Investor Evaluation of Foreign and Domestic Risk

cof. es ENTERNATIONAL FINANCE DISCUSSION PAPERS

”

\ a. ‘ , a Er tes ta a

a ‘ ‘ ) ? ® 4 e Ww Ww

Sa bocnts, Wea

Soe ba,

| 1° +» -\NVESTOR EVALUATION OF FOREIGN AND DOMESTIC RISK

ol, - ee 3 can Alan K. Severn

4 . ‘ss . see

| Discussion Paper No. 34, November 19, 1973.

| a : . — | [

nth

thee ee ‘

. Division of International Finance

” Board of Governors of the Federal Reserve System . fork

j ; :

. The analysis and conclusions of this paper represent the views of the author and should not be interpreted as reflecting

the views of the Board of Governors of the Federal Reserve

System or its staff, Discussion papers in many cases are . circulated in preliminary form to stimulate discussion and

comment and are not to be cited or quoted without the permis~-

sion of the author,

yi? Flyin io yk wee . - : ~ ? : z : . . 5 . "oe | i . - vi, . « e ° e . . ee 7 - Pars ‘ e : _— . - a or ee ee ar ee — . . + een eee = eee : ee TTT er coe me ~ ames te ee ee tee =

ee : INVESTOR EVALUATION OF FOREIGN AND DOMESTIC RISK

“. Alan K. Severn*

‘

In recent years, foreign direct investment by U.S. manufacturing

ye tL DN Ae

firms has grown rapidly, oriented largely toward markets outside the

u.s.2 Levels of economic activity in these markets are less than

' va tf ry

tte

~>perfectly correlated with levels in the U.S. 2 From 1963 on, the U. S.

RRS

: Interest Equalization Tax effectively prevented U.S. portfolio investors

from achieving international portfolio diversification directly, Given

' this constraint, international diversification by firms was of benefit

. 3 to their owners, by.reducing portfolio risk for a given rate of return.

- - In this paper I present an empirical test of investor response

to the respective risks associated with the foreign and domestic income

of U.S.~based direct investors, I then estimate the reduction in

¥.

'. she reholder risk associated with direct investment, oe

‘The theory of capital asset pricing states that the equi ibrium

tyes

return on the ith asset is a function of its systematic visk;:

Ry = rk + R, ~ <*)B;; . | wo of : (1)

eee

Habs Sut Vt Patra

*The author is an Economist, Board of Governors of the Federal ° ‘Reserve System. He wishes to thank Martin M, Laurence for helpful ‘Comments and Cora Flaifel for programming assistance, He retains sole responsibility for opinions expressed,

1

Met t

See R.D, Belli, "Sales of Foreign Affiliates of U.S. Firms, 1961- 65, 1967, and 1968,' ' Survey of Current Business, October, 1970, p. 20,

See, for example, Fred B. Ruckdeschel, "The Determinants of a

. Direct Investment Outflow with Emphasis on the Supply of Funds," Federal vs Reserve Board Staff Study No. 78, 1973, p. 119, .

rr in iS ee

ie

Ry contrast, conglomerate mergers did jot reduce portfolio risk, _ Since corporate diversification merely replaced investor diversification. "s ' See H. Levy and M, Sarnat, "Diversification, Portfolio Analysis and =

the Uneasy Case for “R02 erate Mergers," Journal of Pinence, 25

{ ! | t ' i t ' ’ } ' '

Stag ie

'. Where g* is the excess of firm i's growth rate over that which would

and r* is the riskless rate of return,4 Application of this theory

“more than Rm. Therefore, the equilibrium capitalization rate applied

where b is the proportion of earnings retained,°

- deviations’ around the growth rate of earnings per share during the

“Capital Asset Prices: A Theory of Market Equilibrium Under Conditions

. and Profitability," Journal of Financial and Quantitative Analysis _-

-§ (March, 1974), forthcoming,

PO ee renames © enmnnegnes aeennm een eemmeneens a arene eermer =. te emma ae me me a a a ek

.if earnings were reinvested at the equilibrium rate of return.” Hence,

a

“rgd,

at a

where 8 = p;,0;/o,5 Ri is the equilibrium rate of return in the market,

to common stocks requires recognition of market imperfections which

cause the earnings of some firms to grow faster than would be expected

~~,

be expected if it had no opportunity for investment in assets yielding

to a firm's current earnings is:

-*

(B/P); = rt + Ry - r*)B, - (@ - BRy), 3)

1 e , . ot . . . . a a oy . ey os . a Sau ao 7 . * 4: F a 1 . ee / a a PR EK OR TEE RG AALS GE BO lene gids lee ig odin

Systematic risk of the firm's earnings, By» is calculated from

og ae geal, ae

Y

‘» IN t

NX - . . . oe .. Calculation period, in relation to the earnings per share of the a Standard and Poor Composite Index. Earnings are used directly to “

4 see J. Lintner, "The Valuation of Risk Assets and the Selection ve of Risky Assets in Stock Portfolios and Capital Budgets," Review of _ Economics and Statistics 47 (February, 1965), 13-37, and W.F. Sharpe, - *

of Risk," Journal of Finance 19 (September, 1964), 425-442,

>such opportunities for above-normal growth are especially prevalent among the direct investors under consideration here, because accumulated research and development establishes barriers to entry to the industries in which direct investors are located. See A.K: aioe Severn and M.M, Laurence, "Direct Investment, Research Intensity, me:

_ The observed growth rate, g, is ‘adjusted by retentions times Re rather than R;, since the risk differential of the i firm is already accounted for by Bj» which incorporates fluctuations around the above-normal rate of return and hence in g.

5 eee

'.ealculate B;, rather than returns to the investor, since we are concerned with the market's evaluation of earnings, Earnings per share,

in each year, are normalized by their geometric mean for the calculation period,’ 7 co, . oe Systematic risk of foreign (domestic) earnings is obtained by

- multiplying deviations of earnings, as described above, by the trend __wwalue of the ratio of foreign (domestic) to total earnings, Hence, the sum of foreign and domestic risk is constrained to equal the systematic risk of the firm as a whole, .

-Earnings-price ratio is reported earnings per share, averaged

-

"over two years, divided by share price at the end of the first of the ‘two years, Thus, investors are assumed to forecast accurately the —

firm's earnings in the second year,® oo oo a . 7 Past growth of earnings cannot be taken-as a proxy for expected

future growth of earnings.” Past growth of book valye cannot be used | ree - .

"The ratio of earnings to book value has been used as an alternative to this normalization and calculation of trend. See R.H. Litzenberger and A.P. Budd, "Corporate Investment Criteria and the Valuation of Risk Assets," Journal of Financial and Quantitative Analvsis 4 (Dec., 1970), 395-419, But their procedure builds in an errors-in-variables

bias, because of divergences between earnings power and book value. While the more complicated procedure used in this -paper was necessary - because data on book value of foreign (domestic) Operations are generally not available, it gives far Superior results for the firm as . /a@ whole than does risk calculated from the ratio of earnings to book value,

Spreliminary tests indicated that use of a Single year's earnings Save poorer results than the average of earnings of the cross-sectional "year and the following year, The dependent variable used in this study implies perfect forecasting, by investors, of earnings of the following year. Use of a given year's earnings, divided by price at the end of the year implies zero forecasting by investors, All that is asserted . here is that perfect forecasting is closer to reality than zero fore- “+ - easting ability. tO me

osisegied a see J. Lintner and R. Glauber, "Higeledy Pigeledy Growth in

‘America;" in J, Lorie and R, Brealey, eds., Modern Developments in - ~dnmvestment Management (New York: Praeger, 1972), 645-662.

° : . . H bd bd y , , , . Sct eee -

since, in the-.absence of new equity financing, it is determined only

tm Te

by retained earnings. Given retained earnings and therefore book

“ walue, the firm can either increase assets or reduce liabilities,

-

Growth is equal to the return on assets times the retention rate. The firm with unusually profitable investment opportunities should have a high rate of investment, rather than using transitory income to redude liabilities, Therefore I use past growth of assets as a ‘proxy for expected future growth of earnings, 7 - The market return, Ry is calculated as the sum of the dividend

-yield and the growth rate of dividends (from 1956 through the last

year of the calculation period). The riskless rate was taken as the

_ market yield on 9-to-12 month Treasury bills, since § is based on

ye . ee 10 annual, rather than quarterly, fluctuations in earnings. ; 11

N

Data were available for two cross-sectional years, 1965 and 1966,

“Three firms were eliminated from the original sample of 63 because

of negative earnings in at least one year. Preliminary tests showed

that a uniform computation period of seven years was unworkable,

7

because the 1960-61 recession caused "cyclical" and "growth" firms .

to have similar values of B. Therefore. a. computation period of eight

; years (1959-66) was used for the 1966 ‘cross section, and seven years

10Risk which is measured from annual data is not the same risk as that which is measured fron shorter periods, See H.E. Phillips and J.P. Seagle, "Data: A Blessing or a Curse in Portfolio Selection,"

@roceedines of the Fifth Annual Meeting of the American Institute

for Decision Sciences,

h . line data were originally developed for the study reported 4 in A. K. Severn, "Investment and Financial Behavior of American Direct Investors in Manufecturing,' Universities - National LDureau Conference

"on International Mobility and Movement of Capital (New York: Columbia a tty ang «ovement of Capital

University Press, 1972), 367-396,

. a a . ‘ ‘

: | | | |

,

ay Bhs...

fe whe vies

’ “as 4g

re i & eet Orin

voy ot Tie

web Lae eo

" . $92" 612°

ete”

"eve"

z8z°

-(€88°z) 861°=

* (642°)

£9¢°=

(996°z) Z61°= (926°Z) : T6T°- (SS9°4) 9LZ°=

TO vee meee «a,

Pe EE GA A TRDRRRBE 2 IES EM BOONE BES HG AE tte I Se ty at attain, (84T°y)-. — (eeere) (Sa T)__ . ere? S61°= €60° 80° (S61°4) (zz$°¢) 6e° 61° = - 160° (26€°9) - , 092° _0L27= > "8 . (ove"z).* (goe*t) ~~ (tT0°T) ' egze 961"= LL0° 10°

— (€06°T) £L0°

'

., CS8L°%) . (TLE°T) SOT’ 980° (€56°Z) 10)

Low e-mpenpsenmenpuameeeend

dFIsowog US}fosa0q

UBS DFAQOWOSS

wat J

WIAorD puw “ysty ‘orey uoz3ezzyeazdeg 37 STQPL

(6 yTO)

" (6S%°t) Sloat L€0° wf

soe

900° * ,

(€1L°0z) $97 °9 (€08°0Oz) © *947°9 (000° 0¢€) 868°9

.

(O8S°€T) L1z°9

(T€S°6T) 1z8°9

(97° ST) , 262°9 (8S2°T) (296°2Z) 950° O£0°Z (antea yoog) d Juez euog

'

%

"+" pajoog

PeOrErrrerery Seitiit kt 2 ds

ee ee

So a Oe Ae ne ee ee nee oe me eee 8

ee

+ meme we meee en.

ES eee = gee -

“used in other studies, !> Second, investors are likely to perceive be less-than-unitary reaction to income risk, even if there is unitary

Suggesting that growth itself may be overrated when risk is not taken

14 a » _. into account. of aan

" Tevel of 2.68.

_.for the 1965 cfoss section, ce _ \

The results (Table 1) show that the estimated effect of foreign

and domestic risk is similar. Since the results were stable from one

year to the other, the two cross sections were pooled, !? The cocfficients of foreign and domestic risk show no evidence that investors discounted foreign risk to a _greater degree than domestic risk,

despite the unrealized potential for expropriation, currency restrictions, or devaluations greater than warranted by relative prices.

The fact that the estimated coefficients are significantly below

. the unitary effect predicted by the theory comes about for two reasons,

oo First, the estimated risk premium, (Ry - r*), is larger than that - 7

extreme values of incouc as transitory; if SG, reiurns to investors will have smaller variation than will earnings, Hence, there would

reaction to investor risk, :

ene pg eee eee hee ne eee

The coefficient of growth falls as the estimate of risk improves,

s

12the Chow test gave a value of F of .196, relative to the 5% Hence, there is no evidence that pooling is not vesteinates

13, H.. Litzenberger and A.P. Budd, "Corporate Investment Criteria,' used the market earnings-price ratio to represent R,; hence, they failed to allow for inflation and for above-normal- growth by some

_ firms in the market average, and thereby imparted an upward bias to _their estimate of the coefficient of risk,

14, both samples, growth and risk were inversely correlated,

er od ® .

4. “8

Tee

of risk which it experienced domestically,

-6-

Therefore, 7 assume that investors discount foreign and domestic risk |

equally, and. estimate the reduction in total risk resulting from

. international diversification « on the part of U. S. firms. The major

drawback to such an estimation is that the proportion of foreign to total earnings, W, must be taken as given. By contrast, maximization Of- shareholder wealth should have led firms to require a | lower rate

of return on foreign investment than on domestic investment (at least

15 after the IET was imposed). _ If so, any estimate of risk reduction

"represents an upper limit to the improvement in shareholder wealth,

Within the limits of taking W as given, risk reduction can be

sieasured in terms of how much larger the firm! s total risk would have

been if foreign income had involved the same risk as domestic income, Based on sample means, total risk was lower by .0088% in 1959-65

and .0113% in 1966, for each 1% of the firm's total earnings arising

c - | from foreign, rather than domestic, investment, Given.the risk

Coefficients of Table 1 and the risk premia on which they are based,

these estimates imply that the. capitalization rate @/P?) was lower by 038% at the beginning of 1965 and -0031% at the beatoning of 1966, for each 1% of total income , which was earned abroad.

The sample mean of W was 277%. Hence, a typical firm in my sample : :

_Was accorded a capitalization rate about ten basis points lower than

what it would have had if‘all income had been subject to the > degree

—

‘157, addition, there is indeterminacy (ekonomic as well as

accounting) in the allocation of earnings to foreign vs. domestic “operations, because of joint costs and transfer pricing. Furthermore,

exports make possibie some amount of foreign income even in the

absence of foreign investment. \

Ry

- .

seme sem 0 000 ODEO SHO .

i a

ry ye Spek

Bea WY Pana eh

ye x

era,

a

a

-To this point, I have discussed the risk-reducing aspect of foreign investteit ‘fa terms of the firm's maximizing the net worth of its shareholders, - ‘Undoubtedly, the relatively high rate ‘of

growth of foreign investment of the 1960's increased share prices

of direct investors, as their capitalization rates fell relative to those

of domestically-oriented firms, At the same time, part of the in-

crease in direct investment was probably a response by firms to the

‘low systematic risk of foreign income (relative to fluctuations of

| earnings in the U, S. ). After the Interest Equalization Tax was

. Imposed. in "1963, investment in shares of direct investors was the

only means by which U. Ss. portfolio investors could diversify away

part of the risk of U.S. earnings. Hence, the relatively low

capitalization rate of foreign earnings may have stimulated direct

tnvestient, thereby providing a partial offset to the gross balance

of payments improvement brought about by the IET. °

For this and other reasons, U.S, balance-of-payments deficits

subsequent to the imposition of the IET brought about the mandatory

Foreign Direct Investment Program (FDIP) at the beginning of 1968,

After 1968, the FDIP had little effect on the amount of direct in-

vestment by U. Ss. " manfacturing firms,” Rather, -it forced U.S, firms

- to obtain part of their funds for direct investment abroad, at rates

AY

generally above those in the v.s2? Therefore it offset all or part

nemo .

a

6 - . ; See G, Stevens, 'Capital Mobility and the International Firm," in F Machlup, et al, eds, International Mobility of Capital, p. 342.

a ee A. Severn, "The Financing of the Multi-? ‘ational Firm: Comment, "Kyklos 26 (1973), forthcoming.

Apes

SN AEE PRO Os Bah abe

‘

heat PER tas!

Saray

arr Ta) wed RD ae Des

Way yee Sea al

4 1»

: - wae - : i

oe a i . . : : . of the incentive to invest abroad which was provided when the IET effective- ¢ ly prevented * international diversification by U.S. investors, | Reduction of risk by means of direct investment should also affect oo “ ‘allocation of investment funds within the U.S. Horizontal direct in-

H . ° e

vestment involves the exploitation of some resource, within the firn,

e ay

which is not readily sold to a foreign-owned firnl® The best example of such a resource is the results of research and development activity, Therefore the lower capitalization rate accorded to direct investors should foster investment in research and development, relative to ; i other forms of investment which are intended for domestic use only. ot The rising ratio of company-funded research and development expenditures to sales, during the sample period considered here, supports | a

- this contention!® . Lo : i ee

"- Wiewed in this light, risk reduction by means of direct in- Pree vestment should remain relevant even after the planned termination - o£ the IET in 1974 brings about a greater degree of perfection in

international capital markets. To the extent that implementation

CE ee rf

of the results of research activity continues tc motivate direct investment, risk reduction within the multinational firm will continue to benefit welfare, by reducing the risk associated with © research activity, At the same time, any increase in interna tional economic integration will limit the degree to which risk can be ! OF,

-. reduced by foreign direct investment, oO Le 7 i ane

: oe

- 18 | | : i 3

~ See R. E. Caves, "International Corporations: The Industrial nor “" . «. ° Economics of Foreign Investment," Economics NS 38 (1971), 1-27 =

won. that is, the high profitability of research and development is my supplemented by relatively low risk in its application, See A. Severn and M, Lawrence, "Direct Investment, Research Intensity, and Profitability," ;

° : e . ie ° : ' ; i ‘ me NN eS ye sy SN RN + 8 cee Sen. ee gy ae meee eee meee ee No oe .

LF TT he SRR se wera gees e