetarist .. who aspires te explain the entire balance of payments with 2 - a

’ single yecuecd= sem equation may feel he has mireled the > dilence entirely. Everyone else must suffer te a greater or

lesser cegrec somewhere in between.

Lowe ee

1-11 , Hoe ,

“as Little tire as possible to the tedious job of carcful examination

of all potential sources of data. Yet: innovative, diligent searching can often turn Up snfoxmation that will either excatly improve a data

set or, alternatively, prevent an investigator from drawing unwarranted

“- inferences from statistics that purpoxt to measure one thing and 4 .

actually measure something clse.—

‘Careful serekiny and compilation of the deta are especially

smportant in the field of international capital flows because far too

° feo data are available anda much of what i's available is of poor > : P

quality. Aay researcher who has used palanca-of-payszents data for

the United States in an empirical, study but hes not read the reper 1d be put “e . oy? 4. . . 41 an 1 os 2! =

wna straitjacket until he has done sO 77 peraayps tuice.—- Yora

. of the Bernstein Comittee (Review Committee, 1965) shou

study that discusses data problems in some detail and attempts to

construct an integrated matrix of capital Llous across individveal countries for the 1950-54 period, sec Michael (1971). Smith (1957)

gs another paper that highlights problens with the consistency and

reliability of the data. For sLlustrative cnemples oe: s

~ rs

data problenis sa the context of an empirical study, see the discussion

in Bryant and Nendershott (1970, Appendices y and c).

pe 1/. Compare vachlup's Novumble" (1972, P- 4): “too often researchers - do not question eho meanings of the terms with which they WOTES they are diving into piles of third-rate statistical data which

they believe, OF assume, to be suitabie proxies for the vague ec ° . ambi gucus concepts with which their supposedly first-rate models are furnished." , . 2! ‘Por anether useful docaunent, thet contains sone of the details Ginc hud iy, YEPEveing forme) of how ULS. data on internation! capital Jlews are collected and processed, see the guide te internstional finane ial cratintses compiled by the yedoral erro ae Yank of At lants CVor2). ve

men: ~~ we rasta oy: _ TRE eR . oP wep wrens

. . 4 .

ed ak ee, cee ned “N grew Se letheoe is eg Pe ee ee we ee Ae meee 7 Lue Ne ete + Re me moe

' emp

‘ - ne allie “encase ‘ .

: . . 4 . .

sttle doubt in my mind whether

- There is more than 4 1

° e ° = : ¥s \:

} > oa - be Od . . 20 P en ant wht n

v oO . cd . .

a way fo squeeze blocd out of a turnip. a

. . . ; . ° . ; . | . ° ) . : ; : | % . ‘ . . . - . . . . . . . . . et , . ° ° . . ° . . . . . . . -. . eo. . ; . . . -. . - . ; .? ; 2 . ; . . . . . . ' . i . . : . ‘. . - - . ° . - . ; : . - ; . . . . : ; ° . . a . . ; . ° . . . ; ; . . - ° . " . . . . | . * . . . . - . . fe . °. . . . ° ° ‘, tet -* . - 4 . . ; . .

(1972), the studics of Willer and Whitman (1972) (1970b), those by

| 1V-1 ke { --

IV. Concluding Co Comment: How Tar Yocs Peisting Erypbirical Knowl edie

‘Toke “Us? JERE bee, .

t . . a wel, ee

7" One way to assess the empirical research in the area

surveyed in this paper is to ask: How inuch do we know nov that we did not know, say, fifteen years apo? My summary answer to this

question is that ve do have some useful knowledge we did not have

then; but we still know much less than ve need or would like to

know. a - . .

_One thing about which we have (eeSinitely learned something

- .

is the responsiveness of international ca pital flows ee) changes in

mnonctary conditions. There is ample evidence from a number of recent studies thet capital flows are guite interest clastic. Representetive 7 : : results may be found jn Branson and Will (1972), Reack's research : |

Bryant ena HWendershott (1972), the peper by Nellivell and Masasell on

- Canadien capital flews (1972p), Black's recent vork (1975), and her ring's ‘Ph.D. dissextation (1973). re . -

‘The evidence from these recent studies contrasts sharply with

the conclusions ad awn fro sone of the earliest research in this

area. Reither Bell 1952) nor Kenen (1963 for example, thought . ; ’? j -% b

‘that changes in interest rates would produce sinable juapacts on the capital account. Today, as Kenen notes in the introducscion to his 1973 monograph, there tends to be little dispere: “scholars and: practitioners alike are theroughly persusced that Lonctary policy

has an irgvediate, pervasive effect on private eay ital MoVeMe nts!

(1973, p. 3). .

ET ee emer rere en eon te me an re - :

Branson and Will, and Kwack, have carried their work far

. enough to be able to put forward estimates of the total net effect

on the overall U.S. palance of payments of a change sn relative monetary conditions an the United States and abroad. Herring, following the lead of Branson and Nill, has attempted the ambitious

objective of estimating equations for both interest rates and the

net capital account for six major sndugtxial countries. In his -

ry .

———eooroooo

. .

i/ For an overview of Haack’ s quantitative results, sec Ywack and Schink (1972), especially PP. 16-27. For an overview of the quantitative results for the United States jn Branson and Hill (i971), see the table of stock-shift multipliers en pase 25 and the associated discussion in the tent. The following quotation Sadicates the order of magnitude of the pranson-iill : estimated effects on the net finencial private capital account “e .Quy) es a whole: "For example, a one-poing inerease in the U.S. short-term rate, by stself, will ter -orarily seduce the KA deficit (or inercasc the surplus) by @ one-time stock-shitt effect of $2.6 billion over three quarters. With assets erowe c3 5

o sng at 2bout 7 percent per years this would give a con nuings jmproveinent in the KA balance equal snitially to $180 ni. Llion “at an annu2), rete. Buc if the Evrocoliar rate adjusts point-

for-point to movements in the U.S. rate, the total coefficient of a one-pot chance it rhe U.S. rate on KA is $1.5 (= 2.6-1.1) billion, elude the effact cf a one-peint change in the Rurozetter rate yoelative to the U.S. rate is $1.1 billion. Yhus in this case the one-point increase in the U.S. biJl rate would give 2 $1.5 ‘billion net stoeck-shist effect over three quarters, with an jni.tial continuing- flow effect of about $77 million at er ennvc li rate. Tf the W.S. bill rate increase was accompanied by an gnerease sn velocity of about 0.1, not unreasonable sn U.S. experic: another 6602 mi}lion vould be added to the stack-shift effect, and $42 million to the initial level of the cont inuing- flow." For an gllustrative discussion of how such results might be - used for projections of the capital account, see Branson (1.879, pp. 253-257); note, however, that the empirical results used jn this 1970 article were superceded by the later estimates in Branson and Will (1971).

banks from the united Stat

» . oe . , ° ’ . . , . e ° ° - ae ; ~IVA3 . . , a. ~s t ; ‘. conc}uding chapter, Herring Nas oud.iled these - Ts

stimates together so as to describe, albeit very roughly, the Linkages that (part5a125) integrate ne tional financial markers

gnto the beginnings of 2 world market for moncy and capital.

Although all these eetimales are subject to the. problems noted jn Section II, and hence pt bably subject to very large margins cf error, they aonatheless represent the best guantit ative evidence ve So far have cn the cntexrelationships between international capital flows and sariena) monetary conditions.

The evidence ts also fairly strong that many sndividual capital flows are associated with trade flows of one sort or another. “pending >¥ hanks oY sntew-company credits of ten have an

jmportant element. of trace financing. For examples, note the

relationship between Japanese imports and barxvGwsns by Japanes

tudicd in pryant ane Rencershett (1970, 1972), ox the dcpencence of several components of the U.S-

capital necount on trade Fiews oF the trade belance in vuack's model

~ (1972). Oa the subject ef the relationship hecween capital moverents

and trade flows, sec also Sent Hansen (1961). .

ec’ £8

We know, rather less about the relationships between capital

dé

‘flows and {luctuations in growth and business activity: it js trove

that variables yopresentins the Level of oF fluctuations 4 business

i .

u/ Yerring estimated 2S single, agyrega tive equation for the capital

faccounl 2S a whole (for the United States &s well as for each

other country). An overview af his results ean be found in Chapter 7, PP: 267-315; cee especrelly tables 7.4 and 7.95

| LV-4 7 wy

activity appear in many of the empirical equations that have been

e ° .

estimated. But the interpretation of these results is often

‘strained. For éxemp)e, as noted above, Miller and Whitman regard

“deviations from trend jin business activity as a proxy for risk.

- ‘Others interpret real activity variables as saying something . . about the demand for money (Zouri and Porter, 1972a). In my

view, it my be some time before ve will know how to interpret

the correlztions that have been observed in the actual empirical

work already done,

J have already stressed in Section I] above the failure

of equation specifications adequately to -incorporete the effects

of exchange-rate changes. It is hardly surprising. thereiore,

that we have 2 very inverfect vagerstanding of tiie relatienships between capital flows, spot and forware exchange rates, and exchange- ‘yate expectations. that Kmowledge we do have is mostly qualitative

and tends to come, not from empirical research, but from the

exploration of thcoeretical wedels. The main exceptions to this

sgpeneralization, as already neted, are Black (1973) and Nellivell

. 1/ . _-and his collaborators (1971, 1972b)~ .«

Iam scill less confident that we have useful knowledge .-

. ‘about the time pattern of all the various responses of capital .- Se nnanaas .

1/ Bleck's analysis of and inferences about private speculative behavier (1973, pre 30.52) are espacially Interesting. He

. — eonclites chat the fienible exchanpe markets of 1936 to 1938 - ‘toere net destabilizing. In most cases, they facdlirated tic respense of participants in Lie markets. to the deshabilizins political events of the period" (p. 53).

ene eer

ow ee we a

2 eee es eee me ~

‘VRAD -

flows to their econonic detonsinants. The estimation of lagacd

responses $s still very difficult and. problematical jn all

empirical research. Yut it is a fair judgment on the vast majority of the work done so far in the area of international capital flows

to call it unsophisticated relative to advanced empirical work in

other areas. The existing studies often resort to erude techniques

‘such as the inctusion of the lesged dependent variable in the.

equation.~ yne potential pitfeils associated with this formulation

. 2/ . woes . are nocvorious.~ Quite apart from problems of statistical bias,

‘ :

@ . ° , - ° e. e « inclusioa of the lagged dependent variable’ in a regress20n equation

can generate superficially plausible results cven though the basic

economic relationship 3s seriously mis-specified. (in an extreme

C&aSC 3 meet wal cn min deoen Shere: ent sens ab Sie wan TERE feos wa thar @Qacy, BGQCa wer’ or coyes- Wren tra Paes Sf '2t 2S oe emer . .

- smooth economic time series ~~ cun alone “Yoxplain” the dependent

variable quite well, suggesting a “significant slow speed of

adjustment even when there ere no theoretically appropriate variables

in the relationship to which the dependent variable is allegedly &

. . adiusting.2/) Jn other cases, yesuureners have wade use of olynonira? .- @ ¢

— gnterpolation techniques, but agein in ways that are rather ad hoc

“and that do not give one much fecling of confidence that the actual

results can be relied on.

——_—_—__——-

Mu Sec, for oxampic, Arndt (1988) and the Miller and Whitman studies.

-. Of For a discussion of the state of the art as of 1266, see 2. b]

Griliches useful survey (1967). 3/ Bryant and V andervrshort 97: 2) give sane s1ius strations of the . consequcnces of migcspecd fee eyon of cauations Sncluding, a

Jaggce dep: mdeut vardable. af Brassen' s carly werk CAGES}. for exanple, made extensive use of the polynesina | interpeiac ton teehniane developed by snirtbey

Alinov. . ; .

yer en tt ee Lhe oe

‘ , . ‘ . , ¢ a \

We are in the poores st position of all to appraise the

effects of novernmental nagtrietions on capital flows; for two

yeasons. First, as dsscussed in Section Il above, the techni.qucs q

used to stusy capital Flows directly subject fo governmental

restraints are ina dequate- We therefore know Little about the quantitative impacts of the restraints even on those gpecific capital flows at shich the vostreints are auinec. Second -7

and this $s just as uch of a difficulty when ve are, inter yested in the net effects of the restvictions on 4 country's capital account

or balance of payments as a whole ~~ we know virtually nothing

about the quantitative impacts of the controls on capital flows

other: than thore Spee te flows ct whieh rhe cenfrols aye aimed. rhe Conky Ct

other than Se ific f20NS Ss — Substitutions OF “Nyoakages" cf funds through yon-conty -clled. channets are typically 2n jmportant pheacmenon ~~ teo important te be safely : . Sgnored. Yet by and large they have becn ignored in the empiricel Literature so far. . :

r the Fol les ing cxonples per yta inns to the effects of tricriens imposed by the ULE. savernment. Because of, the ynterest Equ iligzatien Tax GET), foreigners purchased foreign securit i¢s issued outside the vaited States that other~ wise would have been issucd in New Yous end bought by uU.Sresidents. Similarly, because of the ;cegran administered for direct jnavestors by the Commerce Departrwent office of Foraign Direct Tnavestne ants (OrDT), fore dgnars: purchased a greatly jrereaged volun 12 Of securities issucd outside the United Srates py U.S. direct inv wstors; absent the OFpt pragrar. “direct investors Nowld prebably have bor rroved much of thet needed funds fron U.S. residents. But to what extent did foreign INVvEstors therefore buy Gower US. securities than they other ise sould have purchased? Cjhere were of course HO contrels yeaMervies foreigners Not fo reduce the rate at wien they ° thourine would have secu: faced gacn securities 5.) hyde and Hill Qe7l. pp. 1O-20) do not even hether to test for vhit possibi): ity pa ghehy equarian for U.S. sauyes born piabiLetdes ve foreigners. Analogous problers # aydge iM connecu ben vith the Covent tum ee tee pe ; (cont ined an nent pane)

ro rom oo et eee ge a ~

1 OSE EERE IPT Tot eT ERENT ore rem. neenod: a 1 . 7 a Eee ~ oe oe a a ene,

See ya, Tere NT eee oe Oh? #288 owe — ~ at aid qn eae ee — hie sented wee sane: were: Zee

These Lars gn whak we know

capital. controls would pote t

controls were more the’ exception than the rule, ;

the controls *%

7 howevey such is not the case, § d - Ledge | are Serious: he os AL cnous? the 2 empir 23.00}. yosearch in rhis .2ree has produce™

panded: bat vseft additions to our ienenrh oe

goes not carry Us very fat towards the

the formulation ‘of economic policy and monetary policy in - a,

1/ parcicular Tt is all very Sell, For “exemple s to note that

capitel flows ate sonsitive ro changes in gnterest yatcss

jronce that wonetary policy can have @ poser ful snsluenc® en . .

private capiral movements - But cen this snfluence and the other

‘gLous be gpelied out in @ way

By and 1arée> they cat not:

‘what

yelunt ay y rorel ten veeedit R estraint (YTCR) pr ogran vestrainé oth gneLuce a

gng Wee e, banks: pranson and wih 5 ani Wyte mk exp hi 2inINs U.S.

. . ° Joans bo fore Gr won ident’ “put discuss the poss ibs OES

chat goreianer® reduced thor depos 53.058 sn U.S. panks (ect pare athers2se would heve peen) a5 an emai rect rid

with what qhey . of the ywech pr scyy aU i On tls posne. 1s0 the Jiseussien above on “pe

[seck on 1}, Pre 2- -G/.

“ .

Cxeak

me ee

‘when it is taking place) than the continuing-flow effect.~

"rV-8 oe a me

: To be sure, the existing empirical evidence puts a

modest amount of flesh on the bones of the s -called “portLolio

adjustment" view of cepital moverents. this view, it will be

yemenbered ,~ argues that, Sn response to a change 2n monetary

conditions or some other determinant of a given capital flow,

’ . there is a stock-shift effect which, thoug ch spaced out over as:

much as several quarters ox maybe a year or longer, represents

a once-for-all adjustment in portfolios -- and therefore gives .

rise to a onee-for-all capital flow. (Under a regine of £lexible

exchange rates, the stock-shift effect gives rise to a once~for-all

change in the level of the exchange rate.y Simultancously, the

change in ‘monetary conditions jnduces a continuing: flou adjustment (an altered patter of snvesting inerements to sealth) that persists

more ox Less indefinitely. The better studies in the field recently

have all been based on fomulations which allow both types of effects,

_and the resulting empirical estimates support the view that the size

of the steck-shift effect is a good deal larger (during the period 2/

~Yhis latter point, taken together with the evidence of the

relatively high interes: elasticity of capital flows, ‘has important

“gmp)ications for the use of monetary policy to influence the capital

Se ° . .

1/ See also the discussion above on pp. Tsection IL, PP- 6-S/, 2/ Strictly spec Ling, Uke existing, capirieal estimates do not constitute & test of the ne tfolse= adjustment view of capital

movenents; bie specifications employed force this theory on Lhe data. . .

exchange rates 2re floating is that monetary policy should be

IV-9 - . . . .

account in the balance of payments. Je sugecsts that monetary

a eee

‘capital account and/or the exchange rate. ‘yhus, with ey.change

.xates pegged at a ecxtain level, monetary policy -- 3£ used for

external objectives -~ should be thought of mainly as an instrument

for bringing about a one-time change in reserves (compared with

what they othervise would have been). Te can be much less

effective in bringing about an enduring chenge in the capital

account. The counterpart of this ay gument for the case when

regarded primarily as an jnstrument Lor {Inducing a one tiiac .

shift. in the level of the exchsnee rate (con ared with what it

othexwise ould have been).

Although important, these conclusions about the impacts

of monetary policy on copital flows and exchange vetes tell

policymakers very Little of what they need to know. Consider the

following representative illustration of the choices anc uncertainties

‘that confrent policymakers in the (United States. ‘Suppose the Feéere)

Reserve is contemplating the nossibslity of a tis atening of

monctary policy via open-market sales of Treasury bilis. Suppose

further that a proposal is being considered to lever the Regulation i

_and D xeserve requirements on Eurodollar borrowings by U.S. banks. In

-additiion to the need to project the effects of these preposed actions

: ° ; Tv-10 le . a

on real activity, coploysent , TrivEsy ens financial conditions,

- ve sal .

the Federal Reserve neces to be able to ‘project their net. effects

on the balance of payments and the value of the dollar in exchange 3./ | - os . .

markets 7 . , ° wf ee an ‘ s

What analytical steps are required? In principle, one ye st project the effects of the proposed actions on the portfolio

behavicr of ail categorics of cconoric wnits and thus on the whole

structure of interest rates and credit availabilities in U.S.

1/ Yor the purpese ef the discussion in the text, “the motives

. undexlying these cont empla red actions are not relevent, since «> whatever the motives -7 judaments have to be rede toe about the effects of the nections on all the main macrcet sonomic warjabics. Some of Lie pittabsens chad magnet eonceavatey

tude the followings:

give rise ts such propese ed actions in ‘ly, denastic demand 1s

(2) Prices Have been rising too rapic

thought to be expenc eing at too high a rate, shile the balance ‘of payments has been weak (or, alternatir rely, the dollar nes . been depreciating on exchange markets). (b) Domestic demans ; ig judged net to ha been growing at an exces ssive rate, but

the balance of mayen or the exchange ve alue of the dollar

i1tS have been very wealt. (ce) Domestic demand has been expe nang . . execs se but the balance of payments or the exchange . malue of the dollar we not been weak; in this cuse the 3

he fogulation Li and D actions mipghk be under consideration primaxdly for regulatery reasons (c-8-, tO alter ‘competitive jnequities between U.S. banks and foreign banks), and the ere would be a concern avout both actions strengthening the cess

external py sition exc ively. . \ . . .

assets and Jiabilitices of U.S. benks, and hence on the Evrocurrency a

yV-11 s _ \ fanaricia). rarkets. These changes in financial conditions will

have an inf}uence on domestic activity and prices, which then in

turn will feed back and jnduce further changes in financial

conditions. With the United States such an impertant part of

the world econony ; activity, prices, and financial condi.ti.ons

& . . . gn the United States will have significant jmpacts on, for

example, gnterest rates and activity Levels in other countrics.

LE exchange retes are being held yelativety fixed by off cs$cjal ay intervention, | these impacts will be transmitted jn part througo ‘ . . changes 319 reserves. In the absence of significant official sntervention in exchange rarkets, exchange yates themselves will change and help cransrs.t che impacts. In &ny case, it is clear

° .

that a large number of internationst channels of causation and

feedback oops need to be taken into. aecount. The propesed change

in reserve requirements en Euroe Jollar port owings, for example, might, have particularly {mportant impacts on the international

‘markets and then on conditions in financial centers in other major

: : a countries. The analysis has to be able to cope on the onc hand

“with the many influences OF economic activity and financial

conditions , poth in the United states and abroad, on capital flows and on trade” flows, and on the other hand with the many fecdback

influences of capital and trade Flows on economic activity end

financial conditions. | \ \ e _ ° ‘ a

world move? Economists’ lack of cuantitative knowledge of the . vy &

oe ioe eA

. ae et . . tote ‘ ‘ . . . . . . oe Seen against this background of complex, interdependent | . i . .

. on a . . transactions, how artificial an intellectual experiment it is for . ° ' .

the Federal Reserve to ask how much the capital account of the U.S.

‘balance of payments vould improve as a result of an increase of,

-say, 100 basis points in U.S. interest rates with interest rates and

‘other financial conditions in the rest of the vyorld remaining unchanecd, a a TR ee

Yor any large, open economy -- but a fortiori for the United States --

it can net be plausibly essumed that the rest of the world will stand ?

still. But in what manner, and by how much, wili the rest of the

id . 8 . e

determinants of international capital flows is even exceeded by our

quantitative ignorance abovt the manner in which capital flows link

nattonal financial markets together. 7 . - oy

. . - . ° .

- .

e . .

er em .- . .

i/ Nerring's Ph.D. dissertation (1973), which I have alreacy - mentioned, is @ cosmendsbdle first cffort to tackle this area of ignorance. On the same subject, see also Argy and Hoc jer (1973). In an important recent paper, Girton and Nenderson (1973) investigate these problems in Cerms of a two-country - ‘theoretical model. : °

. es ,

"If Nerring's empirical results are taken at face value, they suggest that 2a tightening ef menctary policy in the United States so as to bring about @ 100 basis point increase in (U.S. interest rates vould exert so much upvard pressure on other national interest rates that a weighted-average interest

. diffexentie) between the United States and the rest of the

world would be increased by only 36 basis points (Chapter 7,

* Wable 7.6)... Kervina's ageregative capital-Llow equations yield the result thst a 100 basis point increase in U.S. interest rates, in the absence of any adjustment in foreign interest rates, would produce a $35 billion capital inflow jnco the United States; the

-resulting, infles to the United States wourd be only $1% billion when jt is assumed that forcign interest rates adjust to their new equilibrium levels (as calculated in Neryine's interest-rate

equations). To huve ceasiderable misgivings, as docs ligrryine hivtae: ‘about his specific estimates. But they do constitute a sien yorves: on a path, as yoo wanarked and uncleared, along which many more

researchers will bewe toa tread.

~~ wens rf

ee ae

Even the meage? theore:icet and empirical knowledge we do have is

IVv-13

“just to complete the picture, there is the siaj.J, matter { of having sanething less than a perfect understanding of the linkages and snterdependence between financial conditions , real activity, and the price level -- for the United States, oF for any other economy. the unfortunate truth, therefore, js that the economics profession docs not have sufficient vnowledpe -7 about alinost all of the

enalytical steps chat must be carried out -- £0 enable the Fedcral >

Reserve OF any other monetary authority te sake reasonably 2ccurate

_ projections of the fixst-round and ultimate effects of any of their

proposed ‘ections..

I yould not waat the preceding remarks to be nisinterpretcd.

often sufficient te prevent policyekess prow Falling Inte Gross

“errors ~~ such 4S, for example, Sgnoring the key snterrelationships

between gorestic monetary conditions ‘ane monetary cond3..3.0ns abroad,

or expecting an alteration in the yelative Yevels of domestic ana uo

forcign interest rates to have as large an enduring impact on capital

flows OY changes 3A exchange rates os the initial, stock-acjustmant

. . .

agnostic as to dowhbt the value of, a major expenditure of research resources jn this area to advance our knowledge further.

that J am arguing, however, docs ameunt to an overall

: , judgment of the empirical literature on international capital flows that is quite critical. Viewed another fiftecn years Lrom now, ec ° e

nee pene ees

a & ~ 4 . . . . . . LV-14 8 | . a | hi ca of 7 j Jat by 1973 this ar 1 te surely Leck back and s¢e that by no } | 5 almost $ - — | e | £ no! c t best beginn2ns to strug! oe t in its infancy, was a ‘ ° $f not in 2 inl “ ae | judy since that rc : ude, 8 I leave it to the reader CO JUCLr> srt, adolescence. . ) : ; _ - ejme preferences | “sve and rate of t) | «spective and | a j " ry own persp —_ ads on his or he | * | — | Snictic aSSCSSMeilt « imisti x pessunist . | his is basically an optimistic oF ay Soy this is basied~ ; | whether - . | | . 7 ; ; : . . | | | . | | . | | cod te, : . | ° ry ". . . . ° . y

R-). a

al Movements: Theory and

Ainono, Aihico, Menternational Capit Estimation," in Rd. Ball (ed.), Internationa) Linkare < of Econonuc viodels 5 (Amsterdam: North-Holland, 1973).

Rats jong) eas ese Bi Balance

An Econs: wnetric Model oF Ene of ? Parments (Yor theomings in the ee oarribucions to Economic. ‘fnalysis series of North-Holland } Publishing Company)

Hovember, 1975.

Awnane, Akihiro,

Victor and 7. Hodjex MPinencial Tnteg avation and Inter 1958- 71 " ne Sta

Rate Linkages in the Lndustxyial Countrics, 20, No. 1 (March 1973).

Papers, Vol. s and the -

Argy,

Argy, V« and Kouri, P.eJdeRes Usrers lization Policies Vola tELaty ay International Res serves, ' Fortheaning in RZ. Aliber, et., Naticnal Monelere Policies and the Treerneti onal Pinancial &ysies any (cingepeeas Conrercnce, 972). ; . hacl CG. Porter, "The Forvard Exchenge forhket s Under

and Wie i cxnal Disturdances

“heey, Victor, Bowmestic and Ext Papers, Vol. 19,

and the Effects oF Alternative Exchange nate Systeins,") Tie Scaki No. 3 (November 1.972). Arndt, Sven Ve, “Wpternational Shor c-cerm capital Movements: A Distributes Lag Mtedol of Speculation in Foreien. Exchange," ’ Feoneaztrica, Vol, 3& (January: L9G8). . . * Ball, R.J-, eG. TInternvetions al Vinkace 0? of Ng National ec Eeononke Be Mode le . (Amsterdain: North-Holland, “{973). eo e . nord Hinanesal Constraints in

G, and J. Wae elbrocck, esented at 1973 annual

Rasevi, Project Link," unpublished paper pr

meetings of LINK.

. . . \ ° . : J . pei) ye; Mpysvate Capi.cal HWovements ‘and the US. Palance of | "oy eckers Arfocting tl the United States > Balance of

payments » FBCLOVS fo ne VAs Joine Econonic . Gomithee, 87th “Congyess, Znd Session

y: aynuntS Gy “(8 , GOVE wnment Printing Office, 1962).

. . : Plack, Stanicy W. , Theor -y and Policy Analysis _ of Short-term pavements intl the ne Balenee of | Payments, unpublis shed Ph.D. disser tacion Yaie Univ exrsicy, 1965, abridecd version . published jn Yale poonomic Essays, Vol. & (Spring 1968). .

Wan Econometric Study of Euro-dollay Borrowing, ad

pyack, SV. Rate az Tuterest ou Turo- coll cam

. Yess YorR Panks and the . ne Journé 2. cf Vineness: Vol, 26, No. 4 (ts ven 3973) ae ——oOooOoO sO lack, Stanley We, Internets Zonal Honey 1 Markets s_and Fl exible . rhe national Finance

. pxehange Rates s princeton Study jn In Lxeh Sse . GJ c No. 32. Princeten University Tress, Narch 1973. + NA Prelinminery Analysis of

Renton, tn Private Long-term fapital,

hoatwright, B.D. and G.A.- UK. Internationa Weansaccions

Piseussien Paner Ho. 3, Econangty jc Yorecasting Unit, Lendon : Graduate Schoo) of Business Studscs (undated: 19722).

Pradshes, Robert C., A Theratical and Empivical - Analys £ Yank RNuro-dolisy Torrovins, 1eo6-71, Univers3 cy of sary land

ene wee 0 err nee ee

rnd. dissertation, i973.

Yrainare, Williss C., and Tobin, James, "Pit fells in Financial Model-Building," Parers and Wroceccinn’ of the Ame rican ,

Tieononis hesociation, Vor. IVUET Gitay Less aS).

a .

" syanson, William i., Financial | Capital F Flows 3m the U.S. Bal ance 1968}.

‘of Payrents, (Amst tervdan: Nor thei Noihand Publishing Co-,

. - _ Tivansoa, Wel, Monetary Policy and the Rew View of im ternational . Capital. Movements" Brookings Papers CN Keonemic Activity, Ko. 2, 1970.

Branson, W.1. and W1), RD, “Gaps tal Hovements in the rcp Area: ECD Economic Ove Jock, Gceas jonal

. An Econometuie Analysis " Studies, (Wee. 1271). . . ; . r) ~ ~ , cepabidihee ide eae . aes .

Ve as .

. °

‘policy Tovar d Short- -¢erm the Portfolio Approach,' al Mobility and fove-

Poranson, W.le and Willett, T.D., —_ Capital, Movements: Sow Taplicetions of gn Y. Machlup ect. al. (eds.), Jnicen nationa ment of Capital, Ketional hurea of Econanai.c “Research (tew York; —_—_—_— ~—- — Pd Columbia University ?ress, 1972). . .

Bryant, Ralph ¢. and patric NW. Nendershott, Financial ( Capital . . Flovs in the o Ualance of Ve avments of th the Mniter_$' States: AN oo Pxplor ater “~~ Dmoirical Le Princeton Studies in Internatione al Y3nanee No. 25. Princeton, 1970.

rshott, “Empirical Analysis . Alternative Speci rics WL.OnS,

oO}. and Movement

a 2 ernational Mobility and }

rajeh C, and Batiic U. Hende

Pete Flows: Some Censegecnce

in ¥, Mechlup et. al. (eds.), Ente yoi) ¢ Rational Bureau of Heononle Kese arch (Sev York: press, 1972).

The ‘he Ince rnational J Remand _{ for Liguid Dellar

. Bryant , Ralph Ce, Assets, unpublis shee m2 nuseript, S71. .

Carry, J.L- and J.A. Sawyer, Wphe Palance cf Trternational Payments and the Foreign Exchange in TRACES Meck {IIN," Lvscitute for > fraivsis of Social and Economic Policy, Univer t, 1973).

the Quanta tat Ve areal You ity of Yoronto Cenpublis shed wa

Caves, Richard Capital Transfers . and Leong Le i Vniversity Press, » 197. . .

E., Grant L. Reuber, and others, Poljcw: Canada, 1951-1962, Cambriese, Harvard

"international Capital Flexible ana Fixed te

: Pins Lee

, Richard b., and Grant L. Reuber, and Canadian Eeonory Yolicy Under . Exchange Rates, 1951-3970," in Canadians “United States

- . mo TF hereerenener a rena : Relationshius (Proceedings of a P EONECTEENCE | held in September 17h}

. . Federal he Reserve pansn cf Boston, 1972, pp. 0-40,

. Cave os ° 2S Hart: OLS

Impact of the U.S.

Yang-Sheng, "fhe Halance-of- ~Payments Econometric

\exol Programs: Evidence from Recant menogrephed (December, 1972).

Cheng, } Capi.tal.- ~Cor

° 7 Studies,"

Bee em

Ghoudghry, N., ¥- Yotowitz, JA.

R-

4 os a

Sasyer, J.W.L. Winder, The TRACE

Econometric Model of the Candia Economy, University of

—

Yoronto Press, 1°72.

Cohen, Penjenin 3., “A Survey of

Creation in the 1960's

ote

Regaraing Tied Traterest SEs ° States Relense of Pavoents, W

Conamittee, Office, 1973).

“bth Congress, 1s

Cooper; Richerd ., “The Intere in the Separation of Cz enital Market " Finang Archiv, Yol. 24 (Deccinb ery,

1965), pp. L477 1.

le act it ye be a tte ae iY Lf.

BEd Zy PIsVUaes ae y ao feed OF -3 ad Tiyan:

Flows, "Tt tional Finsnce

Pederel Rese Interns oh

‘the | Pede

Freedman, Charles, NJnternational Copital Flows ,- Interest Rates and ‘the Meney Sap for Economic Rese

ut =

4 ‘ x1y Discussion Pap

Ciccolo, John HN, Jy. and Ndward F. McKelvey, “Large Rank Eiability mincographed (1971). :

ctewity,! jie oded: in jhe 4

Bear snes before the joint

Session (U.S. Government Print ing 4 .

st Equalization Tax: An Expeximent

Arbityvage and Stwrzt-term CaphuaL Discussion Paper No. 49 (January Washington, D.C.

vve Dank of Atlante, Research Nepartmenlt, Cuide to pal Finnnetei Statistics Collected and Processed by to Raserve Bank of Atlanta, Mey 1972.

»»)

rho. 73-32 of -the Center

? 4, Department of Economics, University of

/ are Minnesota (way, 1973). . . ° . . : Girton, Lance “and Dale Vendersen, “A ‘Wwo- Country Model of _Fananeial Capital movements as Stock Adjustments with Emphas

il Move on the Eifeces of Central Ran! Discussion Paper No. 2

fon of International Painance

Federal Reserve Board Gfarch, 1973).

Girton, Lance and Dona E. Roper,

"Floating

’

unpublish

Nechumee Rates Applic

ae .

‘J } drake (1974).

\ "A Monetarist Model of Piset and Jo te the Canadian Experience

aD e-em nnn =P eum exyeceanenenseumrmrersemenre- rerun eeerereeeeeep ee * . soe ves

eme Meee

Griliches, gyi, “Distributed Lags: A survey" Reonometvi ca, Vol. 35 (Janvery, 1967).

. ~ gyubel, UGE Nontexnationayy y pinersi fied Portfolios: Welfare . Gains and Ca ypital seer American Economic Re Review 58, No. 4

(ecenber , 1965), P - 43299-1314. . .

Nansen, Bent. , Foreten Trade ‘redjts and Exchinec Reserves:

; , Pores en pre ee ae A Contribution be tha theery cf Canital Movements (Amsterdam: . A CONN on te ory soe. Cienhee

re North Hollac Publishing Company, 1$61).

Nawhing, Re wort G., “irre stubility of viexible Exchange mates -> Tre Canadian Experience, The Bullctin, gow York University Institute of Finence, Part IL of No. 50-51 (July, 1968}.

. oy

Heeherman, Donal, “Wppternation:

aj. Short-term Capital Novements: Comment 11,' pmericen reonems.c review, Vol. 57 fe ve”

, No. 3 GJunc, 1°67)

aé

. Nelliwell: John ¥., "A Sexuctural. Hoéel of the Foreign Exchange . - Pearkec," Consdion WouerG OF & Hoonowde ss Sol. 2, Roe 4 . nat on PO Hee

. (February, 3969), .

Velliveil, atchn ¥., Hed. Shapiro, GR. Sparks, Te A. Stewart, . F.W. Gorbect aria DR. Seephensen, ahe Structure 6 of BENZ, parts

4 and 2, Bank of Canada Staff Research ch Studies, No. 7, (1971).

Wellinel), jonn F., “The yFfacts of Rev alu ation on Trade and - . Capita plows wotacn Canada and the taited Steres, in ,

: Caneds 207 ive myjecd Shotes Tinea = Rabath onships, Gre x cne8 Les

. Of a conference hoatd in Septoemocr 1971) “Federal Reserve Bank

. of Boston, 1972. (19724) 7

Velliwell, john F., and “Tem Meewell, "Short-term Capital Flows and the Foreian Exchange Herne," | Canadien Journal oF of Economics, Vol. 5 (ilayi 1972) (1972b). : Helliwell, John r., and Tou Maxwell, “Honetary Interdependence of Canada and the » nd tee States Wider Alternative Fschanec Keke . Systons , H Sorehacstns in RZ. ALiber, cd. , National © Honct arn ‘ policies and le TnharnaLyen: al Pinanedal System (A jnospre fond ap eat wee . "* ~—_——— oo* Si ee ev Cenlerence O72). (ivilc)-. ’ ~ .

aah detrei en den anal eee . feta ae td ee me pee

al thd hatte atten? os >, wee tel

mndexrshott, Pa rntric He, “Comaent IV," American Economic Review,

Vol. IIL (Sune, 1967). (1967a).

. . . .

Viendershott, Patric T., Mphe Structure of Intern national Intexest

Rates : Rate,"

Vexyx sno ,

Flows _ ame ate est rate Relate: no Dnivs

Ration

The U.S. Sreasucy BiJl Rate and the Kurodollar Deposit Journal of Finance, Vol. 22 (September, 1967). (1967b).

Richard J., Internat ional T3 yiner cial Integration: Capital.

wong ee Incusee) ial

y, march, 1973,

Nexxring, Richard ane <.D. Willett, Nyhe Capital Control Progran and U.S. Investmenc Activity Abroad," Southern Economic Journ: wd, pov LON

Vol. 39 (July, 1972). . .

. . . . YMewson, John and Eisuke Sekakibara, "! General Eguidibrium Approach to the buro-dcolli warket,', Unpublished dyvaft, May, 1973. (19734). Heveut, goto Ad siduhe Senekibera, “The miiects of U.S. Contess s on U.S. Commer cial Bank Berrosing in the Duro-éollac Market and the Euvo-dollar Interehe Rate, LP Departwental Momorandun, unpublished (Decesser 1975).

- Hewson,

Mul.tip unpub}

Hodjera, 1963-1 _ Galy-

Hodjexa,

John and Eisuke Sakakiverc, "The Euro- ‘della Deposit lier -- A Portfolio Approach," INF Department Memore ndun, ished (Januery, 1974). .

. ° oe

Zoran, “Short-term Cenital Movements of the United Kingdon, . O67", Journal of Political Neoncuy, Vol. 79, No. 4 August, ~ 197).). oy .

goran, “International. Short-teria Capita Movements:

A Survey of Theory and Empirical analysis," J HP gkaff Papers,

Vol. 2

Johnson,

Theory,’ (March, 1972).

0, No. 3 (Hovenber , 1975 ).

Narr ry G., “The Honetory Approach to Balance of Payments Journal of Pinancial and @ entitative Analysis, vol. Vil

cement Controls,"

tional Capital Mev 973 under

. Johnson, Narry G., “Interne

vo (Report on a conference held in Geneva in June 1 the sponsor ship of the International Center Lor Monctary and Pankiyg Studies), Eunomoneys August, 1973.

“Short-term Capital Movements and the U.S. n united states Balance of

e . ' Kenen, Peter }., i ments," sneludes i Joint Economic Committee,

Balance of Pay

Payments, Voarings before the 68th Congress, lst Session. er Office, 1963). , 8

. ° a : : .

(U.S. Government Printing

Loe eer ee

Worning Wayport ai Tncernacionad

; Rencen, Peter Be, Porticlse Canical ond Mone

. “ yvoluacs, MIMAOSLEPHAG? >»

Fcononics, International Finance

Economics, Princeton University (September , 1673). 4 .

S2ecion, opar tment of

: > Kesselman, J. Nyhe Role of Speculation Sn Forward-Rate Yeterninacion: The Canadian Flexible Dollar 1953-60". The Canadian Journal of Econemics: Vol, 4, No. 3 (August 1971).

Pentel TeRes Onpitel Flevs: envserip’, June, 1975.

J. cr: J unpublishe j

~ - Yours, Pentti 3.¥%., and Porter, .G.; “WIncexnational Capital Flows and Portfolio Equilibyiue,". Journal ef Political. Looney

(forthcenins) (19724). . ss

Yours, Pentts 3.K., and Porter, H.G., “Interest Rates, Capital | _s Flows and Ruchanpe Rate Expectazions," nr Steals Mamerancun py /72/S% (1972b). .

— Ywack, Sung y., NA Disaggregated HWodel of Financiel Flows Betvecn the United States and Foreign Regions," Unpublished mimeograph,

: . " . /1971a. ta Disaggregated vodel of Capit apublished MLMEOGY é

/. Kwack, sung Ye, al Flows Tetween _ Y,S. banks and Foreign Regions,” U inh, 1O71b:

eo. ; \ ; . . ‘, . - . vervevemepmen ss . Ze .

Kwack, Sung Y.

An Extension of Vey UT SHOtL “Vol. 26, ho. & ; {5 Sepiosber 3 4S

Yuack, Sung Y

Mode}. of th . _ and Yests oO

Confercnce

February, 1

R-8

“uphe Struct ure of.

“t vo ca

107

on Econ monric 1972 (fort rhcoming) «

International Interest

ects!, ‘The Journal of 3

Wt). (19 g71c).

and Ceorge R- Schink, e United States Trade 2

£ fTolicy E&fFectivencss. woce) Building end Hevelopmei

end Capital Flows: § " presented at? Broo

Rates: jnNance,

"A piengaregated Quartexly

jrvlations kings ne,

ts = of

yeack, Suny y., WA Rote on the pabeance cf La ments EFEact

the U.5. Capital Controls Prope aris: Simulation Estimate s,'

. mine og graphed (forthconins sn Journal of Pimance)s 1g73,. Rate

Kwack, SYNE u.;

Agreement

sO jmalysiss"

Laffer, Aen

Leamer , y. Ee (Boston: #4

Y D

Leaner, VE. Est imat2o

et. al. (eds

. “yatienal . . Universit

Lee, Gls

the United (Ju a! Jisgust, 1969).

vol. 77

Lindert, Journal

HMachlup, Prits ceo VWobd |

“ic esearch ¢

, Jnternat

——eoor

Jurcaau af Pearse

Pr "Sh,

pou, "The Payments

“the Impact e+ on the U.S. "Ha Unpublished,

uct Dey (fone, 195?).

and RM. StCTH,

and Tle Stern,

- s ek a -, . ayn the Smithsonian exchange

“Comment TIL," Quantitative ae

Wwpyablems in the Theory 1 Movements, "oon F.

nownie Rev

—_

akernetio sak

sllyn ana Becon, 1970).

An Econometric

Lev,

Anord ceo EDC Fcone.t a

Economics

and Empirical

Machlup

n of Internation? a}. Capita od, Tpke atic saa Hone gust ity ent A ovement of C20: ihe al, Columbia,

Bureau c ty Press, 1272).

WAS stock-Ad justnien’ St ates- Canad

of pinence, Vol.

3oN

1972).

’

for eonein ¢

an Case. Ww Journal. ¢

NTHENOG action," jn ¥. ane ovoment of

roseazch Gxcu “York:

Fad

t Analysis of

Impact of yoresgn Invest

26 (Decempen 1971).

New York: “Culumia

Capital Move ements: of Political Boo

Sa) ONnoe aN

rent Controls.

Machlup ct. al. (cds.)>

ene G: anital, Yational University

ned

. ° .

. - \

Makin, J.-M. Mycinand ane Supply yunetions for stocks of Bure~

ek -

do) lax pepositss An Empay seal Study," Who Review | of Economics

and Statistics, Vol. VL CHovEnder 5 1972).

8 Markows.tz, “Waery Wey Port forse selection (ew york: Wiley and

~—_————a on

Sons, 1959 ). os

Marston, Richerd c., “Ancrican Honat ay policy and the Eurodol.lar

“ Market," Brookings Tnecicucion, Unpublished manuserapl, 3971.

(fox theomins as a Princeton Study in International Finance.)

Jarston, Ricalra C. and Surg ¥. yoack, “A U.S. Capit cal Account Sector for che Wherees" Model: Som I yolimiMary Pesulhs."

, Unpublished menuserip’, Apri 1973.

. ° se . ry .

of the Monctary aad Balence of Payments Sectors

Nartiensen, Joern and Cuenter $2 wdermann, "A quarterly Nodel

s of the West

; German Ec oncay,' Diseu 19$i1.0n Paper no. 47, Institul fur . _ Goselischafts-une Wixrtschaftssis eenschaften, vniversity of

. , ponn (August, 1973). . 7

; Massaro, Vincent, ‘AD Econemetyic Seusy of Furo-doller Bor rowirs,

Comment , " Jour) al of Yinences Vol. NTT (S sepeemper , 1972).

- & oa . .

Mastrapasqua, Frank, "ve s. Bank rxpansicn via Forel

en Branching:

. Monetary policy Inplica teyons,' Fhe ul Bulletin, New York University ‘ . Graduate schoal of Lusiness Adwinistzatien, Inst: Leute of Jinence (lanvery, 1973). . .

Michael, Walther Po, Hessuzins_ Internagional, Capitel Moyvenents tes eee —_—— St =}

poneeenens

. Occasional Paper Ho. 114, National Bureau of Economie Researen

. . ae (ew York: Columbio University Press, 1971).

ws Analysis of Uni.tca States Long-term Pe wrtfolio Yoreige Inves ; . ment." Quarter ty Journal of Reon mice Vol. LXXXIV (Hay, 1 1970). " (1970a). .

_ Willer, Borman c. and Nerina V- Nn. Whi.cmaMy WAL Cernative whoories

and Tests of U.S. Shart- earn Foroeigu Jrvestwen manuscript: (Novemler, ya7O). CIO7OD).

Willer, Noxmen Cc, and tarina, VOR. Rhituaasn, WA Moan-Variaace

t," Unpubl is shed

Hiller

Shorts term Puncs from the United and FPlo¥s, >, Machlup ct. al. (cds. ), Interna ational Mobility ment of Capital, Rational Pureau for Economic je Res rarch

and } Hove: “(ew Yor

"Wills, Roa: Hovementinvns

Pedexal

--*

Nills, Redne an We, Jue, Movements in }injor Stafzé Neonoric Study ho. 74 (Octeser,

Joaxd

Mills, Ro Nole atid of ‘Of Inte

Their Control, June

soot Pe ReLO

gt of. Jn ‘

C, and Marina V.N. Whiten, Mrhe Outflow of States: Adjustment of Stocks

“lr: “Columbia University Press 1972).

astern Eurepean Technicucs 3a

wove, dv, tTha Repulation of Short-texm Capital n the 1960's,

Preserve Joard Staff Economic St cudy No. 46 Clay, 1968).

eo ‘ .

<.° a .

car

Mon

dney Me, Sr., An Evaluation of of Measurt

e@ Capital Flows, University of Geneva,

rynational Studies,

Hills, Rodney H., Jr.; Ug eyuceeral Change i Harket: Evidence froma Two: Equation -Nodel Finenee Discussion Paper So, 33, Division Tedcceval nesecve Dead CROWEMLEL, 1973). re

Hundell,

Robert A., Internetional Econon enics (New York:

- Macmillan, 2196S).

-O£Ficer,

Wuctwa Yucharge Rete, Cambridge, 3

ee ee

Press,

., Officer, A Crit Credit

. Porter, Analys

Laerence We, Aw Econone (1 3.¢

L9G

Pho Repulation of Short-term Capital ou Jncustrial Countries, Federal Reserve

072).

. .

sures te Influence

Graduate re Institute

Conference on Capital Movements and 14-15, 1973 (fortheoning) « (J.973a)

n the Euro-dolhar

Internat tonal |

C Model _o of £ Cen2rda tl Under the

Yarvard University

LW. and Willett, T.p., "Lhe Interest Arbitrage Schedule: " Journal of Moncey,

end Panking, Vol. 2, Ko. 2 (say, 1970).

ical Survey of Recent, Developments,

. oe wee

Michael G.,

~ty

"A Pheoretical and Enpixice al Framework fox

ing the Term Structure of Exchange Ra ut Steff Papers, Vol, 18, No. 3 (November,

te Expectations,"

1971).

_ porter, Michael G., "Capital Flows as an Offset to Nonetary policy: fhe Gexman Case," Suit St Shalt Papers, Vol. 19, Ko. 2 QJuly, 1972).

Te, Powric, T.L., "Short-term Capital Movements and the Flexible , -Canadi.an Exchange Rate, 1953- 61," The Canadian Journal 0: of .

Economics and Politica. sejenec, Vol. 30 (February, 1964).

. Prachowny, | Martin Ied-, ASEY svetural Model ef the U.S. Bi: Balence

cae ne cee oe a

of Paynents (Amsterdam: Rox the Holland Publishing “Co. , ~ 1569).

vs,

Review Cemaittece fox Natence of Payronts Statisties CParnstcin Comi tree Pey;ort to the Burcey of the Budget, nie Ralarce etics one be Uatced States, A Review and

en ee ee

. te . .

Rhomberg, N.R., “Cane dats Toreign Exchenge Market: A Quarterly Model," DE Stake Pavers, Vol. 7 (April, 1960). . . , the Cenadian Beonony under Fined and

. Rhembers, RR . h Yodel oi

RRL, A Yluctuating Exchange Retes," Journal of Politics al Eeonony, Vol. LXATL (velsuczy, 1964),

Rhomberg, RR. “cornment ,* in F. Machlup et. al s ? > > Internatioas#l VE ytLity and vovemont: of Capita al se tional

= OU Ah

> 7 Bureau of heone kesearen (New Your: Columbia University

Yress, 1972).

Salop, Jeanne, WA Note on the Monetary Appy oae of Payments," Taternetional Finance S ; Division of Jnternational Finance, Tederal

. .. (§ovender, 1973). .

KOR

Sandermann, Gunter, "A Pala ace mimeogrephes , Tnstitec fur University of Bonn (Augus a

Payments Model for West Germany ;

chafus-und Wirtschaitswissensche?

CG eke

AIM fa

Lis 972). . . Smith, John s., MAsymmetr tes and Exrors in Repar tee ee of-

. Payments Statx reties," JSP Stafe Papers, Wolk. 1 1 (July, 1967).

Spitaller, yateh, MA Survey cf Recont Quantitative Stadics of oe Jong- tern Capital Movements! DME Stark Papers, Vol, 18, No. + . . 7 QJaxch, O78 '

aati eit ec en 5 ta aii iaamea al aeidinns een aaa a

— « . { . 32 . i c . ‘Stein, Jerome. Ley “ynternational Shor t- tera Capital, Movements ," American n Feonove Review, Vol. 55, No. }. (iarch, 3.965).

4 Stein, Jerone L., Ushert-term Capital srovements: Reply," American reonemic Review, Vol. 57, No. 3 QWvune, 1967).

— ee

Stevens, Guy V.G., “Canital Mobility and the International Firm,'

sn F. Hechlup ct al, (eds. Internationa) Mobility and Movement | ’ ntexnabiona) PC A Sa

‘of Capital, ational Bureau of. Economic hes fearch (New Yorh: Columbia University Press, 1972). . .

Stewart, Tan A., “RDM he search Department tyoeximental’ Medel Version 2," Bank of Cenada Revice, (April, 1972). |

Stoll, eRe, “AN Erpie cal study of the Forward pxchanbe Market

>, 3

under Fixed and Yie ble Exchange Rate Systems. Ca dian

Journal of ceoncries, Vol. 1, No. 1 @ebruary, 1965).

fobin, Tames, hroney Conital, and Other S& tores of Value,’ : > t Y Papers and Prescesi ans a’ the American Roo comic Assockarion,

Vol.. LI Chay: 1261) *

fobin, James, “Ihe Theory of Portfolio Selection," in F. Kehn and F.P.R. Brechiing, eds., The Theery ef Intere st Rates

(Rew York: Macmillen, 1965).

Tsiang, S.C., Mrhe Theory of forward Es sche nae and Effects of - Government Tntesvention on the Yorvard Exchange Maret," yer staff Panervs, Vol. 7 (April, 1959).

. . . . . . . oy °

Valentini, Johv J. and Lacy li. Hunt TL, “Evro-dollar Borrowing by Kew York fangs and the Rate of Jntercst on ane dollars : Comment," Journal of Finance, Vol. XXVIE (March, 1 72). °

White, Wu, "Interest Rate pifferences, The Forward Exchanges; Tre and the Scope for Short-term Capital wovements,'! IME Sstaft Papers, Vol. 10, Ne. 3, CSovember , 1963).

Cite this document

APA

Federal Reserve (1974, June 30). Empirical Research on Financial Capital Flows. Ifdp, Federal Reserve. https://whenthefedspeaks.com/doc/ifdp_1974-50

BibTeX

@misc{wtfs_ifdp_1974_50,
  author = {Federal Reserve},
  title = {Empirical Research on Financial Capital Flows},
  year = {1974},
  month = {Jun},
  howpublished = {Ifdp, Federal Reserve},
  url = {https://whenthefedspeaks.com/doc/ifdp_1974-50},
  note = {Retrieved via When the Fed Speaks corpus}
}