Modeling Investment Income and Other Services in the U.S. International Transactions Accounts

International Finance Discussion Papers Number 319

December 1987

MODELING INVESTMENT INCOME AND OTHER SERVICES IN THE _ U.S. INTERNATIONAL TRANSACTIONS ACCOUNTS

by

William Helkie and Lois Stekler

NOTE: International Finance Discussion Papers are preliminary materials circulated to stimulate discussion and critical comment. References in publications to International Finance Discussion Papers (other than an

acknowledgment by a writer that he has had access to unpublished material) should be cleared with the author or authors.

Abstract

This paper presents the services account sector of a model of U.S international transactions (the USIT model) that is maintained in the Division of International Finance of the Federal Reserve Board. Part I presents the models for payments and receipts on direct investment, other investment income, and non-investment services. Part II reports on simulations that indicate the sensitivity of the model’s forecast to changes in its predetermined variables such as interest rates and exchange rates. In particular, we explore the implications of large current account deficits and the resulting accumulation of net claims by

foreigners on the United States for the services balance.

Modeling Investment Income and Other Services in the U.S. International Transactions Accounts

by 1 William Helkie and Lois Stekler

Movements of net services have taken on an increasingly important role in U.S. international transactions over the past two decades. Any effort to analyse and project developments in the U.S. external balance must include an effort to quantify factors influencing net service transactions. This paper presents the services account sector of a model of U.S. international transactions (the USIT model) that: is maintained in the Division of International Finance at the Federal Reserve Board.

As shown in Table 1, U.S. net service receipts grew rapidly in the 1970s, reaching a peak of $44 billion in 1981. This large net surplus on services resulted in a growing gap between the balance on merchandise trade on the one hand and the balance on goods and services and current account on the other. In the 1980s, however, the sharp appreciation of the U.S. dollar had a negative impact on trade in services as well as trade in goods. U.S. noninvestment services became less competitive with foreign-produced services. Net investment income declined because profits earned abroad by U.S. direct investors were equivalent to fewer dollars, and other investment income fell because the

current account deficits were accompanied by a rapid increase in U.S. net

1. The authors wish to thank other members of the staff, and in particular Peter Hooper and Kathryn Morisse, for their contributions to the development of these models. This paper represents the views of the auzhors and should not be interpreted as reflecting those of the Board of Governors of the Federal Reserve System or other members of its staff.

la -

LOEcS €9vT TL86%7- vEeETS LOOOT- 61ST9- ZISTSs TS80€ 978S- L699€ 9861

€6E7~ soot- escsly- €Ss9V 8sit- cC89S- = §=97€9SS 9899¢ 6L09- S99CE S861 67296T sett 9L8E7- vTOSt £099 O61T8S- 69479 S8é6rT 6¢76- LICT? 786T LSSOE 7389S SceLe= LOoOecy 7L66 SLL9H- c7SL9S TO6%7T 86SS- 66702 €g6T 7669€ oces LOSvE= LE8tY 0v70T 8cLTS- s89Tc9 7C78T 9STE- O8eTz c86r SOsey £276 6ESEE- 19627 Teves Tevsy- cggEes TS9Sc 8689- 67SCE TS86T EsTtle 9629 T790€- LEVLE 9L8T Y87EC- O9CSE TIS8Z2 S€98- 9YTLE Og6T cSHVE O8cEe S6E8C~ SLOTE $S9- 7099C- 676SZ2 978TE LSE9~ esTse 6L6T OSvET S8gc LESY~- = ETLLT 789- 89nLT- L8L9T LYCT? TIcv- 8S7Sc 8Z6I $L002 yITe LT8TC- TE6E?S ccrt €selt- SOSeT 6E89T PE8T- €L96T LL6T Sv77el OLNZ SL96T- vvT?te 98 TOzOoT- L8cot 688ST OTTE- 6668T 9L6T Le8vlt 0702 TOzLT- T7261 yLST- O€COT- 9628 T9E4T VECC= S6S9T SL6T ossgT 870T 9HE9T- 76ELT €7EC- €SLOT- O€%”8 9c8LT TeeT- 9ST6I 7L6T covet 672 HTtHT- €974T 6LLCE~ S708- 99S ce6yT OT9T- cvS9ot €L6t €608 66- 780¢T- S86IT cLyt- 88zS- 9TSE $996 78cT- 6760T cL6T LLSL Soe T8sot- 9880T 9CL- TLev- LYSe 9664 79TT- O9T6 TL6T TAS) 262 $996- £566 c90T- 0179%- 8LSE S6cL 7L8- 69T8 0L6T i LIN LIN SINSWAVd SIdI9OdY LIN SINSWAVd SIdIdOae LIN SINSNAVd SIdTdOoe UVaA ea a a a aaa SHOTAWAS AHOONT AWOONT ANOONT ‘IVLOL AOIAWAS AFHLO INANLSYANT YaHLO LNUWLSGANT LOYYIG

(SYVTIOd 40 SNOITTIW) . SINAWAVd GNV SLdIdOaeY FOTAMAS T FIdvL

debt to foreigners. The decline in the foreign exchange value of the dolla: since early 1985 has tended to reverse the decline in net receipts from direct investment and other services and, hence, to counter in part the impact of continued increases in net U.S. foreign indebtedness.

Part I of this paper describes the USIT model of the services sector in the U.S. balance of payments accounts. The components include payments and receipts on direct investment, other investment, and noninvestment services. Part II explores the sensitivity of the services balance to changes in underlying factors such as interest rates, and exchange rates. In particular, we explore the implications for the services balance of the accumulation of net foreign claims on the United States.”

I. The Services Sector Model

The services sector model is divided into three major parts: a) direct investment, b) other investment income, and c) other services. Structural equations were estimated for receipts and payments in each category, using quarterly data. Structural models were preferred over time-series techniques because of our interest in making conditional simulations and forecasts, e.g., assessing the impact of changes in underlying assumptions about economic activity, interest rates, or exchange rates in such exercises. The choice of explanatory variables in these models was constrained to some extent by the need to use variables

for-which projections could be readily obtained.

2. An appendix that provides a documented listing of the model is available on request.

A) Direct Investment Receipts and Payments

This section presents (1) the direct investment receipts and payments equations, (2) the regression estimates for these equations, (3) the direct investment asset stock equations, and (4) the regression estimates for these equations.

A.1. The direct investment receipts and payments equatioas

The general approach used to model direct investment receipts and payments is to assume that earnings depend on the scale of investment and on the variables that cause fluctuations in nominal dollar profit rates.” The models for both receipts and payments are analogous. However, the receipts side is disaggregated into three industry categories: manufacturing, petroleum, and other.” The general model for nominal receipts is presented first, followed by the modifications necessary for individual industries, and finally by the modifications necessary to apply the model to payments.

Actual receipts (nominal) on U.S. direct investment abroad (YR) in a particular year can be characterized as the product of both long-run

or expected receipts (EYR), and temporary or unexpected factors (x; )- If

3. Previous models of direct investment receipts and payments (Bond 1977, 1979 and Kwack 1979) have used the Department of Commerce estimates of the U.S. direct investment position as the measure of the scale of investment and some interest rate (e.g. the rate on long term U.S. government bonds) as the measure of rate of return. We rejected this approach because a dollar added to the direct investment position in 1985 is not likely to have the same impact on earnings as a dollar added in 1975 and because, while movements in interest rates might coincide with movements in profits resulting from cyclical factors or inflation, it seemed preferable to model the impact of these factors directly.

4. Data by industry are available from the Survey of Current Business, U.S. International Transactions, Table 5.

nothing unexpected happens, the Xx, are defined equal to one and actual

receipts equal expected receipts.

YR = (EYR) « n> xat (1) In order to explore the determination of expected receipts, consider a specific direct investment project that requires a certain dollar expenditure this period (I), and is expected to produce a certain level of average earnings, constant in real terms, in perpetuity. The expected real rate of return on the project (r) equals expected annual earnings in foreign currency (EN) divided by the foreign currency value

of the initial investment (I/e,, where e equals the exchange rate when the investment is made) .°

EN = r « Ife, (2) If inflation and exchange rate changes do not alter real rates of return

on investment in a predictable way, then expected nominal earnings in a

. : 7 currency increase at the same rate as average prices.

5. Lags will be considered later. The assumption of a perpetual income stream is not as far-fetched as it may first appear if capital consumption allowances accurately reflect depreciation in the economic sense. The Department of Commerce data exclude depreciation from both income and new investment. The reinvestment of capital consumption allowances maintains the productivity of the capital stock.

6. e equals the exchange rate (cents per unit of foreign currency) in the year when the investment was made. e, equals the current exchange rate.

7. I£ inflation or exchange rate changes alter relative prices, in particular the price of the product produced by the investment relative to the costs of inputs, then real rates of return are affected.

as} rg

t I t

(2a)

La] © rg

(eo) ° oO

Expected nominal earnings converted to dollars equal the rate of return times the initial investment adjusted for subsequent exchange rate and

price changes.

rg ct

rg ct

@ ct

EN « -e =retles (2b)

la] fo} ct ia] ° () °

Total expected U.S. receipts on direct investment (EYR) equal the sum over all previous investments (i=l to n) of the rate of return on each investment multiplied by the initial amount invested, adjusted for

subsequent price and exchange rate changes.

/?P e n “t t\ BYR, = *j21 7 ° Ty -( P els

(3)

To simplify, assume that r is the same for all investments; then expected receipts can be approximated by average r times the "current" dollar

value of previous investments.

tJ

n t St EYR, =re raat I, (= - =| = re AR (3a) ro)

The "current" dollar value of previous investments, AR. dces not correspond to the Department of Commerce data on the U.S. direct investment position. The Commerce data is book value, the sum over past years of nominal investment flows; it is neither in constant nor current dollars. We constructed a time series for the variable AR by starting with the Department of Commerce asset position at the end of 1971 and for

each subsequent quarter (1) inflating the asset position by world average

-6-

price increases adjusted for exchange rate changes and (2) adding the

current direct investment outflow (DI) in that quarter.

P e P e t t t t AR. = AR . ( = . —_— ) + DI . . —_ (3b) t 71 Poi e741 72 Poo e755 P e +... + DIL y > = ) t-1 fe-1

Actual receipts may not always equal receipts that are expected over the long run. Short-run transitory or unexpected factors (X;) influence profits by altering demand for the firm’s output, by changing the xelative prices of inputs and outputs, or by creating inventory profits. While many such factors influence receipts, only a few are included in this study. The influence of the cyclical behavior of aggregate demand on profits is represented by a capacity utilization variable. Capacity utilization (CU) is measured by the average level of foreign GNP relative to its potential. Other factors include the reporting of inventory profits due to inflation or exchange rate changes when first in-first out (FIFO) accounting is used, and the inflation of ‘nominal profits when depreciation charges fail to reflect replacement costs. In order to capture these effects the percentage change in prices adjusted for exchange rate changes (P,- e, / Pea: &t -D is included in the equations for manufacturing and other industries for the years before 1982. After this date a switch in accounting standards (from Financial Accounting Standard (FAS) 8 to FAS 52) meant that in most cases the impact of changes in exchange rates were reported separately as capital

gains or losses.

-7-

Capital gains or losses present special problems; they are included in the all-inclusive concept of earnings used by the Department of Commerce in the balance of payments accounts. For example, if the subsidiary of a U.S. firm bought an asset ten years ago and then sells it today at a higher nominal price, but the same real price, the Department of Commerce data will show higher income for this year. The estimated model would be unlikely to predict accurately the size or timing of such realized gains even though some of the variables in the model might influence such capital gains. The model would also be unlikely to capture unrealized capital gains or losses due to the revaluation of balance sheets or to understated inventory and depreciation charges when the dollar depreciates.° Because of the difficulties of explaining these capital gains we have chosen to exclude them from both receipts and payments, and to estimate them separately.

The capital gains equations do not attempt to model the timing of capital gains or losses resulting from the sale of assets at prices differing from book value. For simulation purposes, these capital gains and losses are assumed to be zero. On the other hand, gains or losses on direct investment receipts resulting from the revaluation of balance sheets when exchange rates vary, or resulting from the impact of exchange rate changes on inventory or depreciation charges taken directly to equity under FAS 52, are expected to vary with percentage changes in the

foreign exchange value of the dollar (e/e(-1)). Because of the switch of

8. These were included in profits reported to stockholders between 1976 and 1981 under FAS8, but now are taken direct to equity under FAS52 and show up as capital gains or losses in the balance of payments accounts.

accounting standards from FAS 8 to FAS 52 in 1982, a dummy is used to divide the sample period.

Finally, we have chosen to separate transactions with Netherlands Antilles finance affiliates from transactions with other affiliates in the other industry category. One of the primary functions of these affiliates was to raise funds in the Euromarkets to finance the domestic operations of U.S. corporations. Neither capital nor income flows between the United States and these finance affiliates is likely to be determined by the same variables that determine other direct investment flows. For this reason payments to these affiliates are projected judgmentally, based upon current levels and projections of net Eurobonds outstanding at these affiliates.”

The framework used in modeling direct investment payments is very similar to that used for receipts. One difference is that payments are not disaggregated by industry. Of course, U.S. variables are substituted for corresponding foreign variables.

A.2, Regression results: direct investment receipts and payments

The regression results for direct investment receipts and payments are shown in table 2. While the coefficients have the expected signs and generally are statistically significant, the standard error of the regressions range from 8 to 18 percent for the receipts categories and is 19 percent for payments. It is very difficult to capture all the

special factors that cause fluctuations in direct investment receipts and

9. Since the repeal of the withholding tax by the United States in mid- 1984, most Eurobonds have been issued directly from the United States. The amounts outstanding at Netherlands Antilles affiliates have been falling.

-~ 8a -

Table 2

Regression Results for Direct Investment Income Equations 1/

Direct Investment Income Receipts

Manufacturing Petroleum Other

Intercept 2.44 -1.44 -0.86

(1.28) (-0.61) (-2.80)

Capacity utilization (CU) 6.42 9.51 3.64

(1.98) (2.45) (4.49)

Current value of stock 0.48 0.88 0.78

(AR) (2.78) (3.89) (27.62)

Price change 2/ 2.99 -- 0.96

(P, - e,/ Pea: &-1) (3.04) (1.63)

Rho 0.70 0.75 0.18

(6.76) (7.98) (1.29)

R2 0.76 0.86 0.97

SER 0.159 0.176 0.082

DW 2.29 1.85 1.87 1/ Equations are in logarithmic form; the numbers in parentheses are

2/ Adjusted for exchange rate changes in the receipts equations.

Direct Investments Income Payments

-2.76 (-2.81)

6.60 (3.35)

0.95 (9.25)

0.75 (10.07)

0.95 0.187 2.28

t-statistics.

payments. Receipts in manufacturing and the petroleum industry appear more sensitive to fluctuations in overall foreign capacity utilization than "other" industries, which includes real estate and finance. The response of earnings to higher price levels, exchange rate changes and asset growth is far less than proportionate only in the case of manufacturing. Price chariges appear to have an important impact on earnings (presumably through inventory profits) only in the case of manufacturing.

The regression results for the capital gains or loss equations in manufacturing and "other" are shown in table 3. No significant relationship was apparent for the petroleum industry, so capital gains are assumed to be zero for that industry in our simulations. Since the capital gains experienced by a corporation as the result of exchange rate changes depend on the firm’s management of foreign exchange risk, changes in exchange rate expectations, attitudes toward risk, or methods of risk management would alter the observed relationship between gains and exchange rate changes over time. Therefore, the coefficients may not be stzble.

A.3. Direct investment asset stock equations

In order to simulate or forecast direct investment receipts and paymerits using the model described above, projections of the current dollar value of previous investments (AR) and therefore direct investment capital flows are also necessary. Conceptually we model these direct

investment capital flows using a sources and uses of funds approach. !°

10. There is a vast theoretical literature on direct investment, suggesting many considerations in addition to the ones discussed below. (e.g., taxes, protectionism, export penetration). However, previous (Footnote continues on next page)

- 9a -

Table 3

Regression Results for Capital Gains/Loss Equations 1]/

Direct Investment Income Receipits Direct Investment Manufacturing Petroleum Qther Income Payments

Intercept -100.9 -- -8.5 -- (-1.28) (-0.21)

Exchange rate/. -25262.4 -- -14152.4 -exchange rate (-1) 2/ (-8.10) (-8.94) Ro 0.65 0.70 SER 461.6 234.6 DW 2.01 2.36

1/ The independent variables are in logarithmic form; the numbers in parentheses are t-statistics. The dependent variable is in levels. 2/ Coefficients for the period 1982Q1 on.

- 10 -

Direct investment capital flows (DI) and the resulting direct investment position are not viewed as objectives in their own right for multinational corporations. They are reviewed instead as a residual resulting from production, financing and location decisions. The size of direct investment flows is expected to be influenced by all factors that, in turn, influence other sources and uses of funds of foreign affiliates.

The main sources of funds for a foreign affiliate would include the following:

DIE - direct investment equity and intercompany account flows

RE - reinvested earnings

DEP - depreciation charges, and

B - net borrowing of funds from unaffiliated external

sources. 1+

The main uses of funds would include the following:

I - investment expenditures in plant and equipment, and

AINV - changes in inventories. Since sources of funds equal uses of funds, the following equations must hold: |

DIE + RE + DEP + B = I + AINV (4)

DI = DIE + RE = I + AINV - DEP - B (5) Each of the variables on the right-hand side of equation (5) is, in turn,

a function of other variables. No attempt was made to estimate these

(Footnote continued from previous page)

direct investment equations incorporated in larger models (Kwack, 1979, for example) tend to view direct investment capital flows as reflecting only investment in productive capacity.

ll. To simplify, the possibility of the affiliate selling equities to

parties other than the parent is ignored.

relationships separately because adequate data is not available. Instead these functional relationships were substituted into equation (5) to derive a single equation for direct investment capital flows. The functional equations for each of these right-hand side variables are described below. For ease of exposition we have presented equations with all relationships contemporaneous, although in estimating we frequently have incorporated distributed lags.

Investment in plant and equipment (real) depends on replacement needs, the growth in demand for output, and the relative attractiveness to the multinational corporation (MNC) of expanding production abroad compared with expanding production in the United States. Replacement needs (REP) are some fraction of the existing stock of assets. The growth in demand for output is assumed to parallel the growth in real GNP. Since the MNC can supply that increased demand by expanding production in the United States or abroad, changes in relative costs are expected to influence the location of investment. Relative costs are measured by changes in relative wage rates adjusted for exchange rate changes cacomp) . 22

This model of real investment can be summarized by the following equation:

I/P = ay AGNP + ay ACOMP + REP (6)

12. Relative wage rates were not adjusted for productivity changes on the grounds that most of the differences in productivity across countries reflect differences in technology and capital intensity. It was assumed that a MNC can use the same technology anywhere. The use of relative wage rates as a measure of relative costs will be in érror to the extent that changes in relative wage rates reflect changes in human capital.

- 12 -

Multiplying through by prices to get an equation for nominal

investment yields:

I= ay P * AGNP + ao P * ACOMP + P * REP (7)

where I = nominal investment in dollars,

P = average foreign prices adjusted for exchange rate changes in the outflow equations, and U.S. prices in the inflow equations,

COMP = average foreign wage rates converted to dollars

divided by the U.S. wage rate (twelve-quarter moving average),

GNP = real U.S. GNP in the capital inflow equations and a weighted average of foreign GNP indices in the capital outflow equations (distributed lag), and

REP = real economic depreciation of the capital stock, assumed to be a constant proportion of the real capital stock (DIK).

The change in the value of inventories (AINV) depends on the change in quantity and the change in price. If actual inventories equal desired, and if the desired volume of inventories is a function of aggregate demand (GNP), then the change in inventories can be approximated by the following equation:

AINV = by P * AGNP + by A P * GNP (8)

The price variable ideally would be the price index for the goods held in inventories. For the petroleum direct investment outflow equation, a price index for crude petroleum was tried, but for other industries, general price indices were used.

Depreciation charges (DEP) depend on accounting rules and on the

nominal size of previous investments; they are assumed to be a constant

- 13 -

fraction of the book value of investment. 1? If depreciation charges were equal to the current replacement costs for worn out or obsolete equipment, then the depreciation term would just cancel out the replacement term in the investment function (P * REP). But inflation is likely to cause differences between depreciation charges and replacement costs. Thus,

DEP, = * DIP. 1 (9)

“1

where DIP. = direct investment position (book value) at ime t.

Net borrowing from outsiders (B) by the foreign affiliates depends on the relative cost to the MNC of borrowing by the home office or by the foreign affiliates. This borrowing is assumed to be a function of the difference between domestic U.S. and Eurodollar interest rates. Both short and long-term differentials were tried.

B= dy (URTBILL - URES) (10)

where URTBILL = rates on U.S. Treasury bills or A rated atility

bonds |

URES = LIBOR or Eurodollar bond interest rates

In addition, if the need for borrowing from outsiders is related to the level of profits and reinvested earnings (as it would be if fluctuations in reinvested earnings were not compensated for entirely by movements in direct investment equity and intercompany accounts), then

there is a simultaneity problem. Moreover, the estimated impact of

13. This approximation is admittedly very crude. Also, the book value of plant and equipment investments is unavailable; therefore the book value of the direct investment position was substituted.

- 14 -

changes in GNP on direct investment capital flows would reflect the impact on profits and reinvested earnings as well as on investment in productive capacity and inventories.

Substituting equations (7), (8), (9), and (10) into equation (5) and combining terms produces equations of the following form:

DI = f; (P * AGNP) + £, (P * ACOMP) + (11) £, (AP * GNP) + £, (DIK * P) +

f. (DIP, 1) + f. (URTBILL - URES) The expected signs of the coefficients f)> f, and f, are positive and f. is negative. The expected signs of f, and f. are positive in the capital inflow equation and negative in the capital outflow equations.

Direct investment capital outflows were disaggregated and estimated separately for three industries: petroleum, manufacturing, and all other. Direct investment inflows were not disaggregated. Many of the independent variables were expected to influence capital flows over an extended period of time. Therefore distributed lags or moving averages were used. The sample period for outflows starts in 1972Q1.

- (Quarterly data were not available for earlier quarters by industry.) Seasonal factors were taken into account by the use of seasonal dummies.

The capital outflows in all cases were adjusted to exclude capital gains. This is clearly appropriate, for example, in the case of unrealized currency translation gains; reinvested earnings are inflated by these gains. It also seems appropriate in the case of a realized capital gain or loss resulting, for example, from the sale of an asset at

less than its book value if the proceeds are not paid out as dividends.

None of the explanatory variables included in the models are like-.y to explain the timing of the realization of these gains or losses. A.4. Regression results: direct investment asset stocks

The regression results for the equations included in the model are shown in table 4. Im general, direct investment capital flows seem to be larger, the more rapid GNP growth in the receiving country, the higher replacement costs for capital, and the less adequate depreciation charges. However, the explanatory power of the equations is not very high; in particular, the variables included explain little of the variation of direct investment outflows in the petroleum industry. Not all the variables theoretically relevant are included in every equation. Variables that were insignificant or had signs inconsistent with theory were dropped. In particular we were never able to estimate a relationship between direct investment flows and the relative costs of producing in the United States and abroad. Rather than conclude that relative costs play no role, we would guess that the negative results reflect the difficulty of accurately measuring relevant comparative costs, particularly because the investment decision depends on forwardlooking expectations of such costs. In addition, the role of price increases in causing increased inventory investment was insignificant in all industries, as was the relative cost of raising funds here and abroad. B) Other Investment Income

Other investment income in the U.S. balance of payments accounts includes four components: U.S. government investment income receipts and

payments, and private receipts and payments on investment excluding

- 15a -

Table 4

Regression Results for Direct Investment Capital Flow Equations 1/

Intercept

Price * change in GNP (P * A GNP)

Price * Real asset (DIK * P)

Nominal Asset (DIP 4 )

Rho

R2

SER DW

Direct Investment Capital Outflows

Manufacturing

2210.2 (3.01)

2663.0 (4.72)

0.044 (2.79)

-0.088 (-2.88)

'-0.37 (-2.78)

0.38 714.4 1.86

Petroleum

3690.0 (2.08)

456.9 (0.35)

0.096 (1.99)

-0.175 (-1.94)

0.15 (1.10)

0.10 1084.3 1.93

1/ The numbers in parentheses are t-statistics.

Other

2425.9 (4.20)

1797.3 (3.27)

0.063 (4.38)

-0.099 (-4.07)

0.02 (0.14)

0.45 520.8 1.97

Direct Investment Capital Inflows

-137.7 (0.49)

0.157 (1.96)

0.003 (2.21)

-0.129 (-1.85)

0.03 (0.11)

0.76 941.9 1.98

- 16 -

direct investment. 4 Only the government receipts and payments reflect actual reported data. The Department of Commerce estimates private receipts and payments on investment by breaking down the appropriate stock of assets (or liabilities) into components and multiplying these components by an assumed appropriate rate of return that is gathered by a survey. Our model of all four components of other investment income mimics this procedure as do most other models of portfolio investment income; earnings are assumed to be the product of the stock of assets times an average rate of return. Historical observations for this average rate of return were imputed as the ratio of earnings (Y) to the aggregate asset stock (ay .2? Regressions were then used to establish the average relationship between these imputed interest rates and observed interest rates. The U.S. Treasury bill rate (r) was used for both receipts and payments because the bulk of both claims and liabilities are dollar denominated. In addition, the lagged dependent variable is included in the regressions to take into account the fact that some of the assets have maturities of longer than a quarter and that their rate of return will depend on earlier interest rates.

¥/A = 8 + &)(4) + B3(Y/A)_| (12)

The regression results for the implicit interest rates aie shown in table 5. The corrected R’s are .98 or higher except in the case of the implicit interest rate on government receipts (R? = .85). This

somewhat lower explanatory power is perhaps not surprising since a large

14. Survey of Current Business, Table 1, U.S. International Transactions, lines 13, 14, 28 and 29.

15. Only annual data is published for the U.S. investment position. Quarterly data can be estimated by using reported capital flows and interpolating quarterly data from the annual valuation adjustments.

- loa -

Table 5

Regression Results for the Implicit Interest Rates on Portfolio Investment Income 1/

Private Government Private Government Receipts Receipts Payments Payments Intercept 0.18 0.07 -0.06 0.03 (2.83) (0.72) (-1.21) (0.83) 90 day T-bill rate 0.50 0.08 0.51 0.19 (12.53) (1.30) (11.83) (7.93) Implicit Interest 0.45 0.82 0.47 0.80 rate (t-1) (12.38) (9.35) (9.81) (25.61) R? 0.98 0.85 0.98 0.99 SER 0.028 0.097 0.049 0.033 DW 2.40 2.75 2.05 2.11

1/ iquations are in logarithmic form; the numbers in parentheses are t-statistics.

part of U.S. government claims on foreigners dre the product of foreign aid and do not reflect market interest rates and because U.S. government interest receipts are reported rather than estimated. The long-run elasticity of the implicit interest rate on government claims with respect to a change in the Treasury bill rate is also much lower than for the other implicit rates, probably for similar reasons; the other coefficients imply a roughly proportionate impact of changes in Treasury bill rates on the implicit interest rates.

When simulating other investment income, in addition to the equations above, assumptions about interest rates and the relevant asset stocks are also necessary. Interest rates are set exogeneously. The details of our method for modeling the asset stocks are still evolving, but the general framework constrains the changes in assets and liabilities to be consistent with balance in the U.S. international transactions accounts. A common method for doing this is to estimate independently equations for the major components of capital flows and to use the statistical discrepancy to balance the accounts. "© We rejected this approach for two related reasons. First, given the poor performance of most capital flow equations, this method tends to produce swings in the statistical discrepancy far outside the range of recent experience. Second, estimates of investment income are based only on recorded capital flows; large errors in the statistical discrepancy would cause large errors in simulations of changes in investment income. We chose instead to use net private claims as the balancing item in the accounts.

The derivation of each component of the capital account is

summarized in table 6 along with the source for each component. The

16. See, for example, Proctor, 1982.

- l7a -

Table 6 Derivation of Changes in Private Claims and Liabilities

Surrent account balance and unilateral transfers less net portfolio investment income. Current account model.

Statistical discrepancy. Exogeneous.

Net direct investment capital inflow. Structural model described in previous section of paper.

Changes in government liabilities to foreigners. Structural model based upon foreign official intervention function adjusted

judgmentally for likely changes in OPEC and LDC assets in the United States as well as private purchases of Treasury securities.

SDR allocations. Exogeneous.

Changes _ in government claims on foreigners. Exogeneous (-indicates increases).

Change in net private claims on foreigners. Line (7) = - 1 * (lines 1+2+3+4+4+5 + 6).

derivation of line 4 requires further explanation. U.S. government liabilities to foreigners are assumed to depend on foreign official claims plus net private foreign purchases of Treasury securities.

Foreign official claims on the United States are assumed to increase when foreign governments intervene to support the dollar in foreign exchange markets; intervention is assumed to depend on exchange rate movements and the stock of official reserves held in the United States. The regression results are shown in table 7. Private foreign purchases of Treasury securities were small until 1984. They are exogeneous, as are (PEC and LDC official holdings.

The change in net private claims on foreigners (line 7 projected by this method has to be split between changes in claims and changes in liabilities in order to project receipts and payments.

Because the rates of return earned on claims and liabilities differ, this split affects net earnings as well. Log-linear trends are used to project both private claims and liabilities. The resulting regressions are shown in table 8. Use of a shorter sample period (e.g., starting in 1973 instead of 1965) would have resulted in faster average growth rates. Using the lower rates, based on the longer sample period, implies a slowdown in the growth of international intermediation through U.S. banks from the very rapid pace reached in the 1970s. Net private claims produced by these trend projections are then compared with the net projected by using the balance of payments constraint, and the discrepancy between the two is calculated. The gross changes in claims and liabilities projected by the trend equations are then adjusted to

eliminate the discrepancy. Since the mean absolute size of the errors in

-18a -

Table 7

Regression Results for Change in Government Liabilities to Foreigners 1/

Intercept 6074 (2.12)

Excnange Rate/Exchange Rate ‘(-1) -71519 (-3.21)

Trend 808 (3.04)

Government liabilities to -2.09 foreigners (-1) (-2.63) Re 0.21 SER 5089 DW 1.47

1/ Exchange rate variables are in logarithmic form; the numbers in parentheses are t-statistics.

- 18b -

Table 8

Regression Results for Private Liabilities and Claims Trend Equations 1/

Claims Liabilities Intercept 9.88 10.47 (246.7) (361.6) Time (1=1965Q1) .0420 .0312 (48.95) (50.26) Re 97 97 SER .177 .128 DW .036 .056

oe 1/ Dependent variables are in logarithmic form; the numbers ir. parentheses are t-statistics.

- 19 -

the claims and liabilities trend equations were approximately equal, the

adjustments used are also equal.

GC) All Other Services

We have aggregated all services other than investment income. Included in "all other" services are travel, passenger fares, other transportation, fees and royalties, other private services, and U.S. government miscellaneous services. -/ We assume that imports of these services are related in general to the same variables that determine imports of goods; the real level of these services imports is a function of relative prices and income. In addition, expenditures on

transportation and insurance are likely to be related to the level of

trade. MS PMS In bys «== hy + hy In Soy + h, 1n GNP + h, 1n MG (13) XS ‘ PXS | In fxs 75, +5, IM ppt j3 1n FGNP + j, In xe (14) where:

FGNP = weighted average foreign GNP

GNP = U.S. GNP

MG = U.S. imports of goods

MS = U.S. imports of other services (nominal)

PF = weighted average foreign consumer prices adjusted for exchange rate changes.

PGNP = U.S. GNP deflator

other services payments implicit price deflator

other services receipts implicit price deflator

U.S. exports of goods

exports of other services (nominal)

aa is an eid

17. See Survey of Current Business, Table 1, U.S. International Transactions, lines 4 through 11 for receipts and 19 through 26 for payments. Transfers under U.S. military agency sales contracts (line 3) and direct defense expenditures (line 18) are forecasted judgmentally.

- 20 -

The regression equations for all other service receipts and payments are shown in table 9. The R*s are reasonably high and the coefficients have the expected signs and are statistically significant. The regression results seem to indicate that the income elasticities of foreign demand for U.S. "other" services and U.S. demand for foreign "other" services are approximately the same size. The price elasticity of foreign demand for U.S. services appears slightly lower than the price elasticity of demand for foreign services. The impact of trade volume on other services is approximately symmetric. In addition to the other services receipts and payments equations, the model includes equations for the import and export price deflators. (See table 10.) These equations are designed to mimic the way these price deflators are constructed. The implicit price deflator for other services exports is related to the U.S. CPI and GNP deflator. The implicit price deflator for other services imports is related to the U.S. GNP deflator and weighted average foreign consumer prices adjusted for exchange rate changes.

II. Simulation Results

The model presented in Part I can be used to analyze the impact on the U.S. services balance of changes in various explanatory variables. To do so, the services account model is simulated under the following; shocks:

a) a 100 basis point increase in the U.S. Treasury bill rate,

b) a 10 percent depreciation of the weighted average dollar,

c) 1 percentage point (annual rate) higher foreign GNP growth rate.

d) 1 percentage point (annual rate) higher U.S. GNP growth rate.

- 20a -

Table 9

Regression Results for other Services (Excluding Investment Income) 1/

Receipts Payments

Intercept -3.08 -8.40 (-9.26) (-9.27)

Relative Price -0.48 -0.62 ; (-9.60) (-7.88)

GNP 2/ 0.75 0.73 (14.73) (5.11)

Real U.S. Exports (Imports) 3/ 0.22 0.27 (5.36) (5.25)

Rho . 0.42 0.58 (4.07) (6.14)

Ro 0.99 0.97 SER 0.017 0.024 DW 2.14 1.84

1/ Equations are in logarithmic form; the numbers in parentheses are tstatistics.

2/ Foreign GNP for receipts; U.S. GNP for payments.

3 Exports for receipts; imports for payments.

- 20b -

Table 106

Regression Results for Other Service (Excluding Investment Income) Price Equations 1/

Receipts Payments

Intercept -0.001 3.06 (-0.01) (6.38)

U.S. GNP Deflator 2/ 0.389 0.433 (4.03) (4.63)

Consumer Price 0.611 0.676 (7.73) (7.87)

Rho -- 0.87 (17.7)

Ro 1.0 0.999 SER 0.008 0.013 DW 0.67 1.49

1/ Equations are in logarithmic form; the numbers in parentheses are tstatistics. 2/ Fixed-weight price index for receipts.

e) 1 percentage point (annual rate) higher foreign inflation rate,

f) 1 percentage point (annual rate) higher U.S. inflation rate, and

g) A $100 billion increase per year in U.S. net indebtedness to foreigners, evenly divided between increases in U.S. portfolio liabilities to foreigners and reductions in U.S. portfolio claims on foreigners. These simulations include only partial equilibrium direct impacts of changes in the above variables. For example, a change in U.S. interest rates affects investment income only through its direct impact on the rates of return earned on claims and liabilities, and not indirectly through its possible effects on U.S. GNP and exchange rates. The simulation results are shown in tables 11-17. The simulation results indicate the deviation from the base case solution for the years 1985 and 1986. Since the simulations are partial equilibrium in nature, these simulations should not be treated as estimates of what actually would have happened historically under the counter-factual assumptions. They are designed merely to illustrate the partial-equilibrium properties of the model. A) Higher U.S. Interest Rates (Table 11 A 100 basis point increase in the U.S. Treasury bill rate, ceteris paribus, increases net service receipts over the base case solution by $0.5 billion in the first year when the United States was a net: creditor to the rest of the world on portfolio investment, but recluces net receipts by $0.9 billion in the second year, when the United States shifted to a net debtor position on portfolio investment. The size of the impact depends not just on the size of the interest rate shock, but also on the U.S. net asset position. In addition, because of

the method we use to constrain the balance of payments to sum to zero,

- 2la -

26°0-

TZ°L sl°9

986T

£s‘°0 Ly’ 00's 00°0 00°0 00°0

£$°0

Se6t

Is*tgL°3 92°L 00°0 00°0 00°0 Is*t- 82°83 92°L 00°0-

bb 9861 Sb 986T Z> 986T TD 986T

ST°T- 22°0- 30°83 2c°k £6°9 09°9 00°0 00°0 00°0 00°0 00°0 00°0

GT°T- 22°0-

go's 4 aerd $6°9 09°9 00°0- 00°0- 00°'0 00°0

2 °0s9°9 e¢°9 00°0

2T°0 56°S 90°9 00°0 00°0 00°0 2T°0 b6°S 90°9 00°0-

Is°0 £T°s s9°S 00°0

sjuewAeg pue Szdtecey AJEPTTEW SEPNTOXZ /T

Zz°0 bT*b 16° 00°0 00°0 00°0 Z2°0 yT*d T6°9 00°0-

£2°0 99°? 62° 00°0 00°0 00°0 £2°0 99°2 62° 00°0-

bb G86I cb S86T z> S86T Th S86T

©2ey TITI-1 “S'N JOUSTH 2UTOd Stseq OOT

UOT}NTOS TO4zU0D Wosy EGUeYD 4eTTOG “TTd Cayys ‘s4etTtod °S'N 4° SUCTTTTE)

pUNODSY YUSJIND UO SUOTPIeSUeI] SESTAIOS “S'Nl

“TT @T9eL

ewoouy }ON

sjpuewAed Sjdte9ey TEPOL s@wWOoUT BOTAJES

@wooUT EN syuewAed S}dTel8Yy

ZT s@WOOUT BOTAES J040 . @wooUT 3EN spuewAed Sd Te9eYy

ZT sowooUuy YUSWZS@AUT 40470 @wooUuTy ON spuewAed

S}dTe0ey r@WOSUT PUEWZSEAUT ZEIT

- 22 -

any change in net service receipts implies a simultaneous change in portfolio claims and liabilities, and hence an additional change in net portfolio income receipts. This small factor affects all the simulations.

B) Dollar Depreciation (Table 12)

A 10 percent depreciation of the dollar relative to the base case throughout the two-year simulation period increases U.S. net service receipts by $6.5 billion in the first year and $3.7 billion in the second relative to the base case solution. There are several channels through which exchange rate levels affect service income receipts and payments. U.S. government interest payments on foreign official holdings of Treasury securities go up because the model relates the size of official holdings to exchange market intervention and exchange rate developments. Direct investment income is larger when the value of the dollar is lower because of capital gains resulting from balance sheet translations and because the same amount of income in foreign currency now translates to a larger number of dollars. Direct investment capital flows are also affected by the exchange rate. Other services receipts net are larger because depreciation improves the competitive position of U.S.-provided se:cvices, but this gain is partially offset in nominal terms by the increase in the dollar cost of foreign services as a result of the depreciation. In this partial analysis, net income from private portfolio investment is essentially unaffected because both claims and liabilities are assumed to be denominated in dollars, and because no

account is taken of the feedback between the exchange rate, the

~ 22a -

sjuowAeg pue SzdTedey AYeQTTTW SOPNTOXZ ‘i

sus 1S°9 69°S (96SEC ob°'z) «19ST SséOB VT. @woouy 3°N

68° 6b" I6°t “2e't 06°T gait i. th Ge) on Oe 2 sjuewAed

z9°S «00° 09°S 60 8°S) Sia99"S SES g2°b SS*b COS «OL ET S{dTe00y ; [eyo] sewooUT SOTAIES

LL°0 £2" 0- STI 96°0 99°70 2£°0 So'0- 92°0- = 0- = :92 = @wooUT ON

ss't 80°T 95° 2S'T = $S*T 6b T lool «Of T—sCé TS sjuswAed

6z'2 = S80 69°2 Bhee.SsCST2—C*CT VT Ibot s0°T 99°0 TE°O S{dTe00y yI s@WoOoUT BOTAYES 40430

gs‘0- I£°0- b9°0- 6S°0- -9S'0- = SS0- = bb'O- = LE*O- = OF O- ET" O- @wooul 3°N

9¢°0 0b°0 sc'o 0s SEOs" Ib'0 = bb'0s—«iOS Ose syuewhed

I2'0- 60°0 G2'0- 2°0- «02'0- «= ST'O- «= 00-9 2070 OZ" TO S}d T2004 ZI s@WOOUT YUSWYZSEAUT 49420

os°s = SOL oz*s oss sos TL's 06°z 46Sz's 4 «= og*z@-—s«éSDT* GT. ewoouy ON

00°0 00°0 00°0 00°0 00°0 00°0 00°0 00°0 00°0 00°0 sjuewAed

os°s = SOL oz7s 6s's 29's L's 06°2 Sze o8'2 S2°6T S}d T2004

t@WOOUT PUOWZSEAUT PSCITT

986T Sg6T bb 9861 Sb 986T 2> 986T Th 986T bb Se6T cb S86T zb S86T Th S86T

JeTTIOg “S'N JO UOTzPETOeUdeg XOT

UOT}NTOS [OuzZUOD wou SBueYD 4eETION “Td CavWS ‘S4BTTOd “°S°N JO SUOTTTTA)

punosoy yUeIIND UO SUCTZOeESUeI] S@DTAWES *S’N

ZI @T9@L

- 23 -

merchandise trade component of the current account, the interest rate,

and the levels of portfolio claims and liabilities. }®

C) Foreign GNP Growth (Table 13)

A 1 percentage point higher (annual rate) growth rate in foreign GNP increases net service receipts by $1.5 billion in the first year and $4.3 billion in the second relative to the base case solution. The largest impact is on direct investment receipts: more rapid foreign growth means both higher profit rates and, over time, increased investment. Other service receipts are also affected because higher foreign GNP increases demand for U.S. services. In these partial estimates no account is taken of the interaction between foreign GNP, prices, exchange rates, or trade. D) U.S. GNP Growth (Table 14)

A l percentage point (annual rate) higher growth rate of U.S. GNP reduces net service receipts by $0.6 billion in the first year and $1.5 billion in the second year relative to the base case solution. U.S. GNP affects net service receipts through several direct channels. Direct investment payments are larger, both because profits of foreign-owned affiliates are higher and because more rapid growth of the U.S. economy encourages foreigners to invest more in the United States. Payments for cther services (excluding investment income) also increase with U.S. income. Direct investment receipts, portfolio investment receipts and Fayments, and other services receipts are not directly affected by the

rate of growth of U.S.. GNP. Keep in mind that the simulation does not

18. At least 77 percent of U.S. private portfolio claims and 95 percent of liabilities were clearly denominated in dollars at the end of 1985. In addition, a large percentage of foreign bonds held by U.S. residents are also dollar denominated, increasing the total for claims to as much as 90 percent.

23a -

Tg *o T0‘0os’ 19°0

00°0 19°0

£0°0

10‘°O0- 20°0

9° 00°0 L9°s

986T

os‘*T 00°0 os*T 22°0 00°0 22°0 00°0- 00°0 00°0- 62°T 00°0 62°T

Se6T

23° 20°0- So" 32°0 00°0 32°0 90°0 20°0- <0°0 £0"

00°0 £0°>

bb 9861 sb 986T

“ay “dNO UBFesoy soy OZEY YIMOUN 4OUSTH X T

ss°» 20°0 £s*b 99°0 00°0 99°0 50°0 20°0 20°0 se°s

00°0 ses

yunosoy zUesIND uo SUCT?

- 10°0-

95°0 00°0 95°0 Z0°0 - 10°'0t0°0 49°

00°0 29°

Sb°0 00°0 s>°0 00°0

mo

T° 0° T°

mo ™

UOTPNTOS ToO4zUO

‘sueTTog ‘S'N 4° SUOTTTTA)

62°2 63°T 00°0 00°0 62°2 06°T s¢°0 92°0 00°0 00°0 s<°0 92°0 oo°0- 00°0- 00°0 00°0 00°0- 00°0b6°T 9°T 00°0 _00°0 96°T o9°T

4 wouy e6uey) 4eTTON

eT eTaeL

‘

SjuewAeg pue SPdTEceY AVEZTTTW SEPMTIXa /t

Lt’0 00°0 t‘o 00°0- 00°0 00°0- 20°T

00°0 20°T

zb 9861 Tb 9861 > Se6T Sb SS6T 2b Ss6T

*ttd

oesues, SeOTAUBS ‘S°N

60°0 00°0 60°0 00°0- 00°0 00°0- 5s°0 00°0 oS°0

Th Se6T

qwoouy 2eN

sUueWAeY S}dTel8y [BPO], sewoouyT STAIES

qwoouy 2°N S},USwWAeY S}dTes9y

7T s@woouT edTAVES 49430 @woouy ON spuewAed Sd T9084

7T sewoouy 2USWASeAUT 4O4ID @woouy ON s,UCWABY

Sj, dTel8y sQWOOUT PUEWZSOAUT 290ITC

- 23b -

ss°Tts‘T 20°0- 65° 0- 6S°0 00°0 50°0- 20°0 20° 0- 06°0- 06°0 00°0-

986T

09°0- 09°0 00°0- T2°0- T2°0 00°0 T0°0- 00°0 00°0- 8 °0es°0 00°0-

Se6T

T9°Tes‘ T ¢0°0bL°0- 52°00 00°0 90°0- ¢0°0 £0°0- T3°0-

Te°0 00°0-

bb 9861 Sb 986T 2b 986T TD 986T

sjpuewAeg pue SzdtTesey Ase TTIW sepn[oxq /T

96°T- so't- 2t°t- b6°0- f2°0- O0S°0- S2'0- ewoouy 3°N

b6°T To°T Itt £6°0 22'0 65°0 sz°0 szpuewAed

zo‘'0- 20°0- 10°0- To’0- 00°0- 00°0- 00°0- S}ATID0y TSPO_ sQwoouT eITAIIS

¢9°0- SS°0- sb°0- sc‘0- GSz‘'o- LT°0- 80°0- qwoouy 3°N

£9°0 ss°0 S>°0 s¢‘0 S2°0 Lt‘0 $0°0 syuewAed

00°0 00°0 00°0 00°0 00°0 00°0 00°0 S}d T9994 7U s@wOdUT BOTAJAS 40420

b0°0- £0°0- 20°0- 10‘0- 1T0°O0- 00°0- 00°0- qwoouy 3°N

20°0 10°0 10°0 To°0 00°0 00°0 00°0 szuewAed

zo‘0- 20°0- T0°0- 10°0- 00°0- 00°0- 00°0- SpATIIeH ZT s@MoOoUT PUEWZSEAUT 428470

6z°T- Sss'0- s9°0- gs°0- b°0- ze°0- LTO- qwoouy }eN

6z°T s32°0 s9°0 8s°0 L>°0 2<°0 LT°0 szuswAed

00°'0- o00°0- 00°0- 00°0- 00°0- 00°0- 00°0- SpAdTesey

sQWOsUT PUEWZSEAUT 7994TO

bb G61 sb S6T z> S86T Tb S86T

“yy ‘AND “S'N Jos OFEY UPMOID 4OUBTH % T ceryATog Toayue woay eGue4D JeTION “T¥d

pUNO.Sy ZYUS4IND UO suoT}pOesuesy SBDTAUSS *S*N

yt eTqeL

take into account, indirect effects through the impact of GNP on the merchandise trade component of the U.S. current account, interest rates, and prices. E) Foreign Prices (Table 15

A 1 percentage point (annual rate) increase in the foreign inflation rate raises U.S. net service receipts by $0.1 billion in the first year and $0.4 billion in the second relative to the base case solution. Foreign prices have a direct impact on other services; the higher the relative price of foreign services, the larger the demard for U.S. services by both foreign and U.S. residents. The nominal value of imports does not fall by the full reduction in volume because forei.gn prices have increased. Direct investment receipts increase in nominal terms when foreign prices increase because nominal profits earned abroad are larger. Other investment income is not directly affected. E) U.S, Prices (Table 16

A 1 percentage point (annual rate) increase in the U.S. inflation rate leaves U.S. net service receipts approximately unchanged in both years relative to the base case solution. Several effects appear to cancel each other out. U.S. prices have a direct impact on other services; the higher the relative price of U.S. services, the higher the quantity imported and the lower the quantity exported. On the other hand, the nominal value of U.S. exports rises if the price increase on exports outweighs the quantity reduction. Direct investment payments increase in nominal terms when U.S. prices increase, because nominal

profits are higher. Other investment income is not directly affected.

- 24a -

05°0 2T°0 22'0 90°0 29°0 8T'0

20°0 20°0- 020 90°0 T2°0 90°0 so°o- t0°0-

20°0 00°0 £0°0- To°0-

5>°0 Sst’o 00°0 00°0 b>°0 Sst‘0 986T SS6T

95°0 62°0 S2°0 so°o 92°0 T<°0 40°0- £0°0 50°0- 8>°0 00°0 8>°0

bb 986T

S$>°0 sf °0 o2°0 020 69°0 3s°0

£0°0 00°0-

12°0 sT‘0 52°0 g1°0 90°0- s0°0- 20°0 z0°0 $0°0- £0°0g>°0 2>°0 00°0 00°0 95°0 25°0

£b 9861 2b 9861 I> 986T

Ts°0 8T°0 ST°0 It°0 95°0 os ‘0

20°0- 20°0-

bT°0 Or’O 2T°0 80°0 £0°0- 20°0- T0°0 T0°0 20°0- T0°0- 9¢°0 £2°0 00°0 00°0 9¢°0 £2°0

9T°0 40°0 £2°0 20°0- 40°0 s0°0 T0"0- T0°0 To*0- 6T°0

00°0 61°0

SpuewAeg pue SzPdTe0eY AYEXT [TH SEPN[OXZ /T

60°0 50°0 sT°0 To°*0- 50°0 20°0 T0°0- 00°0 00°0-

TT*0

s0°0 10°0 40°0 To°0- T0°0 T0°0 00°0- 00°0 00°0- 90°0

00°0 90°0

yb S86I Sb S86T 26 S8é6T Th S86T

"y°y ‘seoTUg UBTeu0y 4Oy 028y YZMOUQ YOUGTH X% T

UOTINTOS [OuzZUOD Wous SGBUueYD vETIOG ‘Tg ‘sueTTog “S°N 4° SUOTTTTE)

CavVs 2uNOooY 1Ue4uN} UO SUOTPOeSURBLI SBDTAUES “S°N

ST 9TqeL

@WOOUT EN

S}uewAeY S}d T9094 [SPO] s@wooUT VOTAUES

@woocur PEN s},UuewAeY S}dTel8Yy ZT s@WOOUT BOTAUES 429470 @woouy ON spUuowAeg SPATeDeYy ZU s@wOoUT ZUEWZSEAUT 4040 @woour ON s,uawAeg

Sj}dteley SQWOOUT PUSWYSOAUT POUT

~ 24b -

40°0- £2°0 99°0 £0°0- 59°0 T9°0 60°90 50°0- $0°0 £T°0- £Tt°0o 00°0-

986T

10°0 92°0 92°0 z0°0 €2°0 S2°0 20°0 10°0to’0 b0‘0b0'0 00'0-

S86T

40°0- 03‘0 £2°0 2Tt’0o S0°0- 40°0 2T*0-

2T°0 00°0-

S}UeWARY pUe S}dTESeY AUEPTTIW SOEPNTOXZ /T

¢T°0- 90°0- £0°0- zo°0- 00°0 20°0 20°0 £8°0 39°0 5S°0 £5°0 T<°0 02°0 60°0 02°0 T9°0 2s°0 Tb°0 T£*0 22°0 Tt’°o 50°0- zco°o- 00°0 To°0O £0°0 £0°0 20°0 39°0 6$°0 85°0 g2°0 2°0 8T°0 60°0 s9°0 4s°0 85°0 6£°0 o<*0 T2°0 TI*0 60°0 30°0 90°0 50°0 20°0 To°0 00°0 50°0- £0°0- 20°0- zo°0- T0°0- 00°0- 00°0s0°0 50°0 £0°0 20°0 T0°0 T0°0 00°0 6T'0- 2T°0- 60°0- 20‘0- S0°'0- 20°0- 00°0- 61°0 2t°0 80°0 40°0 s0°0 20°0 00°0 00°0- 00°0- 00°0- 00°0- o00°0- 00°0- 00°0-

bb 9861 Sb 9861. 2b 986T Tb 986T ob S86T £> S86T 2b S86T To Segé6T

‘yy ‘SeoTUd “S'N JOJ ©2BY YZMOUQ JEYSTH % T

UOT}NTOS [04zZUCD Wo4y EGUeYD YETTOT “TT Cavvs ‘s4eTTOd “S'N 3° SUOTTTTE)

puNOD.y PUe4IND UO SUOTPOeSUBI, SBDTAUOS “S*f

9T ST9BL

@wooul ON

spuewAeg Spdtesey TePO, sowooUT BdTAUES

@woouy }°N sz,uswAeg S}dre.ay

ZT s@woouT BOTAUaS 4042G ewoouy }°N spuswAed | Sjdteoey

ZT s@mooUT ZUCWZSOAUT 48420 @woouy ON s}uewAed

S}dTe90y 1@WODUT PUEWYSEAUT 7991TG

- 25 -

G)_ W.S. Net Indebtedness (Table 17)

A $100 billion increase per year in the U.S. net foreign indebtedness divided evenly between increases in portfolio liabilities to foreigners and reductions in U.S. portfolio claims on foreigners reduces U.S. net other investment income by $3.8 billion in the first year and $10.0 billion in the second. (In addition, if the increase in net | indebtedness were the result of a deterioration of the current account and higher imports, then payments to foreigners for other services such as freight and insurance would also be higher.) These simulations assume that the marginal assets acquired by foreigners earn the same rate of return as the average. Historically the average rate of return earned on U.S. claims has been higher than the rate paid on liabilities. If massive foreign inflows to finance large U.S. current account deficits were concentrated in higher return assets, net other investment income woulcl be even lower than indicated by these simulations. The simulation results indicate that, holding other things constant, it would take about four years of debt accumulation of this magnitude ($100 billion per year) to turn the U.S. services balance from a surplus of $25 billion in 1986 to a deficit. The longer the United States runs large current account deficits, the larger the swing in the merchandise trade balance that woulc. be necessary if a balanced U.S. current account position were to be

achieved in the long run.

- 25a -

$6°6- S2° TZ°s- 00°0 00°0 00°0 S6°6- S2* TL°S- 00°0- 00°0 00°0-

926T

T8°¢- 09°T 02°2- 00°0 00°0 00°0 TS*s- 09'T 02°2- 00°0- 00°0 00°0=

SS6T

92°TT- It’s s9°9- 00°0 00°0 60°0 9L°TT- It’s s9°9- 00°0- 00°0 00°0-

bb 986T

SpuowAeYg pue SYdTESeY AVERT ITH SEPNTOX |

be°OT- OS°6- T2°s- €S°9- bl°b- G6°2- 00°T-

62° To*s Lo°s g2°2 00°2 Sso'°T 25°0 G6°S- 65°S- £l°d- Os's- yl'2- OL°T- 8S°0-

00°0 00°0 00°0 00°0 00°0 00°0 00°0

bs°OT- OS°6- IT2°8- $S°9- bl°*b- S6°2- 00°T- 6£°> 10° Lb°¢ £L°2 00°2 S2°T 25°0 G6°S- 65°S- £2L°d- os's- v2°2- OL°T- 89°0- 00°0- 00°0- 00°0- 00°0- 00°0- 00°0- 00°0- 00°0 00°0 00°0 00°0 00°0 00°0 00°0 00°0- 00°0- 00°0- 00°0- 00°0- 00°0- 00°0-

sb 9861 2b 986T Tb 986T ob S86T Sh S86T 2b S86T Tb S86T

GuTMoss0g UBTeu04 “S'f UT eSees9UT UOTTIT OOTS

UOT}INTOS T[O4zZU0D Wouy eGUeYD 4ETTOG ‘“Tt¢ CavVS ‘S4BTTOg “S° 4° SUOTTT TE)

pUNOSZY YUSUIND UO SUOTPIESUB4] S@DTAUES “Sf

Lt etqey

@woouy ON

spUueUAeY S}dTe0ey TePOL sowoouy SdTAse

@wooUy 20K spuSWARY S}dTeley

/U S@WOOUT @STAUES YEU @woour 3eN suewAeY _- SPdTeDeYy

7 s@wooUT YUSWYSEAUT 40470 @woouy ON spueWAeY

S}dTejey S@WOOUT PUSWPZSOAUT P0ITYG

- 26 -

REFERENCES

Bond, M., "A Model of International Investment Income Flows," IMF Staff Papers 24 (July 1977) pp. 344-379.

Bond, M., "The World Trade Model: Invisibles," IMF Staff Papers papers 26 (June 1979 pp. 257-333.

Kenen, Peter, A Model of the U.S. Balance of Payments, D.C. Heath and Company, Lexington, Mass., 1978.

Kwack, Sung Y., The U.S. Sector of the Multi-Country Model and its Multipliers, Federal Reserve Board, International Finance

Discussion Papers, No. 146, July 1979.

Proctor, Alan, A Model of the Services Account of the U.S, Balance of Payments, Unpublished paper, Federal Reserve Bank of New York, 1982.

Sawyer, J.A., "The Invisible Components of the Current Account of the Balance of International Payments" in R.J. Ball (ed.), The

International Linkage of National Economic Models, North- Holland, New York, 1973.

Stekler, Lois E., U.S, Direct Investment Receipts and Payments: Models

and Projections, Federal Reserve Board, International Financial Discussion Papers, No. 140, May 1979.

IFDP NUMBER

319

317

316

315

314

313 312

311

310

309

308

307

306

- 27 - International Finance Discussion Papers

TITLES 1987

Modeling Investment Income and Other Services in the U.S. International Transactions Accounts

Improving the Forecast Accuracy of Provisional Data: An Application of the Kalman Filter to Retail Sales Estimates

Monte Carlo Methodology and the Finite Sample Properties of Statistics for Testing Nested and Non-Nested Hypotheses

The U.S. External Deficit: Its Causes and Persistence

Debt Conversions: Economic Issues for Heavily Indebted Developing Countries

Exchange Rate Regimes and Macroeconomic Stabilization in a Developing Country

Monetary Policy in Taiwan, China

The Pricing of Forward Exchange Rates Realignment of the Yen-Dollar Exchange Rate: Aspects of the Adjustment Process in Japan

The Effect of Multilateral Trade Clearinghouses on the Demand for

International Reserves

Protection and Retaliation: the Rules of the Game

Changing

International Duopoly with Tariffs

A Simple Simulation Model of International Bank Lending

A Reassessment of Measures of the Dollar's Effective Exchange Value

AUTHOR (s)

William Helkie Lois Stekler

B. Dianne Pauls

Neil R. Ericsson

Peter Hocper Catherine L. Mann

Lewis S. Alexander

David H. Howard

Robert F. Emery Ross Levine Bonnie E. Loopesko

Robert E. Johnson

Ellen E. Meade

Catherine L. Mann Eric O'N. Fisher Charles A. Wilson

Henry S. Terrell Robert S. Dohner

B. Dianne Pauls William %. Helkie

' Please address requests for copies to International Finance Discussion Papers, Division of International Finance, Stop 24, Board of Governors of the

Federal Reserve System, Washington, D.C.

20551.

IFDP NUMBER

305

304

303

302 301

300

299

298

297

296

295

294

293

292

- 28 - International Finance Discussion Papers

TITLES

Macroeconomic Instability of the Less Developed Country Economy when Bank Credit is Rationed

The U.S. External Deficit in the 1980s: An Empirical Analysis

An Analogue Model of Phase-Averaging Procedures

A Model of Exchange Rate Pass-Through

The Out-of-Sample Forecasting Performance of Exchange Rate Models When Coefficients

are Allowed to Change

Financial Concentration and Development: An Empirical Analysis of the Venezuelan

Case

Deposit Insurance Assessments on Deposits at Foreign Branches of U.S. Banks

1986 The International Debt Situation

The Cost Competitiveness of the Europaper. Market

Germany and the European Disease

The United States International Asset and Liability Position: A Comparison of Flow of Funds and Commerce Department

An International Arbitrage Pricing Model with PPP Deviations

The Structure and Properties of the FRB Multicountry Model

Short-term and Long-term Expectations of the Yen/Dollar Exchange Rate: Evidence from Survey Data

AUTHOR(s

David F. Spigelman

William L. Helkie Peter Hooper

Julia Campos

Neil R. Ericsson David F. Hendry Eric O'N. Fisher Garry J. Schinasi

P.A.V.B. Swamy

Jaime Marquez Janice Shack-Marquez

Jeffrey C. Marquardt

Edwin M. Truman Rodney H. Mills

John Davis

Patrick Minford Guido E. van der Ven John E. Wilson

Ross Levine

Hali J. Edison Jaime R. Marquez Ralph W. Tryon

Jeffrey A. Frankel Kenneth A. Froot