The Implications of the Internationalization of Banking for the Definition and Measurement of U. S. Credit and Monetary Aggregates

International Finance Discussion Papers Number 177

February 1981

THE IMPLICATIONS OF THE INTERNATIONALIZATION OF BANKING FOR THE DEFINITION AND MEASUREMENT OF U.S. CREDIT AND MONETARY AGGREGATES

by

Michael P. Dooley

The Implications of the Internationalization of Banking for the Definition and Measurement of

U.S. Credit and Monetary Aggregates

by Michael P. Dooley*

1. Introduction

Macroeconomic analysts often assume that a "closed economy" model of a country or region captures almost all important economic relationships. They presume that a "foreign" sector can be added to the model but that, at least for a large country such as the United States, qualitative results are unchanged and quantitative results are only slightly affected. The use of national monies and other political, geographical, and technological barriers to trade in goods and financial assets provide some basis for this fundamental simplification.

Many important :contributions in international finance can be viewed as attempts to determine degrees of independence provided by national currencies. These contributions include comparisons of fixed and flexible exchange rates, proposals for identification of optimum currency areas, suggestions for the appropriate assignment of policy instruments, and statements of the policy implications of the monetary approach to the balance of payments, or currency substitution.

This paper explores a feature of international economic relations that may become as fundamental as the use of different national currencies. That feature is the increasing lack of correspondence between the location of real economic activity and financial transactions related to that real activity./ This phenomenon is important regardless of the currency

denomination of financial transaction relevant to economic activity in

* The views expressed herein are solely those of the author and do not necessarily represent the views of the Federal Reserve System. Portions of this paper were adapted from a paper with Jeffrey Shafer. I am indebted to Lance Girton, Allen Frankel, Dale Henderson, Samuel Pizer, and John Watkins for comments and criticism. oo.

1/ For a comprehensive discussion of this problem see Bryant (2, 1980).

a country or region. Difficulties introduced by this lack of correspondence in defining, measuring and controlling financial variables such as monetary or credit aggregates could prove to be thecrucial problem for economic policy in the 1980s.2/

This paper focuses on dollar-denominated transactions of U.S. residents in banking markets and markets for government securities. This emphasis is motivated by the fact that a large part of U.S. international financial transactions take place in these two markets. Moreover, domestic open-market operations and foreign exchange market intervention involve these financial markets. Finally, special attention is paid to U.S. banks' assets and liability structure in evaluating financial conditions in the United States.

The conclusions reached are that the international transactions of U.S. residents are likely to continue to expand rapidly. Such traasactions are related to monetary policy and can affect the interpretation of both domestic monetary and credit aggregates. For this reason, it is important that a somewhat detailed analysis of international transactions be a part of any interpretation of U.S. financial markets or financial policy.

In the next two sections data from recent years is presented which shows the rapid growth in both net private capital flows and roundtrip flows. The remaining sections of the paper present an analysis of the

17 This phenomenon is in some ways similar to the problem addressed

in the currency substitution literature. That literature explores the implications of substitution among non-interest-bearing monies issued by different governments. This paper argues that U.S. residents can hold dollar-denominated assets that do not reflect the cost of non-interestbearing high-powered money. Thus, the burden of the inflation tax and incentive for currency substitution is considerably reduced. See Girton and Roper (6,1980) for a discussion of currency substitution.

relationship between these two types of capital flows and U.S. monetary policies. It is shown that international financial transactions induced by a change in U.S. monetary policy can distort the behavior of U.S. residents holdings of monetary and credit aggregates at domestic financial institutions.

2. International activities of commercial banks

The consolidated balance sheets of large money-market banks are perhaps the best indicators of the degree to which banking markets have become international. More than half of the assets and earnings of some large U.S. banks are related to foreign activities. In the aggregate, Federal Reserve member banks had assets of $312 billion at the end of 1979 in about 800 foreign branches. In addition, foreign banks have become important in the U.S. domestic market; their standard banking assets now total $130 billion constituting almost ten per cent of all such assets in the United States.

The increased volume of international transactions of U.S. banking offices is related to two fundamental developments .=/ The first is the role that private capital movements have played in the managed floating exchange rate regime as it has developed since 1973. The floating rate period has been marked by large current account imbalances that do not seem to adjust rapidly to exchange rate changes and by a very high degree of private capital mobility. Moreover, massive exchange market intervention by governments has allowed desired changes in private portfolios to result in both net private international transactions and

a 1/ For further discussion of the development of international banking see Aliber (1,1976), Freedman (4, 1977), Grubel (7, 1977), Giddy (5,1981).

exchange rate changes. In this environment, total net private capital flows reflect the reactions of the private sector to relative changes in monetary policy and other shocks to the system. The international transactions of banks, in turn, reflect the competitive position of banks in financial intermediation between U.S. residents and nonresidents.

The second fundamental development has been the rapid growth of round-trip capital flows. These are transactions essentially among U.S. residents that are motivated by lower costs in transactions involving a nonresident at some point in the chain of financial intermediation. The primary cost considered in this paper is maintaining non-interest-bearing reserves against domestic bank liabilities. These transactions are neither directly related to payments imbalance nor to changes in the demand for dollar assets, but instead are influenced by the level of nominal dollar interest rates, a measure of the cost of non-interestbearing bank reserves.

3. Banks' international transactions

and net private capital flows

Table 1 shows annual data for four major components of the U.S. balance of payments: the current account, foreign official capital flows, bank-reported capital flows, and all other private capital transactions. As shown in the Table, the international positions of U.S. banks are a large and volatile component of U.S. international transactions,

For example, in 1977 and 1978 the U.S. current account deficit was more than financed by official capital inflows related to official intervention purchases of dollars. The counterpart of this was large private capital outflows. Bank-reported flows made up the bulk of this outflow

as banks in the United States increased their net claims on the rest of

the world by over $20 billion over the two-year period. In 1979 this pattern was reversed as a generally stronger dollar led to official capital outflows and private inflows. U.S. banking offices again played a dominant role in total private capital flows as they reduced their net

claims on nonresidents by $15 billion during the year.

Table 1. Selected Components of the U.S. Balance of Payments

(Millions of $, + = surplus or increase in net liabilities)

Changes in Foreign

Current Official Assets in Private-Capital Transactions Year Account the U.S. Banks All other 1973 7,140 5,090 -1,278 -9,401 1974 2,124 10,244 -3,499 -8 , 068 1975 18,280 5,509 -12,904 -8 ,080 197€ 4,384 13,066 -10,378 -4,927 1977 ~14,068 35,416 -4,708 -13,730 197 -14,259 31,072 -17,372 -2,251 197 -788 -13,556 15,529 3,278

i a

The apparent relationship between banks' international transactions and conditions in foreign exchange markets is not attributable to changes in U.S. banks’ foreign currency positions. At times during 1977 and 1978 the public was unwilling to hold dollar assets at existing interest differentials and exchange rates; consequently, foreign central banks became the residual buyers of dollar assets resulting in private capital flows from the United States. Foreign official institutions purchased Treasury securities, which are held almost exclusively by U.S.

residents, generating a record official capital inflow.

Official inflows in the form of Treasury security purchases were larger than the current account deficit; necessarily the difference was matched by private capital outflows. The "pull" from U.S. banking markets during most of this period clearly was generated by the "public's" desire to sell dollar-denominated assets to foreign central banks in return for foreign currency-denominated assets. However, this did not imply that U.S. banks acquired foreign-currency-—denominated assets! In fact, it appears that U.S. banks and other U.S. residents were induced to make dollar-denominated loans to nonresidents, who in turn sold dollars to foreign central banks. U.S. banks acquired funds by selling deposit and nondeposit liabilities to U.S. residents. U.S. banks in turn advanced funds to their offshore branches. The funds then were loaned to another bank or nonbank that wished to acquire a dollar-denominated liability and to sell the dollars borrowed for a foreign currency asset (e.g., a markdenominated bank deposit). The motive for all these transactions was a nonresident 's willingness to bid a bit more aggressively for a dollar loan, based on the judgment that the dollar's exchange value would fall. In 1979 the opposite flows were induced by a net private demand for dollar-denominated assets and official sales of dollars on foreign exchange markets.

4. Banks' international transactions and round-trip capital flows

Another factor behind the increased international activity of U.S. banks has been so-called "round-trip" capital flows. These transactions essentially involve U.S. residents and are motivated by lower inter-

mediation costs associated with offshore transactions.

1/ See Lowrey and Smith (10, 1980).

Table 2 provides some data that are suggestive of the importance of round-trip capital flows. Columns 1 and 2 show domestic and Eurodollar CD interest rates expressed in terms of cost of funds to a U.S. office of a member bank of the Federal Reserve System. That is, the market yields are adjusted for reserve requirements. Reserve requirements against Eurodollar borrowings were lowered to zero in August 1978 while required reseives against domestic CD's were raised from 2.5 per cent to 4.5 per cent in October 1979. As shown in column 3, the reserve requirement adjusted differences in costs to banks were quite small as banks closely arbittraged these two sources of funds. Column 4 shows banks' utilization of Eurodollar borrowings. U.S. bank borrowings rose rapidly in 1978 and 1979, a rise that accounted for about half of all managed liabilities raised over the two-year period.

The first two columns of Table 3 show the yield available to nonbank investors on Eurodollar and domestic CD's. The differential in favor of Eurodollar deposits, column 3, remained fairly large throughout the period. The rapid buildup in holdings of Eurodollar deposits by U.S. residents, showa in column 4, might indicate an increasing sensitivity to such arbitrage opportunities; the remaining incentive suggests that there is scope for continued growth in this type of capital outflow.

5. International banking and U.S. policy

The very great involvement of U.S. banks and nonbanks in offshore markets has led to questions concerning the adequacy of monetary policy tools. The volume of capital flows, outlined above, has generated considerable unease concerning the effectiveness of monetary restraint. This

section addresses some of these issues. The problem is that the

Table 2. Relative Cost of U.S. and Eurodollar CD's and Outstanding Amounts of Eurodollar Borrowing of U.S. Banks

Domestic Eurodollar Differential Net Liabilities to

CD's CD's (1) - (2) Own Foreign Offices — —— a San? SD Sol Qwn Foreign Urrices (1)= (2)= (3) (4) (In per cent) (In billions of US$) 1978 January 7.68 7.89 -0.21 -2.3 February 7.68 7.90 -0.22 -2.4 March 7.60 8.45 -0.84 -3.3 April 7.82 8.02 -0.20 -0.5 May 8.32 8.46 ~0.15 0.1 June 8.75 9.10 -0.35 1.3 July 9.10 9.36 ~0.26 2.9 August 8.97 9.14 -0.17 2.2 September 9.53 9.72 -0.19 4.3 October 10.49 10.86 -0.37 6.9 November 12.18 12.36 -0.18 8.1 December 12.32 12.58 -0.26 6.3 1979

January 12.11 12.28 -0.17 6.9 February 11.58 11.53 0.05 12.0 March 11.41 11.36 0.06 17.3 April 11.27 11.22 0.04 19.8 May 11.38 11.37 0.01 23.4 June 10.87 10.95 -0.09 27.0 July 11.14 11.28 -0.14 28.5 August 11.85 12.03 -0.18 33.1 September 13.14 13.26 -0.12 35.5 October 16.19 16.14 0.05 34.4 November 16.77 16.87 -0.09 34.6 December 16.09 16.32 -0.22 28.1

1/ Adjusted for reserve requirements applied to Federal Reserve Member Banks.

Table 3. Relative Yields to Investors of U.S, and Eurodollar CD's and Holdings of Eurodollar Deposits of U.S. Nonbank Residents

Overnight

Domestic Eurodollar Differential Eurodollar Eurodollar CD Deposits (2) - (1) Deposits Deposits (1) (2) (3) (4) (5) (In_per cent) (In billions of US$) 1978 January 7.49 7.89 0.41 14.1 1.1 February 7.49 7.90 0.41 15.3 1.1 March 7.41 8.45 1.03 16.1 1.2 April 7.62 8.02 0.39 16.7 1.3 May 8.11 8.46 0.35 17.8 1.4 June 8.53 9.10 0.57 19.3 1.5 July 8.87 9.36 0.49 20.5 1.6 August 8.74 9.14 0.40 21.6 1.7 September 9.29 9.72 0.42 21.0 1.7 October 10.23 10.86 0.63 21.2 1.8 November 11.63 12.36 0.73 23.0 1.9 December 11.77 12.58 0.81 22.8 2.0 1979

January 11.56 12.28 0.71 24.5 2.3 Februa‘ry 11.06 11.53 0.47 27.1 2.6 March 10.90 11.36 0.46 28.4 2.8 April 10.76 11.22 0.46 29.1 2.8 May 10.87 11.37 0.50 29.6 2.8 June 10.38 10.95 0.57 29.9 2.9 July 10.64 11.28 0.64 31.4 3.0 August 11.32 12.03 0.71 33.9 3.3 September 12.55 13.26 0.71 33.4 3.6 October 14.52 15.23 0.71 33.2 3.5 November 14.68 15.52 0.84 34.0 3.2 December 14.08 15.01 0.93 31.9 3.6

rr

-10-

relationship between geography, currencies, and financial intermediation is becoming more and more tenuous. In this environment it is important to distinguish between a change in the demand for dollar-denominated assets and a change in the demand for financial intermediation located in the United States. Although both these phenomena can create capital flows, they have very different policy implications. Even more perplexing is the possibility that a monetary policy action can affect both the demand for dollars and the relative cost of onshore and offshore intermediation. Because they are at the centers of both policy actions and international capital movements, the behavior of commercial banks provides the natural focus for the analysis. It should be remembered, however, that changes in the competitive position of banks will have profound impacts on the role of banks in international markets.

A monetary shock has two short-run impacts on capital flows. First, a restrictive monetary policy generates both gross private capital inflows and outflows. The motivation for such round-trip capital movements is straightforward. It is often possible for U.S. residents to avoid reserve requirements and other costs by introducing a foreign name into the chain of financial intermediation. The benefits of this procedure are narrower intermediation spreads and, therefore, either lower loan rates or higher deposit rates or some combination of the two depending on the usual elasticity conditions for sharing a tax burden. In general, for a given reserve-requirement differential, a rise in interest rates resulting from a restrictive monetary policy will increase the incentive to avoid reserve requirements, i.,e., the opportunity cost of non-interest-bearing reserves

is higher.

-1l1-

If offshore deposits are reserve-free, the 4.5 per cent reserve requirement on domestic deposits (as existed in 1979) generates about a 23 basis point advantage for offshore deposits when the deposit rate is 5 per cent, a 45 point advantage when the deposits rate is 10 per cent, and a 90 point advantage when interest rates reach 20 per cent. The increasing yield to investors of six months CD's, shown in Table 3 during 1979, roughly reflects this relationship. Thus, we should expect to see increased use of offshore markets by U.S. residents during periods of monetary restraint or rapidly rising inflationary expectations.

The second effect of a restrictive monetary policy on international financial transactions is to generate an incipient private capital inflow (an excess demand for dollar-denominated assets) and in turn upward pressure on dollar exchange rates. If this pressure is met by official sales of dollars (an official capital outflow), then some form of private capital inflow will be necessary to clear credit markets =! It appears that until quite recently commercial banks were the pre-eminent channel

for such private capital flows.

6. The Balance of Payments Constraint

Evaluating the effect of international financial transactions on the

U.S. economy requires placing a given transaction, such as a loan from abroad obtained by a U.S. resident, in the context of its counterpart transaction in the balance of payments. A foreign loan may have as a counterpart a U.S. current account deficit. For example, the current account deficit requires U.S. residents (including the U.S. government) ~1/ The market clearing conditions for stocks of assets, interest rates, and exchange rates are the focus of a large number of open economy models.

For reviews see Henderson (8, 1977), Isard ( 9,1979), Dornbush (3, 1980). This paper focuses on the definition and measurement of endogenous variables.

- 12 -

to raise an equivalent amount of funds, net, from nonresidents to finance the net imports of goods and services. Such credit may be obtained by selling private claims, net, Treasury securities, or both to nonresidents.

Net U.S. indebtedness to nonresidents will not change as a result of Federal Reserve policy unless the net U.S. expenditures on foreig: goods and services (the current account deficit) also changes. One would not expect a program of monetary and credit restraint to have a significant impact on these expenditures in the very short run. Hence, U.S. residents in the aggregate cannot evade a restrictive credit policy by borrowing more, net, from nonresidents. Eventually, as restraint curtails domestic economic activity, thereby tending to reduce the current account deficit, net U.S. borrowing from abroad would be reduced correspondingly.

7. Implications of "round-trip" capital flows for measures of domestic monetary

and credit aggregates

While the balance of payments identity insures that changes in net indebtedness must be equal to the current account, there remain difficult problems for measuring money and credit aggregates in a world witha integrated banking markets. Some of these problems can be illustrated with the following simplified set of balance sheets. In the analysis to follow the assumption of a balanced current account is maintained so that there

is no transfer of net worth between the nonbank sectors:

-13-

Country 1 (United States) Country 2 (Rest of the World ) Nonbanks 1 Banks 1 Nonbanks 2 Banks 2 put pat ) pal? pu ywot

where D = bank deposits

L = loans G = government securities H = high powered money NW = net worth (assumed constant)

Superscripts indicate holder, residence of holder and residence of issuer, respectively. For example prl2 = bank deposits held by U.S. nonbank residents at Rest of World Banks.

For simplicity we assume:

(a) no excess reserves so that: r(p@t + pl) = yl r(p"2? n prl2) = 2

where r is a reserve requirement. (b) only banks hold government securities (c) yr is the same for all bank deposits

(d} there are no interbank deposits

-14-

In this simplified world total bank credit available to U.S. residents consists of loans from domestic banks, ppl and foreign banks, pn? | U.S. residents also hold domestic deposits, pall | and foreign deposits, pal | It is clear from the Nonbank 1 balance sheet that the supply of bank credit available to U.S. nonbanks cannot change unless their deposits at U.S. and foreign banks change. While this seems obvious, it has the important implication that a measure of monetary aggregate that includes offshore and onshore deposits is also a comprehensive measure of bank credit available to U.S. residents.

In order to evaluate the effects of an open market operation initiated by the U.S. monetary authority we can focus on the monetary or credit aggregates that appear on the balance sheet of U.S. nonbanks.

A restrictive domestic open market operation implies:

(2) - AnDL =ac>lt

e and given assumption (a):

(3) - L pet = —apPlt _ qph22 2 —apBll _ yyn2l

which states that the sum of U.S. banks" deposit liabilities to U.S. and foreign nonbanks must fall and the sum of U.S. banks' loans to U.S. and foreign nonbanks must fall by an amount equal to the open market purchase — divided by the reserve ratio. In order to know the ultimate impact on U.S. nonbank residents we would have to add behavioral relationships to the system of identities and derive equilibrium conditions, a task beyond this paper. But by simply manipulating the identities it can be shown that:

1 bl nll 12

(4) - = aH = -aL™™ - AL? mnl2 _ apntt

+ (AD AD)

-15 -

which states that the effect on bank credit available to U.S. residents,

apnlt + arnt? | is equal to the full reduction in credit issued by U.S.

banks, - 2 an’, unless U.S. residents increase their holdings of deposits

in foreign banks, ap? | or foreign nonbanks decrease their domestic holdings

in U.S. banks, ~ap@? |

An important determinant of the scope for shifts in deposits by nonbanks in the United States and abroad is the nature of the offshore banking system. If it is subject to a binding reserve requirement:

(5) p-- + pt = + y>2

and since 2 y?? is fixed, it follows that in order to avoid a reduction of bank credit implied by a restrictive open market operation, U.S. residents would have to induce foreign nonbanks to reduce their holdings of bank deposits and bank credit. The scope for this is probably quite limited.

If the offshore banking system is not subject to a binding reserve requirment, or if U.S. nonbanks deposits in foreign banks are not subject to reserve as is currently the case in most instances, U.S. residents can avoid the reduction in credit available at U.S. banks implied by a restrictive open market operation by selling loans to foreign banks if they are willing to hold additional deposits at foreign banks. This is a more precise description of the round-trip capital flows discussed

1/

above.— If this sort of transaction is quantitatively important, as we

if Ultimately, the control of such a system depends on a stable relatioaship between the world monetary base and the total size of the world's banxing system. For a demonstration that this is possible in a closed economy model without required reserves see Johnson (10, 1976).

-16-

showed earlier it has been in recent years, it is likely that measures of domestic credit, pil and the domestic monetary aggregate, pull | are insufficient as a quantitative measure of U.S. nonbanks financial position.

An implication of the above analysis is that the propensity of U.S. residents to obtain credit from offshore sources depends in part on the attractiveness of offshore deposits. As insitutional changes occur making offshore deposits good substitutes for a wider range of domestic deposits, the potential measurement problems increase correspondingly. Throughout the 1970s, the growth in offshore deposits and other domestic alternatives to bank liabilities was concentrated nontransactions-type assets. As shown in Table 3, holdings of term Eurodollar deposits by U.S. nonbank residents more than doubled in the two-year period shown. While the level of such deposits remains small relative to similar domestic aggregates, the rapid growth suggests an increased sensitivity of U.S. residents to the higher yields available on reserve free deposits.

A potentially more serious measurement problem would be the continued expansion of transactions-type balances at offshore banking offices, Data now available suggest that demand deposits of U.S. banking offices remained the dominant transactions asset for U.S. nonbanks in the 1970s. But there are no theoretical or technical reasons for this to remain the case in the 1980s. Indeed, the change to same day settlement for Eurodollar transactions expected in 1981 will be an important step in this direction. As shown in Table 3, U.S, nonbank holdings of overnight Eurodollars at the Caribbean offices of Federal Reserve member

banks has grown rapidly although the level of such deposits was only

-17 «

about $3.5 billion by the end of 1979. The overnight Eurodollars shown here are included in the Federal Reserve measure of m2 2/ Such deposits are available the next business day without penalty and earn interest near the federal funds rate. As reserve requirements on domestic trans~ actions accounts outside of Federal Reserve member banks are raised over the next eight years (as the Monetary Control Act of 1980 is implemented) and. as international communications and corporate cash management techniques improve further, there will be incentives for U.S. residents to make increasing use of transactions balances in offshore banking offices, Round-trip transactions can generate particularly difficult

measurement problems during periods when the pattern of bank participation in the process is altered. An example of such a period was October 6, 1979 through March 15, 1980, when the Federal Reserve imposed a marginal reserve requirement on managed liabilities, including Eurodollar borrowings, of U.S. banks. If binding, the marginal reserve requirement would have increased the cost of funds to U.S. banking offices and, in turn, resulted in less favorable lending terms. Rather than absorbing this cost, U.S. borrowers were apparently able to deal directly with foreign sources of credit .2/ It seems likely that this has contributed to very large unrecorded capital inflows estimated to have been about $50 billion in the four quarters ending in June 1980. Under these circumstances it seems likely that domestic bank-credit aggregates were a smaller than

2/ While such borrowing made it possible for U.S. residents to avoid the costs of the marginal reserve requirements, it did not allow the U.S. nonfinancial sector to avoid the effects of the sharp increases in

market interest rates which followed the October 6 and subsequent measures.

- 18 -

usual share of total credit available to U.S. residents over this period, and that the normal relationship betwen bank-reported and total private capital flows was significantly distorted.

8. Implications of official intervention

in foreign exchange markets for

measures of domestic credit

A program of monetary and credit restraint generates higher (relative) interest rates on dollar-denominated assets, an increased demand for such assets and for net borrowing in foreign currencies by U.S. residents, and upward pressure on the exchange value of the dollar. If such pressure is met with official sales of dollars in foreign exchange markets, the private purchasers of the dollars will invest them, by definition, in dollardemoninated assets.

The dollars might be invested in a loan to a U.S. nonfinancial corporation apparently adding to credit extended to U.S. private residents. But it would be incorrect to conclude that credit to U.S. private residents had expanded since one must also consider the other side of the foreign exchange market intervention. The foreign central banks first must have sold some dollar asset, in the most common case a U.S. Treasury security, to obtain the dollars sold in the exchange market intervention. Hence, the increase in dollar funds available for lending to the private sector is matched by an increase in the supply of Treasury securities that must be held by the private sector.

The balance sheets presented in the previous section can be utilized to show that capital flows associated with official intervention in foreign exchange markets do not, in themselves, affect total credit

available to U.S. nonbanks.

- 19 -

Assume an official intervention sale of dollars is initiated by the

foreign central bank and that it is fully sterilized so that ~agmet =

+ acm? and yt and xP? are unchanged .=/ Utilizing the balance of payments identity:

(6) ap=?2 _ apo22 _ apn? 4 apt _ ag2t =0

and the U.S. Bank Balance Sheet Identity:

(7) -apt™) ~ ap?4@t 4 arttt + au21 4 ago? + achtt = 9 And adding equations (6) and (7) and noting that -agn2t = act, (3) apt? 4 apDt = ap? 4 apt

which states that the condition noted above, that total bank credit available to U.S. nonbanks depends only on the total bank deposits held by U.S. residents, is unaffected by sterilized intervention.

The ultimate holders of the U.S. Treasury securities will be detarmined by the relative preferences of different holders for various dollar assets. Most often a U.S. resident, and frequently U.S. commercial banks, have acquired Treasury securities as foreign central banks have sold them. To the extent that U.S. banks increased their holdings of Treasury securities, the amount of bank credit that they can provide (with a given level of deposits) to private U.S. residents is reduced by an amount equal to the increase in credit extended to U.S. residents from abroad.

However, data covering business loans extended by U.S. banking offices and other funds raised in U.S. credit markets will not encompass —I7 Im the case of U.S. intervention sales of dollars, the Treasury

(or the Federal Reserve) would be selling additional Treasury securities in the market.

- 20 -

all the funds obtained by U.S. private borrowers. Thus, data on domestic banks credit and international capital flows must be analyzed together to obtain a complete picture. In practice, information on the international component is less comprehensive, and available with a longer lag, than information on the domestic component. 9. Conclusions

The factors identified in this paper that have led to the expansion of banks' international transactions suggest a continued rapid growth of such transactions in the 1980s. The implication of this process focused on in this paper is the increasing difficulty it generates for measuring monetary and credit aggregates relevant for evaluating the financial positions of U.S. nonbank residents. It is ironic that in a time when quantitative monetary and credit aggregates have played an increasing role in the implementation of monetary policy in many countries that the measurement of such aggregates is becoming increasingly difficult. In this regard the recent and potential growth of transactions-type deposits in offshore banking offices and the apparent inability of the balance of payments reporting system to capture recent capital inflows are

particularly troublesome.

10.

ll.

12.

-?21-

BIBLIOGRAPHY

Aliber, Robert Z., "Towards a Theory of International Banking,"

Federal Reserve Bank of San Francisco Economic Review, (Spring 1976), pp. 5-8.

Bryant, Ralph C., Money and Monetary Policy in Interdependent Nations, ®rookings Institution, 1980).

Dornbusch, Rudiger, "Exchange Rate Economics: Where Do We Stand?", Brookings Papers on Economic Activity, (No. 1, 1980), pp. 143-206.

Freedman, Charles, "Micro Theory of International Financial Intermediation," American Economic Review, Vol. 67 (No. 1, February 1977), pp. 172-179.

Giddy, Ian, "The Theory and Industrial Organization of International Banking," in Robert Hawkins et al, Editors, The Internationalization

of Financial Markets and National Economic Policy, (Forthcoming, 1981).

Girton, Lance and Roper, Don, "Theory and Implications of Currency Substitution," Journal of Money, Credit and Banking, (Forthcoming).

Grubel, Herbert G., "A Thoery of Multinational Banking," Quarterly Review, Banca Nazionale del Lavoro, (No. 123, December 1977), pp. 349-363.

Henderson, Dale W., "Modeling the Interdependence of National Money and Capital Markets," American Economic Review, Vol. 67 (No. 1, February 1977).

Isard, Peter, The Process of Exchange-Rate Determination: A Survey of Popular View and Recent Models, Princeton Studies in International Finance (Princeton, N.J., No. 42, 1978), International Finance

Section, Princeton University.

Johnson, Harry G., “Reserve Requirements and Monetary Control," Canada Council Discussion Paper (No. 66, October 1976).

Lowrey, Barbara and Smith, Ralph, "U.S. Banks, Exchange Market, and the Dollar, September-November 1978," International Finance Division Discussion Paper (Federal Reserve Board, March 1980).

Simpson, Thomas D., "The Redefined Monetary Aggregates," Federal Reserve Bulletin, Vol. 66, (February 1980), pp. 97-114.