\'V0HK Y(, IPAP H", DOM tWmid tmauoeul phinee Intemnational Econoiics Dpeprtment The World Bank October 1993 WPS 120 Empirical Perspectives on Natmional Index Funds Ishac Diwan Vihang Errunza and Lemma W. S1"enbet U.S. investors could benefit from diversification that involves national index funds, particularly funds originating from coun- tries to whose local markets they have limited access. Country funds also improve pricing efficiency in local capital markets and help local finns mobilize local capital at lower COSt. MhsP4SyRsueahaki%d.Pap.udiu.uinawcflndofwa&inprrmadmawaeoopAchanaofidcas=#amAODi*a , amid i 0nh8 imd indwdopmmurm su.T ppq diwm by - Rh.uh Adviy S;:. :M; don na o-dhe £i.ucvO. ud.. hsw dud=W*guIdhum mdbWw=Wb*ay.7Ur-bi-V mpiim adeihea .eleah,owaL They .u~bs~Umb **- Widd D.e d. .mis,dmi .mamii. of y of ia 1 amweui 2J, $. ~ ~ ~ ~ ~ w ,,. :;... Ddl ntwhbo F«7o2 WPS 1206 This paper -a product of the Debt and Litemadonal Finance Division, Intemational Econemics Departmnent-is a companion paper to the heoretical analysis of country funds by the sane authors, 'The Pricing of Country Funds and Their Role in Capital Mobilization for Emerging Economies," Policy Research Working Paper 1058. Copies of this paper are available frec fom the Wodd Bank, 1818 H Street NW, Washngt DC 20433. Please contact Azeb Yldeiu, room H7-03, extension 36067 (October 1993, 49 pages). Closed-end nadonal index uds (NIFs or compare rslts across emergin and industrial "country fmnds") invest primarily in the stocks of markets and, where appropriate, over different the originating countries, such as Brazil, India, subperiods. and the Republic of Korea. They are typically traded in the organized exchanges of industrial Their evidence suggests that U.S. investors countries, such as the Urited States and the could benefit significantLv in diversification that United Kingdom. Although NIFs have not raised involves NIFs, particularly funds originating large amounts of extemal funds, recently they from countries to whose local markets they have have expanded rapidly. limited access. In a companion paper ('The Pricing of Diwan, Errunza, and Senbet investigate the Country Funds and Their Role in Capital Mobili- pricing of NEFs, testing their principal theoretical zation for Emerging Economies," WPS 1058), predictions about the relative significance of the Diwan, Errunza, and Senbet develop a theoreti- home market, host market, and global closed-end cal model to compare the pricing of country fund factors. They analyze initial (public- funds in the reference malkets (say, the United offering literature) and after-market returns, and States) with the pricing of the underlying compo- explain the behavior of fund premia/discounts. nent assets (or net asset valuation) in the origi- The evidence shows that variables that pioxy the nating securities market under various assump- degree of access and substitution effects show up tions about market structure. as significant determinants of country fund premia/discounts. In this paper, they empirically investigate the hypotheses that emerge from the model. They The empirical study supports their theory first analyze country fund pricing and associated about the welfare implication for emerging premia, or discounts, and then explore the issue economies that originate country funds. The of diversification services provided by NIFs model suggests that country funds can improve from emerging markets. The emphasis on pricing efficiency in local capital markets and emerging markets is important as many markets promote local capital mobilization by firms at are otherwise closed to foreign investors. They more favorable tenns (lower costs of capital). The Policy ResearchWorking PaperSeries disseminates the findings of work under way in theBank Anobjective ofthe series is to get these findings out quickly, even if presentations are less than fully polished. The fndings, interpretations, and conclusions in these papers do not necessarily re?resent official Bank policy. Produced by the Policy Research Dissemination Center Empirical Perspectives on National Index Funds by Ishac Diwan, Vihang Errunza, and Lemma W. Senbet* *The authors are, respectively, from the World Bank, McGill University, and University of Maryland. Note: This is a companion paper to "The Pricing of Country Funds and Their Role in Capital Mobilization for Emerging Economies, Policy Research Working Paper 1058, World Bank, 1992. The authors thank Ked Hogan, Art Moreau, Vikram Nehru, and Morty Yalovsky for comments. Financial support was received by the au.hors from the World Bank, Faculty of Management at McGill Univer- sity, and the William E. Mayer endowment at Maryland. We thank Joseph Salva tore, Santiago Galindez, and Philip O'Connor for able research assistance. TABLE OF CONTENTS Page SUMMARY ....... I NATIONAL INDEX FUNDS: EMPIRICAL PERSPECTIVES. . 3 1. THE DATA AND SAMPLE ................................ 4 11. GAINS FROM DIVERSIFCATION INTO EMERGING MARKETS ... 5 I. 1 Pairwise Correlations . . . .............. 7 II.2 Components of Risk ........... 9 11.3 Mean-Variance Efficient Frontiers ........................ I l 11.3.1 Idealized Diversification Gains .................... I 1 11.3.2 U.S. Based Diversification ....................... 12 ilI. PRICING OF NIFs ..................................... 13 III. I Imperfect Substitution .14 111.2 Methodology .14 111.3 Results ......................................... 16 IV. IPOs OF NIFs ..18 IV.I Initial Returns .19 IV.2 After Market Returns .19 IV.3 The Seasoning Effect .20 V. PREMIUMS/DISCOUNTS ON NIFs ..20 V.1. Determinants ol Premium/Discount .21 V.2. Variable Definilions .21 V.3. The Test Procedure ................................. 23 V.4. Test Results ..................... 25 VI. CONCLUDING REMARKS .................... 26 REFERENCES ............. 28 ANNEX TABLES ............. 29 l i -JRES ............................... I ................ 47 SUMMAR I Closed-end national index funds (NIFs or "country funds") primarily invest in the stocks of the issuing or originating countries, such as India, Korea, Brazil, and are typically traded in the organized exchanges of the developed countriLs, such as the US and the UK. AlthoLigh the external funds tapped through NIFs have been small to date, they have expanded at a rapid rate over the recent past. This raises the issues of their role in providing pricing efficiency in the originating stock markets of emerging economies and enhancing capital rnobilization by local firms of such economies. (Since country fiNnas themselves remain a very small fraction of the stock of external capital available to emerging economies, external capital mobilization is less important.) This paper is a companion paper to a theoretical analysis of countrv funds (Diwan, I., V. Errunza, and L. Senbet. "The Pricing of Country Funds and Their Role in Capital Mobilization for Emerging Economies," PRE Paper, 1058, The World Bank, December 1992.) The first paper focussed on the pricing of country funds in the reference markets (say the US) relative to the pricing of the component underlying assets (or net asset valuation) in the originating securities markets. Thai paper identified several variations of mnarket segmentation structure and arbitrage restrictions. This paper provides an empirical investigation of country funds with particular attention to their diversification benefits and their pricing behavior. Based on the theoretical analysis of the companion paper, this paper analyzes the empirics of country fund pricing and the associated premia or discounts. The paper explores the issue of diversification services provided by the NIFs from emerging markets. The emphasis on emerging markets is of particular importance, since many of them are otherwise closed to foreign investors. We compare the results across emerging and developed markets and report them for sub-pexriods, where appropriate. The evidence suggests a significant diversification benefit to US investors arising from NIFs, particularly those funds originating from countries with limited access tJ their local markets. The paper also investigates the pricing of NIFs. Specifically, it tests the principal theoretical predictions regarJing the relative significance of the home market, host market and the global closed-end fund factors. It analyzes the initial and after market returns, as in the initial ,.ublic offering (IPO) literature. It also attempts to explain the behavior of premia/discounts of these funds. The evidence for this question supports the predictions of the companion theoretical paper with regard to the pricing of country funds, relative to their net asset values. Variables that proxy the degree of access and substitution effects show up as significarnt determninants of country fund premia/discounts. The empirical support for the theoretical analysis is particularly useful, since the companion paper has advanced a number of welfare implications for emerging economies that originate country funds. The model suggests tha. vountry funds can enhance pricing efficiency in the local capital markets and promote local capital mobilization by firms at more favorable terms (lower costs of capital). Thus, we are encouraged that the policy implications drawn from the theore'ical analysis for the promotion and support of country funds have an empirical basis. 2 NATIONAL INDEX FUNDS: EMPIRICAL PERSPECTIVES Although the benefits of international diversification have been apparent for quite some time, they are impeded by barriers (explicit and implicit) to cross-border portfolio flows. In recent years, National Index Funds (NIFs) that specialize in assets of a given country (or region) and trade on a developed market, such as the NYSE, have been offered as an alternative investment vehicle to foreign markets.' Some of these funds special',,e in developed markets (DMs) that are largely aevoid of explicit barriers but may involve high transaction, information and other costs. Other funds invest in Emerging Markets (EMs), many of whom have imposed prohibitive barriers to foreign investments. Thus, many EM funds may provide the only opportunity to investors who wish to diversify in these closed markets.2 From the perspective of the home country, the NIFs may serve several purposes: (a) serving as a means of attracting external funds, (b) developing and liberalizing local capital markets, and (c) providing pricing efficiency through globalization, thereby enhancing local mobilization of investment capital. As argued in the companion paper, although the external funds tapped through NIFs have been small to date, their contribution to the local economy can be substantial through pricing efficiency and local capital mobilization. Despite the significance 'This is primarily a late 1980's phenorren'rn, although a few funds were available prior to 1985. 2lndirect investments through the multinationuis and a handfuli of American depository receipts are also available. See Errunza and Senbet (1981) for the valuation effects of indirect diversification through multinationals. 3 of the NIFs and their phenomenal growth since 1986 (see Figure 1), though, important empirical issues remain largely unexplored.3 This paper provides an einpiical investigation of country funds with particular atteniion to their diversification benefits and their pricing behavior. The theoretical analysis of the companion paper will provide a basis for the empirics on country fund pricing and the associated premia or discounts. We begin with a description of the data and sample. In Sectior II, we exrlore the issue of diversification services provided by the NIFs from emerging markets. The emphasis on EMs is of particular importance, since many of therl are otherwise closed to foreign inve,tors. The theoretical gains from diversification are con pared with the achievable NIF-based diversification. We compare the results across EMs and [)Ms and report them for subperiods, where appropriate. Section III reports on the pricing of NIFs. Specifically, it tests the principal theoretical predictions regarding the relative significance of the home market, host market and the global closed-end tund factors. Section IV analyzes the initial and after market return., as in the initial public offering (IPO) literature. Section V attempts to explain the behavior of pr!mia/discounts based on theoretical predictioris of the companion paper. Section VI provides concluding remarks. I - THE DATA AND SAMPLE The study covers all closed-end single country funds publicly traded in New York by the end of 1990. Table I lists the thirty-two funds in the sample, their offering dates and the 'Bailey and Lim (1989, 1990) analyze the diversification henefits and some issues related to their initial ofterings. 4 number cf weeks of trading.4 Eighteen funds are fronm EMs ard 14 from DMs. Table 2 provides the initial size, number of shares, price and value at the end of 1990. It should be noted that the initial size represents total capital raised by the fund tnrough one or more offerings up to the end of 1990. About 39% of the global market valuL :n 1990 corresponds to EM funds. Figure 1 documents the dramatic increase since 1981 througih 1990. The data base contains wcekly data for each fund since its inceotion. It comprises: Friday closing prices as reported in the NYSE records; net asset value (NAV) as obtained from fund managers; dividends and distributions of capital gains; local stock market inde; pro"ided by local exchanges; representative indices calculated by the International Finance Corporation for fifteen funds -eleven countries- (series start only in January i989); exchange rates (from IMF) between local currencies and U.S. dollar; and the Standard & Poors Composite Stock Price Index of 500 Stocks. Tables 3 to 6 report returns on the various funds. Table 7 reports returns on the coiresponding local stock market indices. II - GAINS FROM DIVERSIFICATION INTO EMERGING MARKETS The gains from international portfolio diversification documented in traditional analyse. do not take account of barriers to capital flows and the associated costs of accessing capital markets across national boundaries.j It is apparent that portfolio capital does not flow freely 4Twelve closed-end multi-country funds that traded in New York are not included in the study. We also do not consider non-diversified country funds, warrant funds or debt-conversion funds. 'See Levy and Sarnat (1970) and Errunza (1977), among others. 5 among many EMs and a number of developed markets.6 Thus, the evidence based on the assumption of free flow of capital and corresponding market indices may not reflect the true benefits of diversification. The purpose of this section is tn provide new evidence on the benefits of interrational diversification. The emphasis will be on EMs, many of whom are closed to foreign investors. We use the freely accessed National Index Funds that are also freely traded on the U.S. stock ex%zhanges together with various freely accessible developed market portfolios to document gains fro-m international diversification. We also use the EM and DM market indices (many of which are not freely accessible) to provide comparisons with past work and distinguish among theoretically desirable versus practically attainable diversification opportunities.7 Following the tradition, we begin by reporting the pairwise correlations, and then the systematic risks of the various NIFs with respect to the benchmark local and U.S. market indices. We proceed to develop various mean-variance efficient frontiers to document theoretical gains from diversification based on various market indices and the diversification potential based solely on the U.S. market that includes the National Index Funds. Finally, we provide the evidence of attainable diversification gains from the U.S. perspective based on all available assets, including all NIFs from developed and emerging markets as well as all freely accessible 6Stulz (1981) and Errunza and Losq (1989), among others, develop asset pricing models that suggest higher expected returns on securities from markets that cannot be accessed freely. 'Note that the debate as to whether the National Index Funds serve as a substitute for direct portfolio investments in the corresponding markets is moot in the case of markets otherwise closed to foreign investors. In the statistical sense, though, the behavior of NIFs will he compared to the underlying assets and market inc1'ces in this and the next section. 6 national market portfolios. Comparisons are made across DM and EM assets, and where appropriate, the results are reported for subperiods to establish time stability. 1.1 - Pairwise Correlati!nQ We begin with a correlation analysis for the period 1989-1990. The price returns arv based on market clearing prices of the NIFs in the abcve host (U.S.) market. The NAV returns are based on home market clearing p;;_es of the underlying securities that constitute the NIF. Note that we do not include dividends in computing price and NAV returns so as to make them comparable with returns on various market indices (which are not adjusted for dividends) used in the analysis of this paper. The results are presented in Table 8 and summarized below. The coefficients of correlation are: between returns on prices (ROP) and NAVs is CPN; between ROP and returns of local market index in U.S.$ (RLM) is CP,L;8 between ROP and returns on S&P500 is Cpsp; between NAV returns and RLM is CNL; and between NAV and S&P500 is CNSP. 8Table 8 also contains correlation coefficients computed between price and NAV returns with respect to IFC local stock market indexes. Waeneveg there is no IFC index, a N.C. (non computable) messages appears. 7 Table 1: Summary of Correlations of Pricjs and NA Vs Mean Std. Dev. Maxim Minim CP'N 0.4136 0.1780 0.7175 -0.039 CP,- I 0.4525 Li. 1560 0.7941 0.2167 ICpsp 0.3297 0.i465 0.5680 -0.131 CN L 0.7851 0.1764 0.9784 0.1944 CN,SP 0.2143 0.1934 0.5468 -0.161 Source: Table 8. These numbers are (across funds) averages of the correlation coefficients computed between two variables for each fund. Fc example, the correlation coefficients between prices and NAVs are computed first for each fund, resulting in a new series of thirty-two observations. Then a univariate exercise is done on that series, to obtain a mean of 0.4136, standard deviation of 0. 178, and so or, The same procedure is used for the other four pairs of coefficients. The first coefficient (0.4136) reflects that returns on prices and on NAV have recorded some degree of co-movement over the period, in average terms. This provides thb. initial evidence of imperfect substitution between NIFs and their underlying assets (see Case III of the companion paper). Secondly, country funds price returns are slightly more correlated with the home country index returns (average of 0.4525) than to S&P500 index returns (0.3297). Given that the typica! 8 correlation among U.S. securities is about 0.5 to 0.6,9 this is preliminary evidence that country funds, on average, provide diversification benefits to a U.S. investor. On average, the EMs are also less correlated with the S&P530 in comparison to DMs, thus providing some support to the higher diversification potential of EMs over DMs. As expected, returns on NAV are much more closely associated with returns on local stock market indexes (0.7851) than to returns on S&P500 (0.2143). 11.2 - Components of Risk A risk components analysis is conducted by regressing NIF price returns on their corresponding local stock market returns (Model 1) and then on the S&P500 returns (Model 2), respectively. The detailed results are reported in Table 9 and are summarized below: Table 11. Decomposition of Risk Model I (%) Model 2 (%) Average systematic risk 22.8 13.1 Average unsystematic risk 77.2 86.9 Total risk (variance) 100.0 100.0 Maximum unsystematic risk 95.3 100.0 (India) (Turkish) Maximum systematic risk 63.1 32.5 (Emer. Germ) (Irish 1.) Source: Table 9. 9Chapter 8 of Frank K. Reilly, 'Investment Analysis and Portfolio Management", Third Edition. 9 For each model and each fund, the residual variance is computed by squaring the residuals from the estimated model. Systematic risk is given by the square of the product of the beta (the slope of the regression) with the market return standard deviation. The results of all regressions are reported in Table 9. Averages are across all regressions, for each model. aFwud) = p2aMarket) + 2(Residua) where, p= Cov[R(Fund),R(market)] Var[(R(market)] and R (market) = Home market in model 1 S&P500 in model 2 R (Fund) = Price return On average, the proportion of variance, that is attributable to the unique features of country funds, and not by the movements of their local stock markets, is about 77% in the case of Model 1. The average share increases up to 87% when the market measure used characterized by the movements of the S&P500 index. Another interesting feature is that the 10 maximum systematic risk proportion for Model 2 is about one half of the maximum of same risk for Model 1. Again, these results suggest that international diversification via NIFs is desirable for the U.S. investor. It also points out that, on average, the NIFs are not a good proxy for their corresponding local market index (i.e., NIF-based diversification gains would be lower than theoretical diversification gains based on (inaccessible) local market indices). 11.3 - Mean-Variance Efricient Frontiers Although there are 18 EM and 14 DM funds that trade on the NYSE representing 13 EMs and 10 DMs, in this subsection, we include all NIFs that continuously traded over the period July 1989 to June 1991. Thus, the data consists of weekly returns on 8 EM and 6 DM national index funds traded on the NYSE and the corresponding data on 7 EM and 6 DM market portfolios. The efficient frontiers are defined as usual as the set of portfolios that have less risk than any other with comparable expected return, and more return than any other with comparable risk. These frontiers were developed using standard packages using historical time series of returns.' 11.3.1 - Idealized Diversification Gains We first investigate whether the diversification benefits documented in the past studies carry through to the more recent period and for our sample countries. Specifically, we inquire whether the benefits (in terms of mean-variance (M-V) efficiency) to the passive U.S. investor sequentially increase as (s)he diversifies into developed and emerging markets. As in the previous studies, we use various market indices and assume no barriers to international 10 Since the local market indices do not include dividends, we do not include distributions of the sample NIFs in computing returns. We use the S&P500 without dividends as a proxy for the U.S. market return. 11 investments, and hence the diversifications are presumed to be achitved in an idealized, costless manner. Figure 2 plots the M-V efficient frontiers for 6 developed and 7 emerging market indices over the July 1989-June 1991 period. It is apparent that the passive U.S. investor (in S&P 500) would have improved performance by diversifying into other markets. 'ionsistent with past studies, the benefits of such investments into EMs are substantially larger than those of developed markets. The S&P 500 portfolio is sequentially dominated by efficient frontiers based on developed markets, emerging markets and the global markets." Thus, the traditional diversification argument carries through to the most recen1t period. H.3.2 - U.S. Based Diversification The previous results (and past studies) do not account for the possible impediments facing U.S. investors in accessing the sample countries. It is now a common knowledge that there are a host of market imperfections (barriers) that inhibit free portfolio flows across national boundaries, particularly emerging and less developed economies. The barriers may take the form of border taxes, exchange controls or capital flow restrictions. Further, the pricing relationships undergo substantial revision on removal of such barriers. Thus, the gains from diversification documented in past studies and the previous section may be illusory. Finally, due to regulatory restrictions (% of foreign traded assets that can be held in a pension fund) and personal preferences, U.S. investors may wish to restrict their investment opportunity set to securities traded on the home market. Thus, this section restricts the U.S. investor opportunity set to securities traded on the NYSE. "Note that at the lower risk levels, the global frontier dominates the EM frontier. At higher levels of risk, the two (global and EM) frontiers overlap since no developed markets enter the M-V efficient portfolios. 12 Figure 3 plots M-V efficient frontiers based on 6 NIFs from developed countries and 8 NIFs from the emerging markets over the July 1989-June 1991 period.'2 Although the benefits of international diversification are not as dramatic as those in the previous section, they are substantial. Again the S&P 500 portfolio is dominated by developed market NIFs which in turn is dominated by the frontier based on emerging market funds. Thus, the benefits of international diversification are real and the portfolio performance can be substantially enhanced by including emerging markets in the opportunity set. To summarize, the traditional argument of intemational diversification carries through to the most recent period. The benefits reported in this section are real based on freely traded and accessible assets and suggest the advisability of global diversification that includes assets from emerging markets. IH - PRICING OF NIFs The evidence presented in the previous section suggests that, although NIFs provide substantial diversification benefits to U.S. investors, the gains are smaller than if they had access to the originating market portfolios or if the funds had been designed to mimic the local index (i.e., a true national index fund). This raises an important question as to tfle pricing of these funds. Specifically, do these funds behave like domestic U.S. securities or follow the originating country returns? Bailey and Lim (p. 8, 1989) conclude that, "countrv funds are priced more like domestic U.S. stocks than the foreign equities they are invested in." They consider intraday 12 All results are reported for returns based on prices since N1Fs can only be traded (as a unit) on price basis, Note that the risk-return perfonnance based on NAV dominate that hased on prices for all NiFs. Detailed results are available from the authors. 13 correlations and volatilities during trading and non-trading hours. Their tests follow the existing empirical literature on cross-border stock market relationships. Although they attempt to explain these results, we must study this issue further based on theoretical insights of the companion paper. Note that their conclusion is consistent with the prediction of Case 11 of the companion paper which rests on perfect substitution and imperfect arbitrage. [A. - Imperfect Substitution As noted earlier, the return behavior of NIFs in our sample does not qualify them as perfect substitutes for the underlying assets traded in the home market. As further evidence, consider the ratio of standard deviations of price returns and NAV returns for the sample NIFs [ratio (1)/(2)] as reported in Table 10. In all cases the price returns display substantially higher volatility compared to the NAV returns. The mean of the ratio of volatilities is 2.12 with a standard deviation of 0.79. The only exceptions are Turkish and Brazil funds whose portfolios had substantial holdings of the U.S. T-bills during the period studied. This leads us to consider the empirical implications of Case III of the companion paper, that admits imperfect substitution, in what follows. 11I.2 - Methodology The multiple-partial correlation coefficients are used to study the relationship between returns on funds and the given market factor(s) while controlling for the influence of the other factor(s). For example, we first test the importance of the U.S. factor while controlling for the originating country factor. That is, we test the hypothesis, 14 (1) Ho: pR1(R1,5)IRd=O using sample multiple-partial correlation. To test the hypothesis, we calculate the F statistics, F [SSR(Rdi in Model)-SSR(R,,RdI in ModeO]/k SSR(R,J,R& in Model)/n-p where Ri is the price return on ith fund, R", is the return on S&P500, Rd, is return on the ith market index, k refers to number of restrictions (one in this case), n is the number of observations, p is the number of parameters (total number of independent variables plus the constant - 3 in this case), and SSR (-) refers to the relevant sum of squared residuals. We reject Ho at c level if F 2 FK, ,,p .- We test the following other hypotheses, (2) Ho: pRci(Rd,) IRs=O (3) Ho: p Rci(R,is) gRF=O (4) Ho: p RCi(R) IRF=O (5) Ho: pRc1(Rs) IRFRdj=O (6) Ho: pRcg(Rdj) IRRFRss=O where RF iS the return on global fund index based on total sample. This index is value-weighted and calculated for Price and NAV series corresponding to the two (developed and emerging) subgroups of funds and the total sample. Details are available from the authors. Note that the tests using global fund factor that conforms neither to the originating countries nor to the host 15 countries is based on the prediction of the Case III of the companion paper. We argued in the companion paper that there may be other factors that are unrelated to either originating countries or reference countries that affect country fund prices. With such "imperfect substitution", we postulated that there may be a factor related to noise trading activity that would be common to all finds. We attempt to capture this factor through the construction of the global fund index based on our sample of NIFs. m.3 - Results To test the hypotheses outlined above, the following regressions were run for each ith fund since their time of inception: Ri?=a1 +P,(Rdf)+Pl Rci a2 + P2(Rdi) +r2(R,) + 112 Rc =a3 +r3(Rus) + P3 Rc,= a4+84(RF)+l`4 R4c = CS +rS(R,) + 85(RF) + P5 Rci =a 6 + P6(Rdd) + 8 6(RF) + P6 Rci=a7 + P7(Rd!) +r7(R,,) +87(RF) + 117 Tables 1 la and 1 lb report the relative importance of the domestic, U.S. and the (total sample) global factors for the NIFs from developed and emerging countries respectively. Specifically, we report the calculated F values and their significance levels for the six hypotheses under investigation. Similar results are reported in Tables I Ic and I Id with the total sample global factor being replaced by the subgroup (developed or emerging) global factor for the corresponding group of NIFs. Let us first review the importance of the U.S. factor in explaining the price returns of developed market NIFs (Table I la). The U.S. market factors' importance is significant for 8 16 of the 14 funds after taking into account the influence of the corresponding domestic market factor. The U.S. market factor is less important (for 5 of the 14 funds) if we were to take into account the global factors influence. The importance of the U.S. factor almost disappears (except in one case) when we take into account the contributions of the domestic as well as the global factors. Let us now consider the importance of the domestic factor. It is significant for 12 of 14 funds after taking into account the influence of the U.S. factor and for 6 of 14 funds when the impact of global factor is taken into consideration. If both the U.S. and the global factors are included, the importance of the domestic factor in explaining price returns of developed market NIFs is reduced to 5 of 14 funds. The findings remain unaltered when the total sample global factor is replaced by a factor based on only the developed market NIFs (Table I lc). Thus, for our sample of developed market NIFs, the global factor seems to be the most important in explaining price returns followed by the domestic market factor. The U.S. factor does not seem to be important except in the case of Spain fund price returns which seem to be affected solely by the movements in the S&P 500 index return. The results are very similar for emerging market funds (Tables 1 lb and I Id). Although, the U.S. factor is important in the presence of the home market factor, its importance declines precipitously when the global factor is taken into account. Although the home factor also becomes less important given the global factor, it remains significant in a majority of cases. To summarize, the results of this section provide strong support to the theoretical predictions of the companion paper. The case III of the companion paper, which admits imperfect substitution and suggests the presence of an additional factor common to NIFs, is borne out by the importance of the global index factor in explaining price returns of NIFs from 17 EMs and DMs as reported in this section. This finding has important implications for the design of NIFs and policies to reduce imperfect substitutability of the funds and its component assets traded in the home market. IV - IPOs OF NIFs The theoretical and empirical literature dealing with the underpricing of 1POs of individual U.S. firms is extensive.'3 These authors contend that underpricing results from information asymmetry and gaming strategies among various IPO participants. On the other hand, Mauer and Senbet (1992) develop an equilibrium model of IPO's that suggest that the so- called underpricing is a fair price differential based on incomplete access and imperfect substitution of the IPO in the secondary market. Recently, Peavy (1990) tests the IPOs of closed-end funds, and reports a mean initial return not significantly different from zero and attributes it to knowledge regarding the value of the underlying assets i.e. low information asymmetry. His conclusions also hold for the subset of international closed-end funds if three special-access funds that prohibit direct portfolio investments by U.S. investors are excluded. He also reports significantly negative after market returns. In a similar vein, Bailey and Lim (1990) also report statistically insignificant initial returns on average with high positive returns on funds specializing in Pacific Rim and Eastern Europe. With respect to aster market returns, they report poor performance except in the case of Pacific Rim countries. '3See for example, Ibbotson (1975), Ritter (1984) and Rock (1986). 18 IV.1 - lnitial Returns For our sample of NIFs, we calculated the initial returns defined as the offering day's closing price minus the offering price divided by the offering price.'4 The results are reported in Table 12. With respect to the emerging market funds, the initial returns are all positive with the exception of Emerging Mexico Fund and insignificantly negative returns for India and Mexico Equity/Income Funds. The mean return is a highly significant 6.44%, suggesting significant underpricing of EM funds. The information asymmetry and the difficulty of access, coupled with the diversification potential of EMs, may give rise to this initial return. For developed markets, with some exceptions, returns cluster around zero consistent M ith past findings. The average mean return is 5.02% which reduces to an insignificant 0.74% if we exclude the abnormal performance of the New Germany Fund. IV.2 - After Market Returns The after market returns are calculated as percent returns from the first trading day to the ninetieth calendar day i.e. end of the 13th week. The results are reported in Table 12. The after market returns for EMs on average are slightly negative. If we were to exclude the abnormally high return on the Taiwan Fund, the returns become substantially lower. Inclusion of the impact of the opportunity cost (i.e., T-bill interest rate) would further lower the performance. As a subgroup, the Pacific Rim countries outperform the other EMs. Thus, the result is consistent with Bailey and Lim (1990). With respect to the developed market funds, "'Due to uncertainty regarding offering date and data problems, the initial return is based on the closing price of 2nd or 3rd trading day for a few funds. This should not cause any concern since as Peajy (p. 697, 1990) states, .semi-strong form market efficiency implies that closed-end fund returns will not be significantly different from zero on days subsequent to the initial trading day". 19 the returns are more negative (-12.79%), on average, and are similar to those reported by Peavy (1990). IV.3 - The Seasoning Effect As suggested in Mauer and Senbet (1992) and as extended to the NIFs in the companion paper, we would expect a lowering of the underpricing due to the access effect, as new funds that are spanned (by existing funds) in the host market are issued. This is bome out in the case of Mexico - Mexico Equity/Income - Emerging Mexico; Taiwan - R.O.C. Taiwan; Indonesia - Jakarta and New Germany - Future Germany - Emerging Germany. Note that we do not include Germany Fund, which was issued in 1986, wh'ereas the other three German Funds were issued during January-March 1990, a period characterized by German reunification and political changes in Eastern Europe. Only the Spain - Growth Spain Funds do not conform to the seasoning hypothesis. Of course, a rigorous examination of this issue should involve a careful consideration of the issue dates, market environment, the fund investment objectives, relative issue sizes, etc. V - PREMIUMS/DISCOUNTS ON NIFs Premiums/discounts are determined by comparing NAV with closing market prices. When the market value of a share is above its NAV, the fund is selling at a premium, and when the market price is below NAV it is selling at a discount. Both NAV (the current value of the component, underlying assets) and the closing market prices are readily observable, since the underlying assets are marketable securities in the local markets, and the closing prices are obtained from the stock markets of the host countries. Thus, 20 PREMJUWDISCoUNT= (PRICE-NA EX1OO NA V Average weekly premiums/discounts for each of the thirty-two funds and their corresponding standard deviations and coefficient of variation are reported in Tables 13 and 14 respectively. V.1 - Determinants of Premium/Discount On the basis of the existing literature and the theoretical insights of the companion paper, ue can postulate the premium (discount) as dependent on (a) degree of access to the local market, (b) degree of spanning of local assets within the host market, (c) degree of substitution between the fund and its underlying assets, (d) the fund size, and (e) global country fund discount. ' V.2 - Variable Derinitions Premium/discount (PD): The premium/discount for all funds show high fluctuations over time. Detailed data are available from the authors. Given the scope of this project, we do not conduct empirical tests that would explicitly consider the investor sentiment/noise trader hypotheses.'6 Rather, we conduct cross-sectional and time-series analysis as detailed below. Degree of access (ACC): It is very difficult to systematically classify our sample by degree of access. No study exists (to our knowledge) that would provide us with indicators or "5As discussed in the companion paper, the premium (discount) will also depend on differential pnce of risk, differential tax rates, differential real interest rates, changes in market sentiment and noise trading, arbitrage restrictions and expropriation risk. Given the limitations of data and scope of the project, these variables will not be explicitly considered. "6As Long, Shleifer, Summers and Waldmann 1990) state, "The fluctuations in noise trader opinion of the expected return on the funds also explain why the discounts fluctuate". Although there appears to be some corroborating evidence in favor of the hypotheses put forward by Lee, Shleifer and Thaler (1991), a systematic analysis would require some additional data and modelling. 21 benchmarks to construct al index. As the second best alternative, we have divided our sample countries into three categories based on IFC Emerging Stock Markets Factbooks 1989 and 1990. (a) Completely open:- All ten developed markets, Singapore, Malaysia, Portugal (b) Relatively easier access:- Thailand, Indonesia, Mexico, Philippines, Turkey (c) Virtually closed:- Korea, Taiwan, Brazil, India, Chile. We would expect the premium to be relatively higher for funds that invest in markets with difficult access. Degree of spanning within host mark.t (SPN): As discussed in the companion paper, the availability of substitute assets (for a given fun,:) in the host market would determine the potential diversification benefits of the fund uneier nsideration. The natural proxy would be the residual volatility of a fund obtained from time series regressions on the U.S. index. Since this proxy would suffer from measurement error problem, we cor ider another proxy for diversification benefits, namely the host and home market retums correlation coefficient. A priori, we would expect higher premium for funds with the lower correlation. Degree of substitution between the fund and the underlying assets (SUB): The results of the companion paper suggest that the degree of substitution between the fund and the underlying asset has ar, important bearing on the premium/discount. We use the ratio of volatility of price and NAV returns as a reasonable proxy to capture the degree of imperfect substitution. We would expect a higher ratio to have a lower premium effect. Fund size (FSZ): The initial size of the offering could proxy investor demand, since the size is usually determined by the investment bankers to reflect market interest and conditions. We would expect a larger initial offering to command a larger premium. 22 Global country fund premium/discount (AGB): Given the theoretical prediction of a common NIF factor and the results of Section III that suggest the significance of the global factor, we incorporate the global premium/discount as an independent variable. We would expect a positive relationship between the individual fund premium/discount and the global premium/discount. V.3 - The Test Procedure Following the preceding discussion, we postulate the following relationship: [(PD),,] =Constant+ Pl(A CC) + P 2(SPN)j,, +P 3(SUB),1 + P4(FSZ) + P (A GB), +error term The variables in the above relationship are as defined previously. Subscripts i and t denote the ith fund and t' period, respectively. It should be noted that the [(AGB)] discounts vary over time whereas the degree of Access (ACC) and the fund size (FSZ) vary across the funds. That is, depending on the variable, there is time variation, variability across funds, or both. In order to capture these properties, we conduct time series and cross-sectional tests. Time-Series Tests: For each fund, the test uses all available data. Since the global country fund premium/discount can be calculated from January 1989 (based on 13 available NIFs) the test period begins in January 1989 or later. We use correlation coefficient between the host and home market returns as the proxy for (SPN) and the ratio of price to NAV return volatility as the proxy for (SUB). In both cases, we use a proxy formation period of 26 weekly observations (host-home market returns for correlation and price - NAV returns for volatility) preceding the test period to estimate the two proxies (SPN) and SOB). That is, observations for t= -25,... 0 periods are used to arrive at the proxy estimate for the t= I period. Similarly, 23 observations for t = -24, ... 0, I periods are used to arrive at the proxy estimate for the t = 2 period and so on. Thus, the proxy formation period for eac", subsequent test period is obtained by replacing the first observation of the previous proxy formation period by the observation corresponding to the last test period. Finally, the following OLS regression is run for eacn ith fund. [(PD),] =Constant + P l(SPN), + ,B2(StUB), + P 3(A GB), + error term Cross-Sectional Tests: Seventeen of the funds in our sample have complete data beginning in Deeember 1989. The sample increases to twenty-nine funds in July 1990. Since 6 months (twenty-six observations) are used to calculate proxies (SPN and S0B) for each fund as described under the time-series tests above, we have a total of 57 periods (27 for a seventeen fund sample and 30 for a twenty-nine fund sample) for cross-sectional tests. The following OLS regressions are run for each of the 57 periods (weeks): [(PD) -A GB] = Constant + p1 (SPN)I + p2(SUB)1 + p 3(FSZ)1 +error term where all variables are defined as before. Since the global country fund premium/discount is invariant across funds for a given period, we use the NIF premium/discount net of the AGB as the dependent variable as in Errunza (1991). These regressions result in weekly OLS values of the coefficients (p's) for each of the 57 periods. Note that although it would be interesting to include the access variable in these regressions, the use of two dummy variables to capture three categories would add two intercept and six slope terms (on R.H.S.). Given the very small sample sizes, we have found it appropriate to exclude this variable. 24 V.4 - Test Results The time series test results are reported in Table 15. Since tne residuals indicated significant auto correlation, we used the Cochrane-Orcutt procedure. The resulting residuals are well behaved. In all cases, the adjusted R2 are reasonably high - they range from a low of 29.6% for France Fund to a high of 95.98% for Korea Fund. The spanning variable is significant for only four sample NIFs. The substitution variable does somewhat better being significant in 9 of the 29 cases. Further, 3 out of 4 and 5 out of 9 significant coefficients for spanning and substitution proxies respectively suggest positive relationship with the premium/discount. Thus, the results are not very encouraging with respect to these two variables. Potential difficulty may lie with the choice of proxies. For example, both the proxies are very volatile in case of most funds. During our test period, the correlation coefficient (SPN) for Thailand moves over time from positive (high of 0.34) to negative (low of -0.32) to positive (high of 0.58). With respect to the global premium/discount va,iable, the results are as predicted by the theory, the only exceptions being the Brazil and Turkish Funds. Coefficients for all other funds are positive and very significant. The cross-sectional regression results on a weekly basis are very weak. In most cases, the adjusted R2 are zero and coefficient estimates are not significant. This is neither very surprising nor contrary to our theoretical rnodel which states the equilibrium relationship between prwi!iurn/discounts and the various independent variables. The postulated relationship should hold on average and not necessarily, week by week. Thus, we report the average values and summary measures of the time-series properties of coefficients Pl,, P2, and 03, below in Table IIl. 25 Table 11. Determinants of Premiums/Discounts on NIFS. OLS Estimates pit P2. i Mean -0.0434 -0.0162 0.0001 Standard Deviation 0.1726 0.0307 0.0003 t statistic -1.898 -3.983 2.5 17 Significance Level 0.062 0.001 0.014 The above result provides strong support to the predictions of the theoretical model. The spanning and substitution effects are negative and significant. whereas the size effect is positive and significant. To summarize, the results of the time series regressions strongly support the theoretical prediction of a common NIF factor. The week results for the spanning and substitution effects need to be further studied based on better proxies and more powerful time series models. The results of cross-sectional regressions are very encouraging, even though a larger sample should strengthen the results. Filially, efforts to quantify some of the variables (e.g. access, taxation effects, political risk, asymmetric information and valuation) not included in the above tests should prove ;nteresting and useful. VI - CONCLUDING REMARKS This paper has provided an empirical investigation of national index funds on the basis of the available data. The available data are quite limiting, both in time series and cross- sectional terms, but the results are, nonetheless, encouraging. The evidence suggests a 26 significant diversification benefit to US investors arising from NIFs, particularly those funds originating from countries with limited access to their local markets. Further, the evidence supports the predictions of the companion theoretical paper with regard to the pricing of country funds, relative to their net asset values. Variables that proxy the degree of access and substitution effects show up as significant determinants of country fund premia/discounts. The support for the theoretical analysis is particularly useful, since the companion paper has advanced a number of welfare implications for emerging economies that originate country funds. The model suggests that country funds can enhance pricing efficiency in the local capital markets and promote local capital mobilization by firms at terms more favorable (lower costs of capital). Thus, we are encouraged that the policy implications drawn from the theoretical analysis for the promotion and support for country funds have an empirical basis. 27 REFERENCES Bailey, W., and J. Lim. "Evaluating the Diversification Benefits of the New Country Funds." Working paper, Cornell University and National University of Singapore, May 1990. Bailey, W., and J. Lim. "Initial Public Offerings of Country Funds: Evidence and Implications." Working paper, Cornell University and National UniversLcy of Singapore, 1990. Delong, J.B., A. Shleifer, L.H. Summers, and R.J. Waldmann. "Noise Trader Risk in Financial Markets." Journal of Political Economy, 98, 1990, pp. 703-738. Diwan, I., V. Errunza, and L. Senbet. "The Pricing of Country Funds and Their Role in Capital Mobilization for Emerging Economies." Working paper, No. 1058, The World Bank, December 1992. Errunza, V. "Gains from Portfolio Diversification into Less Developed Countries' Securities." Journal of International Business Studies, Fall-Winter 1977, pp. 83-99. Errunza, V. "Pricing of National Index Funds. " Review of Quantitative Finance and Accounting, 1, 199 1, pp. 91-100. Errunza, V. and L. Senbet. "The Effects of International Operations on the Market Value of the Firm: Theory apd Evidence," Journal of Finance, 36, 2, 1981, pp. 401417. Errunza, V., and E. Losq. "Capital Flow Controls, International Asset Pricing, and Investors' Welfare: A Multi-Country Framework." Journal of Finance, Vol. XLIV, No. 4, September 1989, pp. 1025-1037. lbbotson, R.G. "Price Performance of Comrnon Stock New Issues." Journal of Financial Economics, 2, 1975, pp. 235-273. Levy ";., and M. Sarnat. "International Diversification of Investment Portfolios." American Economic Review, September, 1970, pp. 668-675. Mauer, D., and L. Senbet. "The Effect of the Secondary Market on the Pricing of Initial Public Offerings: Theory and Evidence." Journal of Financial and Quantitative Analysis, 27, No. 1, 1992, pp.55-79. Peavy, J.W. "Returns on Initial Public Offerings of Closed-End Funds." Review of Financial Studies, 3, 1990, pp. 695-708. Ritter, J.R. "The 'Hot Issue' Market of 1980." Journal of Business, 57, 1984, pp. 215-240. Rock, K. "Why New Issues Are Underpriced." Journal of Financial Economics, 15, 1986, pp. 187- 212. Stulz, R. "On the Effects of Barriers to International Investment." Journal of Finance, Vol. 36, September 1981, pp. 923-934. 28 Table 1 LIST OF CLOSED-END COUNTRY FUNDS PUBLICLY TRADED IN NEW YORK Listed in Inception of fund Weeks trading Exchange (initial offering) thru 12/31/90 Emerging Stock Markets 1 Brazil Fund NYSE March 1988 144 2 Chile Fund NYSE September 1989 66 3 Emerging Mexico Fund NYSE October 1990 13 4 First Philippine Fund NYSE November 1989 60 5 India Growth Fund NYSE August 1988 125 6 Indonesia Fund NYSE March 1990 44 7 Jakarta Growth Fund NYSE April 1990 38 8 Korea Fund NYSE August 1984 332 9 Malaysia Fund NYSE May 1987 191 10 Mexico Equity/Income Fund NYSE August 1990 20 11 Mexico Fund NYSE June 1981 500 12 Portugal Fund NYSE November 1989 61 13 R.O.C. Taiwan Fund NYSE May 1989 86 14 Singapore Fund NYSE July 1990 23 15 Taiwan Fund NYSE December 1986 211 16 Thai Capital Fund NYSE May 1990 32 17 Thai Fund NYSE February 1988 150 18 Turkish Investment Fund NYSE December 1989 57 Developed Stock Markets 19 Austria Fund NYSE September 1989 67 20 Emerging Germany Fund NYSE March 1990 40 21 First Australia Fund AMEX December 1985 263 22 France Growth Fund NYSE May 1990 34 23 Future Germany Fund NYSE February 1990 44 24 Germany Fund NYSE July 1986 233 25 Growth Fund of Spain NYSE February 1990 46 26 Irish Investment Fund NYSE March 1990 40 27 Italy Fund NYSE February 1986 253 28 Japan OTC Equity Fund NYSE March 1990 42 29 New Germany Fund NYSE January 1990 49 30 Spain Fund NYSE June 1988 132 31 Swiss Helvetia Fund NYSE August 1987 176 32 United Kingdom Fund NYSE August 1987 178 Multi-Country Funds 33 Alliance New Europe Fund NYSE March 1990 34 Asia Pacific Fund NYSE April 1987 35 Europe Fund NYSE April 1990 36 First Iberian Fund AMEX April 1988 37 G.T. Greater Europe Fund NYSE March 1990 38 Latin America Invest Fund NYSE July 1990 39 Pacific-European Growth Fund AMEX April 1990 40 Scudder New Asia Fund NYSE June 1987 41 Scudder New Europe Fund NYSE February 1990 42 Templeton Emerging Markts NYSE February 1987 43 Templeton Value Fund NYSE October 1988 44 Worldwide Value Fund NYSE Augus5 1986 Notes: (a) NYSE = New York Stock Exchange - AMEX = American Stock Exchange (b) Exact dates can be found in the 1ECDI database. (c) Equal to the number of observations in the IECDI database. (d) Listed in the AMEX until December 1988. 29 Table 2 CAPITALIZATION OF COUNTRY FUNDS (ORIGINAL SIZE AND MARKET VALUE AT END OF 1990) Initial Size Shares Price Value (Mill.) (Mill.) ($/Sh.) (Mill.S) Emerging Stock Markets 1 Brazil Fund 150.00 12.04 6.750 81.27 2 Chile Fund 80.50 5.37 15.500 83.29 3 Emerging Mexi:o Fund 60.00 5.01 9.000 45.08 4 First Philipptne Fund 107.64 8.98 6.37S 57.25 5 India Growth Fund 60.00 5.01 10.750 53.84 6 Indonesia Fund 69.00 4.61 10.000 46.07 7 Jakarta Growth Fund 60.00 5.01 6.750 33.81 8 Korea Fund 150.05 20.84 12.375 257.90 9 Malaysia Fund 87.00 7.26 11.000 79.85 10 Mexico Equity/Income Fund 72.00 6.01 9.875 59.34 11 Mexico Fund 134.60 19.72 12.625 248.97 12 Portugal Fund 79.35 5.30 9.375 49.66 13 R.O.C. Taiwan Funi 375.57 25.78 7.750 199.76 14 Singapore Fund 60.00 5.01 8.750 43.83 15 Taiwan Fund 81.92 4.07 21.250 86.49 16 Thai Capital Fund 72.00 6.01 6.750 40.56 17 Thai Fund 115.00 9.60 15.375 147.57 18 Turkish Investment Fund 84.00 7.02 6.750 47.41 Developed Stock Markets 19 Austria Fund 111.50 8.26 10.000 82.59 20 Emerging Germany Fund 168.00 14.01 7.625 106.81 21 First Australia Fund 60.00 6.01 7.250 43.54 22 France Growth Fund 120.00 10.01 8.125 81.32 23 Future Germany Fund 243.00 13.51 11.500 15'.32 24 Germany Fund 140.98 13.04 11.125 145.06 25 Growth Fund of Spain 216.00 18.00 8.000 144.00 26 Irish Investment Fund 60.00 5.01 6.750 33.81 27 Italy Fund 76.01 6.33 9.875 62.56 28 Japan OTC Equity Fund 102.00 8.51 8.250 70.20 29 New Germany Fund 431.25 28.76 11.375 327.11 30 Spain Fund 120.00 10.01 10.875 108.86 31 Swiss Helvetia Fund 120.00 8.01 11.875 95.08 32 United Kingdom Fund 48.00 4.01 9.000 36.07 Multi-Country Funds 33 Alliance New Europe Fund 252.00 21.00 8.250 173.25 34 Asia Pacific Fund 86.50 8.66 10.000 86.60 35 Europe Fund 108.75 7.26 11.250 81.64 36 First Iberian Fund 65.00 6.51 7.750 50.46 37 G.T. Greater Europe Fund 240.00 16.00 9.250 148.00 38 Latin America Invest Fund 60.00 4.01 10.750 43.08 39 Pacific-European Growth Fund 36.00 3.00 8.000 24.00 40 Scudder New Asia Fund 84.00 7.03 12.125 85.24 41 Scudder New Europe Fund 200.00 16.00 8.375 134.00 42 Templeton Emerging Markts 115.00 11.52 13.125 151.20 43 Templeton Value Fund 170.00 17.30 7.375 127.57 44 Worldwide Value Fund 60.00 3.00 12.125 36.38 30 Table 3 WEEKLY PRICE RETURNS EXCLUDING DIVIDENDS ARITHMETIC MEAN ANi STANDARD DEVIATION IN PERCENTAGE Country Fund Since Inception (a) Period 89/90 Year 1989 Year 1990 Name Mean Pr>D Std D Mean Pr>D Std D Mean Pr>D Std D Mean Pr>D Std D 1 Brazil -0.27 0.01 6.82 0.13 0.02 7.37 1.24 0.15 7.54 -0.98 0.04 7.09 2 Chile 0.06 0.04 6.96 b -0.45 0.92 8.82 c 0.19 0.02 6.51 3 Emerging Mexico -0.11 0.47 5.05 b -0.11 0.47 5.05 c 4 First Philippine -1.06 0.02 6.50 b 1.23 0.90 15.03 c -1.36 0.15 4.55 S India Growth 0.07 0.15 5.28 0.27 0.15 5.24 1.43 0.15 4.81 -0.89 0.15 5.43 6 Indonesia -0.91 0.41 6.58 b -0.91 0.41 6.58 c 7 Jakarta Growth -1.33 0.88 6.52 b -1.33 0.88 6.52 c 8 Korea 0.50 0.01 6.32 -0.52 0.02 6.40 0.61 0.15 4.25 -1.65 0.01 7.94 9 Malaysia 0.28 0.01 8.19 0.67 0.01 7.90 2.00 0.08 6.94 -0.67 0.02 8.61 .o Mexico Equity/Income -0.77 0.51 6.35 b -0.77 0.51 6.35 c 11 Mexico 0.30 0.01 7.37 1.02 0.01 6.33 1.53 0.01 5.88 0.52 0.15 6.77 12 Portugal -0.70 0.01 6.12 b 0.79 0.26 2.48 c -0.93 0.01 6.49 13 R.O.C. Taiwan -0.49 0.01 7.47 b -0.21 0.01 6.13 c -0.67 0.15 8.26 14 Singapore -1.28 0.78 5.52 b -1.28 0.78 5.52 c 15 Taiwan 0.06 0.01 10.05 -0.10 0.01 8.64 0.77 0.13 5.35 -0.96 0.03 10.98 16 Thai Capital -1.85 0.22 5.45 b -1.85 0.22 5.45 c w~ 17 Thai 0.13 0.01 6.92 0.55 0.06 7.81 2.21 0.01 7.44 -1.11 0.02 7.88 18 Turkish Investment -0.83 0.15 6.44 b 0.40 0.29 13.21 c -0.93 0.15 5.86 19 Austria 0.10 0.01 10.14 b 4.32 0.97 14.46 c -1.03 0.02 8.45 20 Emerging Gerrmany -0.80 0.01 6.30 b -0.80 0.01 6.30 c 21 First Australia 0.05 0.01 5.56 0.03 0.01 5.46 0.21 'j.01 3.85 -0.16 0.01 6.73 22 France Growth -1.27 0.02 7.85 b -1.27 0.02 7.85 c 23 Future Germany -0.89 0.18 6.13 b -0.89 0.18 6.13 c 24 Germany 0.29 0.01 7.04 0.77 0.01 9.32 2.24 0.01 9.54 -0.69 0.01 8.94 25 Growth F. Spain -0.88 0.49 5.09 b -0.88 0.49 5.09 c 26 Irish Investment -1.34 0.56 4.90 b -1.34 0.56 4.90 c 27 Italy 0.12 0.01 6.24 0.40 0.01 6.53 1.23 0.01 5.14 -0.42 0.02 7.64 28 Japan OTC Equity -0.71 0.01 7.00 b -0.71 0.01 7.00 c 29 New Germany -1.35 0.08 6.37 b -1.35 0.08 6.37 c 30 Spain 0.19 0.01 7.54 0.35 0.01 8.30 2.55 0.01 9.68 -1.86 0.05 5.94 31 Swiss Helvetia -0.02 0.01 4.54 0.29 0.02 4.24 0.96 0.01 4.19 -0.38 0.02 4.22 32 United Kingdom -0.03 0.01 5.00 0.05 0.01 4.27 0.32 0.01 3.38 -0.22 0.01 5.02 Notes: (a) If fund is incepted before year 1989. (bi Because inception occurs during year 1989 or 1990, parameter is estimated for less than 104 weeks. (c) Because inception occurs during year 1989 or 1990, parameter is estimated for less than 52 weeks. Table 4 WEEKLY PRICE RETURNS INCLUDING DIVIDENDS ARITHMETIC MEAN AND STANDARD DEVIATION IN PERCENTAGE - Country Fund Since Inception (a) Period 89/90 Year 1989 Year 1990 Name Mean Pr>D Std D Mean Pr>D Std D Mean P,->D Std D Mean Pr>D Std D 1 Brazil -0.07 0.01 6.73 0.33 0.02 7.26 1.26 0.15 7.57 -1.44 0.01 11.69 2 Chile 0.22 0.02 7.00 b -0.25 0 82 9.11 c 0.34 0.03 6.47 3 Emerging Mexico 0.15 0.65 4.49 b 0.15 0.65 4.49 c 4 First Philippine -0.90 0.01 6.43 b 1.32 0.90 15.03 c -1.20 0.15 4.43 S India Growth 0.15 0.15 5.34 0.34 0.15 5.29 1.55 0.15 4.95 -0.87 0.15 5.40 6 Indonesia -0.76 0.36 6.48 b -0.76 0.36 6.48 c 7 Jakarta Growth -1.28 0.86 6.50 b -1.28 0.86 6.50 c 8 Korea 0.60 0.01 6.36 -0.37 0.02 6.60 0.72 0.15 4.31 -1.47 0.01 8.18 9 Malaysia 0.32 0.01 8.22 0.69 0.01 7.89 2.02 0.07 6.97 -0.63 0.02 8.58 10 Mexico Equity/Income -0.48 0.31 6.22 b -0.48 0.31 6.22 c 11 Mexico 0.42 0.01 7.41 1.11 0.01 6.37 1.65 0.02 5.91 0.57 0.15 6.81 12 Portugal -0.67 0.01 6.12 b 0.85 0.24 2.41 c -0.90 0.01 6.49 13 R.O.C. Taiwan -0.42 0.01 7.45 b -0.11 0.01 6.16 c -0.62 0.12 8.22 14 Singapore -1.19 0.78 5.65 b -1.18 0.78 5.65 c 15 Taiwan 1.01 0.01 9.70 0.46 0.01 7.97 1.21 0.07 5.00 -0.28 0.01 10.11 16 Thai Capital -1.85 0.22 5.45 b -1.S5 0.22 5.45 c ta.~ 17 Thai 0.28 0.01 7.05 0.74 0.04 7.96 2.39 0.01 7.90 -0.90 0.03 7.86 18 Turkish Investment -0.83 0.15 6.45 b 4.77 0.28 13.29 c -0.93 0.15 5.86 19 Austria 0.17 0.01 10.13 b 4.36 0.97 14.51 c -0.95 0.01 8.44 20 Emerging Germany -0.74 0.01 6.28 b -0.74 0.01 6.28 c 21 First Australia 0.19 0.01 5.62 0.15 0.01 5.40 0.35 0.01 3.67 -0.05 0.01 6.73 22 France Growth -1.10 0.01 7.77 b -1.10 0.01 7.77 c 23 Future Germany -0.84 0.16 6.09 b -0.84 0.16 6.09 c 24 Germany 0.39 0.01 7.02 0.81 0.01 9.34 2.28 0.01 9.59 -0.66 0.01 8.94 25 Growth F. Spain -0.79 0.49 5.14 b -0.79 0.49 5.14 c 26 Irish Investment -1.23 0.57 4.81 b -1.23 0.57 4.81 c 27 Italy 0.24 0.01 6.40 0.51 0.01 6.45 1.25 0.01 5.14 -0.23 0.02 7.52 28 Japan OTC Equity -0.52 0.01 6.97 b -0.52 0.01 6.97 c 29 New Germany -1.30 0.11 6.42 b -1.30 0.11 6.42 c 30 Spain 0.30 0.01 7.46 0.47 0.01 8.21 2.62 0.01 9.62 -1.69 0.09 5.83 31 Swiss Helvetia -0.02 0.01 4.53 0.29 0.02 4.23 0.96 0.01 4.19 -0.37 0.02 4.20 32 United Kingdom 0.11 0.01 5.04 0.18 0.01 4.24 0.37 0.01 3.37 0.00 0.01 4.98 Notes:(a) If fund is incepted before year 1989. (b) Because inception occurs during year 1989 or 1990, parameter is estimated for less than 104 weeks. (c) Because inception occurs during year 1989 or 1990, parameter is estimated for less than 52 weeks. Table 5 WEEKLY RETURNS INCLUDING DIVIDENDS : GEOMETRIC MEAN IN PERCENTAGE Country Fund Price Returns NAV Returns Name Incep.(a) 89/90 1989 1990 Ineep. (a) 89/90 1989 1990 1 Brazil -0.298 0.063 0.974 -0.840 -0.314 -0.587 0.746 -1.902 2 Chile -0.019 b -0.638 c 0.136 0.517 b 0.732 c 0.464 3 Emerging Mexico 0.056 b 0.056 c 0.464 b 0.464 e 4 First Philippine -1.108 b 0.352 c -1.303 -0.196 b 0.116 c -0.238 5 India Growth 0.012 0.201 1.428 -1.011 0.265 0.208 0.513 0.096 6 Indonesia -0.960 b -0.960 c -0.417 b -0.417 e 7 Jakarta Growth -1.492 b -1.492 c -0.566 b -0.566 c 8 Korea 0.399 -0.586 0.631 -1.788 0.486 -0.123 0.509 -0.751 9 Malaysia -0.014 0.393 1.789 -0.984 0.088 0.334 0.841 -0.171 10 Mexico Equity/Income -0.668 b -0.668 c 0.279 b 0.279 c 11 Mexico 0.144 0.911 1.489 0.337 0.201 0.740 0.996 0.484 12 Portugal -0.858 b 0.825 c -1.114 -0.367 b 0.112 c -0.440 13 R.O.C. Taiwan -0.695 b -0.295 c -0.954 -0.397 b 0.274 c -0.821 14 Singapore -1.327 b -1.327 c -0.033 b -0.033 c 15 Taiwan 0.584 0.153 1.089 -0.774 0.756 0.151 1.038 -0.729 16 Thai Capital -2.000 b -2.000 c -0.924 b -0.924 c 17 Thai 0.046 0.448 2.118 -1.194 0.304 0.491 1.436 -0.445 18 Turkish Investment -1.306 b -0.188 c -1.100 -0.500 b 4.468 c -0.872 19 Austria -0.313 b 3.409 c -1.292 0.222 b 0.887 c 0.043 20 Emerging Germany -0.926 b -0.926 c -0.317 b -0.317 c 21 First Australia 0.028 0.015 0.288 -0.256 0.086 -0.121 0.069 -0.310 22 France Growth -1.384 b -1.384 c -0.069 b -0.069 c 23 Future Germany -1.020 b -1.020 c -0.369 b -0.369 c 24 Germany 0.164 0.415 1.863 -1.010 0.18, 0.325 0.799 -0.146 25 Growth F. Spain -0.919 b -0.919 c -0.123 b -0.123 c 26 Irish Investment -1.344 b -1.344 c -0.375 b -0-375 c 27 Italy 0.034 0.304 1.127 -0.514 0.117 0.254 0.541 -0.033 28 Japan OTC Equity -0.746 b -0.746 c -0.366 b -0.366 c 29 New Germany -1.500 b -1.500 c -0.147 b -0.147 c 30 Spain 0.037 0.150 2.204 -1.863 0.106 0.095 0.533 -0.342 31 Swiss Helvetia -0.120 0.206 0.873 -0.456 -0.030 0.177 0.331 0.024 32 United Kingdom -0.025 0.098 0.315 -0.118 0.083 0.082 0.118 0.045 Notes: (a) If fund is incepted before year 1989. (b) Because inception occurs during year 1989 or 1990, parameter is estimated for less than 104 woeks. (c) Because inception occurs during year 1989 or 1990, parameter is estimated for less than 52 weeks. Table 6 TOTAL REINVESTED CUNULATIVE RETURNS IN PERCNTWAGE Country Fund Price Returns NAV Returns Name Incep.(a) 89/90 1989 1990 Incep.(a) 89/90 1989 1990 1 Brazil -34.73 6.74 65.56 -35.52 -36.19 -45.79 47.17 -63.17 2 Chile -1.24 b -7.98 c 7.33 39.85 b 9.95 c 27.20 3 Emerging Mexico 0.67 b 0.67 c 5.71 b 5.71 c 4 First Philippine -48.1? b 2.49 c -49.43 -10.92 b 0.82 c -11.64 5 India Growth 1.49 23.26 109.07 -41.04 38.76 24.08 30.46 -4.89 6 Indonesia -33.95 b -33.95 c -19.52 b -19.52 c 7 Jakarta Growth -42.66 b -42.66 c -18.95 b -18.95 c 8 Korea 273.05 -45.71 38.72 -60.86 398.16 -12.00 30.20 -32.41 9 Malaysia -2.59 50.34 151.47 -40.21 18.27 41.44 54.58 -8.50 10 Mexico Equity/Income -11.96 b -11._- c 5.44 b 5.44 c 11 Mexico 105.54 156 89 115.64 19.13 172.52 115.17 67.39 28.54 12 Portugal -40.37 b 6.80 c -44.16 -19.78 b 0.90 c -20.50 13 R.O.C. Taiwan -44.72 b -9.00 c -3°.25 -28.70 b 9.45 c -34.85 14 Singapore -25.47 b -25.47 c -0.73 b -0.73 c 15 Taiwan 240.02 17.25 75.59 -33.22 386.05 16.94 71.09 -31.65 16 Thai Capital -46.53 b -46.53 c -25.00 b -25.00 c __ 17 Thai 7.17 59.19 197.36 -46.47 57.09 66.40 109.87 -20.71 4b, 18 Turkish Investment -44.17 b -0.75 c -43.75 -24.47 b 19.11 c -36.59 19 Austria -18.69 b 59.89 c -49.14 15.74 b 13.16 c 2.28 20 Emerging Germany -30.44 b -30.44 c -11.65 b -11.65 c 21 First Australia 7.67 1.61 16.11 -12.49 25.21 -11.83 3.64 -14.93 22 France Growth -36 .86 b -36.86 c -2.24 b -2.24 c 23 Future Germany -35.65 b -35.65 c -14.70 b -14.70 c 24 Germany 46.17 53.90 161.17 -41.07 54.18 40.19 51.26 -7.32 25 Growth F. Spain -33.99 b -33.99 c -5.39 b -5.39 c 26 Irish Investment -41.00 b -41.00 c -13.63 b -13.63 c 27 Italy 8.97 37.05 79.13 -23.49 34.21 30.14 32.38 -1.69 28 Japan OTC Equity -26.44 b -26.44 c -13.96 b -13.96 c 29 New Germany -51.61 b -51.61 c -6.82 b -6.82 c 30 Spain 4.98 16.84 210.66 -62.39 14.88 10.37 31.87 -16.30 31 Swiss Helvetia -18.96 23.92 57.14 -21.14 -5.11 20.22 18.74 1.24 32 United Kingdom -4.32 10.74 17.75 -5.9S 15.81 8.85 6.35 2.36 Notes: (a) If fund is incepted before year 1989. (b) Because inception occurs during year 1989 or 1990, parameter is estimated for less than 104 weeks. (c) Because inception occurs during year 1989 or 1990, paraneter is estimated for less than 52 weeks. Table 7 LOCAL STOCK INDEXES IN USS: WEEKLY PRICE RETURNS WITHOUT DIVIDENDS (ARITHHETIC MEAN & STD DEV IN %) Country Index Period 89/90 Year 1989 Year 1990 Mean Pr>D Std D Mean Pr>D Std D Mean Pr>D Std D 1 Brazil Bovespa (1968=0.0001' -0.15 0.14 12.54 1.07 0.04 11.30 -1.37 0.15 13.67 2 Chile IGPA (1/80=100) 0.80 0.01 3.35 b 1.43 0.47 3.92 c 0.64 0.02 3.22 3 India FE Bombay (1979=100) 0.35 0.15 3.47 0.30 0.15 2.73 0.41 0.15 4.10 4 Indonesia JSE Comp (10/82=100) -0.56 0.40 4.63 c Since 4/10/90 -1.40 0.34 4.14 c s Korea KSE Comp (1980=100) -0.26 0.15 2.92 0.06 0.15 2.52 -0.57 0.04 3.27 6 Malaysia KLSE Comp (1/77=100) 0.38 0.01 3.30 0.91 0.01 2.37 -0.16 0.01 3.97 7 Mexico BMV Gral (11/78=781.6) 0.85 0.15 3.52 1.02 0.15 3.17 0.68 0.14 3.86 Since 8/14/90 0.06 0.27 4.68 c Since 10/90 1.05 0.60 2.95 c 8 Philippines Manila Co/In (1/58=100 -1.40 0.01 6.79 b -2.65 0 02 10.49 c -1.24 0.04 6.26 9 Portugal Banco Tota/Aco (77=100) -0.50 0.01 2.82 b 0.21 0.35 1.32 c -O.il 0.01 2.98 10 Singapore Strait Times (1964=100) -1.09 0.39 5.26 c 11 Taiwan TSE Average (1966=100) 0.29 0.15 8.41 1.53 0.15 5.70 -0.96 0.15 10.35 Since 5/12/89 -0.37 0.15 8.87 b 0.55 0.74 5.87 c -0.96 0.15 10.35 12 Thailand SET Index (4/30/75=100) 0.60 0.01 5.57 1.58 0.15 2.16 -0.39 0.01 7.48 Since 5/22/90 -1.02 0.10 9.17 c 13 Turkey ISE Index (1/86=100) 1.23 0.15 9.77 b 7.61 0.99 5.95 c 0.74 0.15 9.87 14 Australia All Ordinary Shares -0.22 0.01 2.33 0.08 0.01 2.51 -0.51 0.15 2.12 15 Austria CA-Share Index (Atkien) 0.55 0.15 6.19 b 1.80 0.46 5.54 c 0.21 0.13 6.36 16 France CAC General Index -0.58 0.35 3.25 c 17 Germany FAZ Aktieni12/31/58=100) 0.30 0.08 3.18 0.70 0.13 2.46 -0.09 0.13 3.75 Since 1/24/90 -0.14 0.29 3.74 c Since 2/27/90 -0.14 0.27 3.76 c Since 3/29/90 -0.43 0.32 3.82 c 18 Ireland IESMISEQ -0.64 0.42 2.95 c 19 Italy Banca Com Ital (72=100) 0.04 0.01 2.66 0.39 0.01 2.13 -0.31 0.04 3.07 20 Japan Nikkei Avg (1/4/68=100) -0.31 3.83 5.62 c 21 Spaia Madrid G (12/30/85=100) 0.01 0.04 2.92 0.23 0.15 1.89 -0.22 0.15 3.68 Since 2/14/90 -0.18 0.68 3.87 c 22 Switzerland Swiss Bank Corporation 0.11 0.15 2.49 0.29 0.15 2.24 -0.06 0.15 2.72 23 United Kingdom FT 100 (4/10/62) 0.27 0.12 2.58 0.37 0.15 2.19 0.16 0.15 2.94 24 United States S&P500 Comp (41-43=100) 0.18 0.01 2.05 U.4h 0.01 1.86 -0.12 0.15 2.19 Notes: la) If fund is incepted before year 1989. Ib) Because inception occurs during year 1989 or 1990, parameter is estimated for less than 104 weeks. Ic) Because inception occurs during year 1989 or 1990, parameter is estimated for less than 52 weeks. Table 8 CORRELATION COEFFICIENTS FOR PRICE AND NAV RETURNS EXCLUDING DIVIDENDS - PERIOD 1989-1990 Country Fund Price Returns Price Returns versus NAV Returns versus Name vs. NAV Return Loc Mkt. IFC Indx S&P500 Loc Mkt. IFC Indx S&P500 1 Brazil 0.3006 0.2749 0.2836 0.2098 0.8626 0.8421 0.0630 2 Chile b 0.3232 0.3042 0.2079 0.3912 0.5486 0.6427 0.0819 3 Emerging Mexico b 0.5366 0.3889 0.4065 -0.1305 0.6879 0.7228 -0.0833 4 First Philippine b 0.2907 0.2793 0.2143 0.1703 0.5671 0.5174 0.1538 5 India Growth 0.1790 0.2167 0.1792 0.2032 0.5880 0.6391 -0.1610 6 Indonesia b 0.5838 0.4974 N.C. 0.3604 0.6019 N.C. 0.2509 7 Jakarta Growth b 0.2838 0.4120 N.C. 0.3126 0.8021 N.C. 0.0141 8 Korea 0.5020 0.5434 0.5766 0.2397 0.7254 0.6518 0.0314 9 Malaysia 0.3935 0.3818 0.3966 0.3289 0.9775 0.9606 0.2795 10 Mexico Equity/Income b 0.2135 0.3263 0.3804 0.4523 0.1944 0.1987 -0.1199 11 Mexico 0.5911 0.5828 0.5450 0.5216 0.9203 0.8039 0.4990 12 Portugal b 0.6179 0.5969 0.5378 0.4768 0.8302 0.8147 0.3748 13 R.O.C. Taiwan b 0.6058 0.5326 0.5531 0.3110 0.9040 0.8057 0.0374 14 Singapore b 0.1209 0.2346 N.C. 0.4021 0.7165 N.C. 0.5142 is Taiwan 0.2207 0.3647 0.4104 0.3203 0.4912 0.4138 0.0257 16 Thai Capital b 0.5634 0.5112 0.5519 0.3836 0.9745 0.9709 0.4481 17 Thai 0.3332 0.4029 0.3660 0.2824 0.9003 0.9132 0.2880 18 Turkish Investment b 0.5008 0.6009 0.5061 0.0063 0.9314 0.9222 0.2008 19 Austria b 0.2582 0.2975 N.C. 0.3546 0.9218 N.C. 0.256S 20 Emerging Germany b 0.7086 0.7941 N.C. 0.3963 0.9602 N.C. 0.3280 21 First Australia 0.3557 0.2928 N.C. 0.1563 0.7235 N.C. 0.3033 22 France Growth b 0.4054 0.6112 N.C. 0.4525 0.7696 N.C. 0.1284 23 Future Germany b 0.6426 0.7289 N.C. 0.3879 0.8547 N.C. 0.1421 24 Germany -0.0387 0.4012 N.C. 0.2164 0.6294 N.C. -0.1017 25 Growth F. Spain b 0.7175 0.7612 N.C. 0.4083 0.9595 N.C. 0.5468 26 Irish Investment b 0.4040 0.3560 N.C. 0.5680 0.9126 N.C. 0.2633 27 Italy 0.4302 0.4673 N.C. 0.3990 0.8727 N.C. 0.4059 28 Japan OTC Equity b 0.3835 0.4147 N.C. 0.4343 0.7747 N.C. 0.3463 29 New Germany b 0.4038 0.4924 N.C. 0.2265 - 8371 N.C. 0.3845 30 Spain 0.3096 0.2604 N.C. 0.3795 0 8776 N.C. 0.2488 31 Swiss Helvetia 0.5322 0.5241 N.C. 0.4376 0.9784 N.C. 0.3582 32 United Kingdom 0.S641 0.6268 N.C. 0.4909 0.8283 N.C. 0.3495 Notes: (a) If fund is incepted before year 1989. (bi Because inception occurs during year 1989 or 1990, parameter is estimated for 3ess than 104 weeks. (c) Because inception occurs during year 1989 or 1990, parameter is estimated for less than 52 weeks. Table 9 SYSTEMATIC & UNSYSTEMATIC RISKS AS PROPORTIONS OF PRICE RETURNS VARIANCE - PERIOD 1989-1990 Country Fund Variance of MODEL 1 YMODEL 2 Name Price Returns Sys Risk Uns Risk Total Sys Risk Uns Risk Total 1 Brazil 0.0054311 7.56% 92.44% 100% 4.41% 95.59% 100% 2 Chile b 0.0048464 9.26% 90.74% 100% 15.37% 84.63% 100% 3 Emerging Mexico b '>0025511 15.16% 84.84% 100% 1.70% 98.30% 100% 4 Fir-st Philippine b 0.0042314 7.80% 92.20% 100% 2.87% 97.13% 100% 5 India Growth 0.0027422 4.70% 95.30% 100% 4.17% 95.83% 100% 6 Indonesi.. b 0.0043279 24.74% 75.26% 100% 13.00% 87.00% 100% 7 Jakarta Growth b 0.0042520 16.98% 83.02% 100% 9.80% 90.20% 100% 8 Korea 0.0041461 29.42% 70.58% 100% 5.75% 94.25% 100% 9 Malaysia 0.0062388 14.58% 85.42% 100% 10.83% 89.17% 100% 10 Mexico Equity/Income b 0.0040372 10.65% 89.35% 100% 20.45% 79.55% 100% 11 Mexico 0.0040085 33.97% 66.03% 100% 27.23% 72.77% 100% 12 Portugal b 0.0037457 35.77% 64.23% 100% 22.90% 77.10% 100% 13 R.O.C. Taiwan b 0.0055830 28.36% 71.64% 100% 9.69% 90.31% 100% 14 Singapore b 0.0030451 5.50% 94.50% 100% 16.23% 83.77% 100% 15 Taiwan 0.0074608 13.30% 86.70% 100% 10.26% 89.74% 100% 16 Thai Capital b 0.0029686 26.14% 73.86% 100% 14.67% 85.33% 100% 17 Thai 0.0060984 16.23% 83.77% 100% 7.99% 92.01% 100% i8 Turkish Inventment b 0.0041453 36.10% 63.90% 100% 0.00% 100.00% 100% 19 Austria b 0.0102785 8.85% 91.15% 100% 12.46% 87.54% 100% 2O Emerging Germany b C.0039628 63.05% 36.95% 100% 15.84% 84.16% 100% 21 First Australia 0.0029779 8.54% 91.46% 100% 2.45% 97.55% 100% 22 France Growth b 0.0061676 37.36% 62.64% 100% 20.35% 79.65% 100% 23 Future Germany b 0.0037624 53.13% 46.87% 100% 15.06% 84.94% 100% 24 Germany 0.0086809 16.10% 83.90% 100% 4.73% 95.27% 100% 25 Growth F. Spain b 0.0025876 57.94% 42.06% 100% 16.61% 83.39% 100% 26 Irish Investment b 0.0023971 12.69% 87.31% 100% 32.53% 67.47% 100% 27 Italy 0.0042637 21.99% 78.01% 100% 16.00% 84.00% 100% 28 Japan OTC Equity b 0.0049003 17.20% 82.80% 100% 18.98% 81.02% 100% 2q New Germany b 0.0040613 24.24% 75.76% 100% 5.16% 94.84% 100% 30 Spain 0.0068813 6.40% 93.60% 100% 18.67% 81.33% 100% 311 Swiss Helvetia 0.0017958 27.56% 72.44% 100% 19.37% 80.63% 100% 32 United Kingdom 0.0018204 39.50% 60.50% 100% 24.14% 75.86% 100% Notes: (b) Because inception occurs during year 1989 or 1990, parameter is estimated for less than 104 weeks. Table 10 Icountry Funds: Comparative Volatilities (Period: Since Fund Inception Unuil 06/28/91) f OBS Fund STO Deviation STD Deviation STD Deviation Ratio Ratio Ratio ________ _____ |______________ Price Returns NA VReturns MKTReturns _ _________ ________________I_____________ (1) (2) (3) (IV(2) (1)(3) (2)1(3) Singapore Fund 0.0459842 0.0106574 0.039639 4.3;477 1.16009 0,26886 2 First Philippine Fund 0.0608341 0.0149597 0.066033 4.06654 0.82127 0.22655 3 Mexico Equity/Income Fund 0.0476027 0.0140794 0.039463 3.38101 1.20626 0.35678 4 Jakarta Growth Fund 0.0622949 0.0185163 0.038196 3.36433 1.63094 0.48477 5 France Gr-wth Fund 0.0647729 0.0216454 0.030433 2.99245 2.12835 0.71124 6 Portugal Fund 0.5900491 0.0220324 0.029354 2.60905 2.01162 0.77102 7 Spain Fund 0.0738364 0.0286853 0.027813 2.57401 2.65472 1.03136 8 Malaysia Fund 0.0628551 0.0248707 0.040010 2.52727 1.57098 0.62161 9 Malaysia Fund 0.0791697 0.0321039 0.038496 2.46605 2.05659 0.83396 10 Chile Fund 0.0638540 0.0272545 0.034339 2.34288 1.85953 0.79370 I I Italy Fund 0.0610389 0.0280369 0.034366 2.17709 1.77613 0.81583 12 _ mergind Mexico Fund 0.0548820 0.0267346 0.030923 2.05284 1.77480 0.86456 13 Korea Fund 0.0615942 0.0301289 0.031360 2.04436 1.96409 0.96074 14 New Germany Fund 0.0621019 0.0308592 0.036176 2.01242 1.71666 0.85303 15 __ Austria Fund 0.0881553 0.0455913 0.0563% 1.93360 1.56316 0.80U42 16 Emerging Germany Fund 0'.0595368 0.0312380 0.036536 1.90591 1.62952 0.85498 oo 17 Irish Investment Fund 0.00.08474 0.0238172 0.029734 1.88298 1.50827 0.80100 18 Swiss lHelvetia Fund 0.0442271 0.0240356 0.026437 1.84007 1.67290 0.90915 19 _ Japan OTC Equity Fund 0.0798827 0.0434208 0.046979 1.83973 1.70037 0.92425 20 United Kingdom Fund 0.0491958 0.0268796 0.029079 1.83023 1.69177 0.92435 21 _ Germany Fund _0 0692195 0.0380740 0.030568 1.81803 2.26441 1.24553 22 Futurc Germany Fund 0.0578084 0.0324509 0.036171 1.78141 1.59821 0.89716 23 First Australia Fund 0.0537279 0.0313747 0.024834 1.71246 2.16352 1.26340 24 Growth Fund of Spain 0.0467491 0.0273317 0.035770 1.71043 1.30693 0.76409 25 Thai Fund 0.067303 0.0417380 0.046891 1.61251 1.43530 0.89010 26 India Growth Fund 0.0511082 0.0321037 0.035580 1.59197 1.43722 0.90279 27 Thai Capital Fund 0.0708192 0.0450254 0.073010 1.57287 0.96999 0.61670 28 Taiwan Fund 0.0964076 0.0633112 0.075332 1.52276 1.27978 0.84043 29 Roc Taiwan Fund 0.0720099 0.0525485 0.084309 137035 0.85412 0.62329 30 __ Mexico Fund 0_ 0.0728115 0.0558739 0.064383 1.30314 1.13091 0.86783 31 Turkish Fund 0.0721610 0.0731441 0.093552 0.98656 0.77135 0.78185 32 Brazil Fund 0.0741052 C 0824737 0.111304 0.89853 _0.66579 0.74098 Note: Funds sorted by Descending Ratio (1)/(2) ]___I___I_I_I TABLE I la: Relative Importance of Domestic U.S. and Global Fund Factors - Developed ______ _ F Significance o' Domestic U.S. and Global Fund Factors Developed Markets #1 #IF #2 #2F #3 #3 F #4 U4F I 1SF #6 16 F Calculated Signifcance Calculated Significance Calculated Signmicance Calculated Significance Cakulated Significance Calculated Significanc F- Value Level F-Value Level F-Value Value F-Value LelW F-Value Level F-Valae Lcvl AustxiaFund 90412 0.0034 4.6539 0.0037 0.0298 0.8633 1.3131 0.2549 0.0138 0.9863 0.6482 0.5255 EmergingGennany 2.763 0.1015 58.6264 0 5.6486 0 0185 11.3553 0.0012 2.0779 0.134 4.818 0.0114 France Growtlh FIund 2.6857 0.1069 17.04 0.0001 0.4255 0.5169 3.6331 0.0618 0.1415 0.8684 1.7128 0.1899 First Australia Fund 2.9383 0.0889 8.6622 0.0039 0-2586 0.612 3.767 0.0545 0.1616 0.851 1.9029 0.1534 Future Gennany Fund 7.0494 0.0099 35.0906 0 2.9645 0.0898 4.3842 0.0401 1.3067 0.3T77 2.0027 014 Gemaiu Fund 2.6443 0.1064 21.7097 0 1.3072 0.2551 0.25626 0.1119 0.7599 0.4717 1.5797 0.2534 Growth Fumd of Spain 2.8443 0.0963 41.0398 0 0 1128 0.7314 17.68 0.0001 0.1641 0.849 8.8427 0.0004 Ifish Investment Fund 14.2924 0.0004 3-2894 0.0746 2.8404 0.0969 0.2543 0.615 1.338 0.27 0.0682 0.9341 Italy Fund 5.8462 0.0174 13.9319 0.0003 4.4483 0.0074 13.3873 0.0004 0.497 0.6099 4.8735. 0.0096 Japan OTC Equity Fund 11.0222 0.0015 4.7488 0-033 2.1377 0.1486 0.1241 0.7258 1.0126 0.5691 0.0227 0.9776 New Gerfnany Fund 3.2066 0.0776 19.4947 0 0.6277 0.4308 2.6789 0.1061 0.3273 0.722 1.3589 0.2688 Spain Fund 20.12 0 2.0681 0.1529 7.8903 0.0058 0.049 0.8342 3.94 0.0219 0.0462 0.9549 Swiss Helvesia Fund 114.0255 0.0003 24.9088 0 0.544 0.0025 21.2807 0 2.4271 0.0924 1.0509 0.0005 United Kingdom Fund 14.0595 0.0003 44.6:96 0 11.6432 0.0009 36.3809 0 2.854 0.0615 14.6041 0 s Percent Significant (At 5% Level) 57.14%f ___ 85.71% 35.71% 42.86%1 _ 7.14%1 ff 35.71% TABLE llb: Relatlive Importance of Domestic U.S. and Global Fund Factors I . I 1-~~~~~~~~~~~ I__ _ _ _ _ _ _ _ _ _ _ _ _ F Significance Level for V rious Hypotheses Testd_ Emerging Mlarkets #I #IF #2F 112F #3 113F 4 54 F i5 #5SF 56 #6 F ___________ Calculated Significance Calculated Significance Calculated Signifcanee Calculated Significance Calclated ignificance Calculated Significance F-Value level F-Value Level F-Value Value F- Value Level F-Value Level F-Value Level Birazil Fund 5.9215 00163 12-8648 0.0003 0.7202 0.3977 9.5304 0.0025 0.4255 0.6544 4.8006 0.0092 Chile Fumd 15.9348 0.0001 9.4936 0.0028 2.7837 0.1988 11.3872 0.0011 1.205i 0.3044 5.4423 0.0059 Emerging Mexico Fund 0.2859 0.5962 7.294 0.0106 4.8049 0.0351 2.4724 0.1249 2.4704 0.0996 1.3295 0.278 First Philippine Fund 2.1323 0.148 7.4145 0.0079 0.105 0.7467 5.3652 0.023 0 I 2.5947 0.0809 India Growth Fund 4.9202 0.0288 5.5533 0.0204 i.6001 0.20S8 6.7015 0.111 0.E641 0.4246 3.3931 0.0075 Indonesia Fund 8.7894 0.0042 12.9661 0.0006 0.1419 0.7076 10.4109 0.002 0.062 0.9399 5.1174 0.0026 Jakarta Growth Fund 6.3163 0.0147 14.6491 0.0003 2.0044 0.162 8.1646 0.0059 0.7191 0.4914 3.7213 0.0301 Korea Fund 4.5532 0.0347 40.4165 0 3.5235 0.0628 20.6325 0 1.7091 0.151 10.1971 0.0001 rhe Malaysia Fund 13.933 0.0003 11.9677 0.0007 2.1539 0.1447 2.9535 0.0881 0.8545 0.4279 1.2498 0.2901 Mexico Equity and Incomne 1.9158 0.1736 5.7449 0.0597 0.11.1 0.734 0.7281 0.3983 0.0471 0.954 0.3453 0.71 The Mexico Fund 11.5277 0.001 23.0433 0 12.4,,, 0.0006 26.6071 0 1.2047 0.1698 8.277E 0.0003 Portugal Fund Inc. 5.2559 0.0244 17.5971 0.0001 3.6799 0.0585 7.71S6 0.0068 0.9673 0.3844 2.9227 0.0594 ROC Taiwan Fund 14.0325 0.0003 56.3208 0 0.0041 0.9491 23.9095 0 0.0023 0.9975 11.8446 0 Singapore Fund 3.2389 0.0878 0.6768 0.4209 3.2141 0.0889 0.0876 0.7705 1.5886 0.2315 0.0923 0.5069 Twain Fund 7.9985 0.0056 11.8249 0.0009 0.1606 0.6899 4.0281 0.0474 0.059 0.9427 1.9729 0.1444 o The Thai Fund 8.4961 0.0052 5.7661 0.0198 0.5626 0.4565 0.1934 0.6619 0.2749 0.7607 O.J937 0.9107 Thai Capital Fund 8.6626 0.0039 18.2093 0 0.8874 0.341 5.9928 0.0157 0.289 0.7495 2.8156 0.0636 Tukish Investment Fund 0.062 0.804 58.7928 0 0.7162 0.4 27.7017 0 0.5574 u.575 13.9424 Personal Significance (At 5% Level) 72 22% = 88.89 -_ =1 n .2 0.00% = 50.0% TABLE 1 Ic: Relative Importance Or Domestic U.S. and Global Fund Factors -Developed __ _ F signliicance Ievels ror Various Hypotheses Tested DEvelope l Markets 41 #IF #2 02F #3 #3 F 41F 14 P 45 55 F #6 16 F Calculated Signif icance Calculated Significance Calculaled Signif cance Calculated Suaiicap e Calad Sigificae Cakulated Signiface F-Value Lecl F-Value Level F-Value Level F- Value Level F-Valre Leel F-Value Level Austria Fwud 90412 0.0034 4.6539 0.0037 0 1 1.041 0.3104 0.0089 0.9911 0.5237 0.5942 Emerging Germany 2.763 0.1015 58.6264 0 1.9329 0.1694 9.1766 0.0036 0.454 0.637 3.9615 0.0241 France Growth Fund 2.6857 0.1069 17.04 0.0001 0.0046 0.9462 3.6084 0.0626 0.0012 0.9983 1.7709 O.i 97 First Australia Fund 2.9383 0.08S9 8.6622 0.0039 0.3739 0.342 4.766 0.0309 0.3144 0.7301 2.4976 0.0363 Future Gennany Fund 7.0494 0.0099 35.0906 0 4.5549 0.0365 0.2631 0.6097 2.1631 0.1232 0.0546 0.S569 GcamanFund 2.6443 0.1064 21.7097 0 4.5852 0.0342 0.1179 0.7319 2.372 0.0974 0.1526 0.8586 Growth Fund of Spain 2.8443 0.0963 41.0398 0 1.2488 0.2677 16.5723 0.0001 0.1563 0.3356 7.603S 0.0011 kish Investnment Fund 14.2924 0.0004 3.2894 0.0746 3.S746 0.0535 0.4554 0.6948 1.8545 0.1652 0.0278 0.9726 Italy Fund 5.8462 0-0174 13.9319 0.0003 10.7775 0.0014 19.4545 0 1.3292 0.2217 5.5293 0.0053 Japan OTC Equity Fund 11.0222 0.0015 4.7488 0.033 1.6785 0.1998 0.0326 0.8573 0.8779 0.4207 0.0665 0.9494 New Germany Fund 3.2066 0.0776 19.4947 0 2.9799 0.0687 0.115 0.7058 1.4642 0.23S3 0.052 0.621 Spain Fund 20.12 0 2.0681 0.1529 4.492 0.036 0.1431 0.0002 2.6495 0.746 0.4783 0.007 Swiss Helvetia Fund 114.0255 0.0003 24.9088 0 8.7128 0.0038 14.557 0 2.3439 0.0996 5.163 0 Unitet Kingdom Fund 14.0595 0.0003 44.61% 0 15.104 0.0002 42.261 0 3.2419 0.0424 15.9012 _ Percent Significant (At 5% Level) 57. 14% 85.71% 42.86%, _42.86% = 7.14% = 35.71% TABLE IId: Relative mportance of Domestic U.S. and Global Fund Facton -Emerzjng | _ _ ._______ F Significance Level for Hypoth Tested Emeo ing Maarets #1 #IF #2 b2F 13 13F 04F #4 F I5 S5F 16 116F __________ Calculaled Significance Calculated Significance Calculated Significance CAluated Sigficaa Calulated ag1Sifieme Calculated Signfiasnce F- Value Level F-Value Level F-Value Value Level Value F-Value Level F-Value Lecl Brazil Fund 5.9215 0.0163 12.8648 0.0005 1.2956 0.2572 9.2666 0.0028 0.7573 0.4802 4.6919 0.010I ChileFund 15.9348 0.0001 9.4936 0.002S 4.2352 0.0426 9.255S 0.0031 2.1115 0.1272 4.5942 0.0127 Emerging Mexico Fund 0 285592 0.592 7.294 0.0106 3.5411 0.0682 2.0537 0.1607 1.8236 0.1769 10.971 0.3454 First Philippine Fund 2.1323 0.148 7.4145 0.0079 0.1692 0.6819 4.U213 0.0309 0.0106 0.9f95 2.3044 0.1063 India Crowth Fund 4.9202 0.0288 5.5533 0.0204 1.0784 0.3015 7.005 0.0094 0.5499 0.5737 3.4348 0.0344 Indonesia Fund 8.7894 0.0042 12.9661 0.0006 0.0267 0.6707 6.6441 0.0045 0.0016 0.99B4 4.2436 0.0185 Jakarta Growth Find 6.3163 0.0147 14.6491 0.0003 1.2374 0.2704 5.S191 0.0189 0.3512 0.7053 2.5647 0.0839 Korea Fund 4.5532 0.0347 40.4165 0 3.0901 0.0812 13.60S6 0 1.3891 0.2531 6.5953 0.0019 lhe Malaysia Fund 13.933 0 0003 11 .967. 0007 3.1419 0.0787 1.5878 0.21 1.3555 0.2616 0.5869 0.5576 Mexico Equity and Income 19158 0.1736 5.7449 0.0597 0.0035 0.8519 0.1452 0.7051 0.0162 0.9839 0.0698 0.9327 The Mexico Fund 11.5277 0.001 23 0433 0 8.0555 0.0035 21.4213 0 1.0554 0.3519 7.3127 0.0011 Portugal Fund Inc. 5.2559 0.0244 17.5971 0.0001 4.4487 0.0668 7.2127 0.0087 0.3766 0.4201 2.7 0.0732 ROC Taiwan Fund 14.0325 0.0003 56.3208 0 0 1 16.1106 0.0001 0.0117 0.9884 7.9942 0.0006 Singapore Fund 3 2389 0.0878 0.676S 0.4209 2.8178 0.10% 0.001 0.9751 1.4261 0.2661 0.08 0.9234 Twain Fund 7.9985 0.0056 11.8249 0.0009 0.0875 0.768 1.5353 0.2182 0.0347 0.9659 0.75131 0.474 The Thai Fund 8.4961 0.0052 5.7661 0.0198 1.1605 0.2821 0.05OS 0.3222 0.5362 056 0.0316 0.9687 Thai Capital Fund 8.6626 0.0039 18.2093 0 1.7897 0.1834 3.9477 0.0491 0.7193 0.4891 1.787 0.1717 Turkish Investment Fund 0 062 0.804 58.7928 0 0.3287 0.3681 29.6104 0 0.306 0.7373 14.3146 0 Personal Sioficance (At 5% Level) 72.22%1 88.8h _ 11.11% 66.67% o.om0_ _ 44.44% ). -,- - 0->  0 2 c iii? e -g 2 gg 5 5  -' - tO  tA K) K) K) - - a _  ,-  - - -  .' - - - - a - S  t tA C' C - 0 K) - - - .') K) _ tO __ - - __ K) - - - 00 -- - -. ** 0 -_ tA  -0 - K) tA 00 00 00   K) YQQtA-Q … Table 13 AVERAGE WEEKLY PREMIUMS (DISCOUNTS) MEAN AND STANDARD DEVIATION IN PERCENTAGE Country Fund Since Inception (a) Period 89/90 Year 1989 Year 1990 Name Mean Std D Mean Std D Mean Std D Mean Std D 1 Brazil -23.51 20.73 -23.90 23.52 -42.95 6.35 -4.84 18.35 2 Chile 2.26 14.81 b 10.68 9.89 c -0.31 14.84 3 Emerging Mexico -15.52 3.15 b -15.52 3.15 c 4 First Philippine -8.00 22.20 b 26.98 16.98 c -13.30 17.93 S India Growth -2.80 20.69 0.60 20.62 -5.23 18.43 6.42 21.20 6 Indonesia -0.64 11.45 b -0.64 11.45 c 7 Jakarta Growth -4.74 16.52 b -4.74 16.52 c 8 Korea 60.52 34.15 65.47 31.02 91.29 14.93 39.65 19.01 9 Malaysia -0.13 20.52 3.58 20.17 -4.86 15.57 12.01 20.84 10 Mexico Equity/Income -12.44 9.71 b -12.44 9.71 c 11 Mexico -7.26 35.80 -10.11 9.54 -16.09 7.15 -4.13 7.73 12 Portugal -0.77 17.49 b 18.33 2.27 c -4.03 16.89 13 R.O.C. Taiwan -4.99 13.26 b 1.82 14.18 c -9.60 10.52 14 Singapore -15.71 9.26 b -15.71 9.26 c 15 Taiwan 37.45 46.41 14.18 16.71 7.04 13.76 21.33 16.44 16 Thai Capital -8.65 10.77 b -8.65 10.77 c 17 Thai 25.83 19.80 23.85 20.55 27.90 12.94 19.79 25.53 18 Turkish Investment -20.52 7.83 b -10.81 5.09 c -21.86 6.12 19 Austria 2.98 28.08 b 22.98 17.33 c -2.70 28.23 20 Emerging Germany -14.70 6.23 b -14.70 6.23 c 21 First Australia -14.28 9.46 -14.82 9.34 -19.06 5.87 -10.58 10.25 22 France Growth -12.61 12.47 b -12.61 12.47 c 23 Future Germany -13.03 6.59 b -13.03 6.59 c 24 Germany 1-77 17.65 8.86 23.87 -1.04 18.67 18.76 24.55 25 Growth F. Spain -15.50 8.54 b -15.50 8.54 c 26 Irish Investment -19.43 9.14 b -19.43 9.14 c 27 Italy -11.97 1,3.46 -8.64 14.27 -11.57 9.67 -5.71 17.33 28 Japan OTC Equity 0.93 19.55 b 0.93 19.55 c 29 New Germany -2.09 22.46 b -2.09 22.46 c 30 Spain 20.81 42.77 28.24 45.31 30.38 53.72 26.10 35.37 31 Swiss Helvetia -7.53 7.41 -5.36 7.44 -8.45 6.70 -2.27 6.88 32 United Kingdom -16.42 5.23 -14.77 3.61 -16.07 2.71 -13.47 3.95 Notes: (a) If fund is incepted before year 1989. Ib) Because inception occurs during year 1989 or 1990, par"meter is estimated for less than 104 weeks. (c) Because inception occurs during year 1989 or 1990, pa.ameter is estimated for less than 52 weeks. Table 14 AVERAGE WEEKLY PREMIUMS (DISCOUNTS): MEAN IN PERCENTAGE - COEFFICIENT OF VARIATION IN ABSOLUTE TERMS Country Fund Since Inception (a) Period 89/90 Year 1989 Year 1990 Name Mean IC.V.i Mean IC.V.I Mean IC.V.I Mean IC.V.I 1 Brazil -23.51 88.16 -23.90 98.42 -42.95 14.78 -4.84 378.75 2 Chile 2.26 655.89 b 10.68 92.54 c -0.31 4754.65 3 Emerging Mexico -15.52 20.29 b -15.52 20.29 c 4 First Philippine -8.00 277.37 b 26.98 62.94 c -13.30 134.80 5 India Growth -2.80 737.64 0.60 3454.33 -5.23 352.40 6.42 330.01 6 Indonesia -0.64 1799.97 b -0.64 1799.97 c 7 Jakarta Growth -4.74 348.33 b -4.74 348.33 c 8 Korea 60.52 56.42 65.47 47.38 91.29 16.36 39.65 47.93 9 Malaysia -0.13 16183.40 3.58 564.03 -4.86 320.52 12.01 173.54 10 Mexico Equity/Income -12.44 78.07 b -12.44 78.07 c 11 Mexico -7.26 493 .03 -10.11 94.32 -16.09 44.j2 -4.13 187.01 12 Portugal -0.77 2272.15 b 18.33 12.41 c -4.03 419.11 13 R.O.C. Taiwan -4.99 265.78 b 1.82 778.23 c -9.60 109.50 14 Singapore -15.71 58.93 b -15.71 58.93 c 15 Taiwan 37.45 123.92 14.18 117.78 7.04 195.48 21.33 77.07 16 Thai Capital -6.65 124.45 b -8.65 124.45 c 4>- 17 Thai 25.83 76.65 23.85 86.17 27.90 46.37 19.79 129.01 18 Turkish Investment _____________ -20.52 38.18 b -10.81 47.04 c -21.86 28.00 19 Austria 2.98 941.36 b 22.98 75.41 c -2.70 1044.81 20 Emerging Germany -14.70 42.38 b -14.70 42.38 c 21 First Australia -14.28 66 25 -14.82 63.01 -19.06 30.76 -10.58 96.90 22 France Growth -12.61 98.89 b -12.61 98.89 c 23 Future Germany -13.03 50.55 b -13.03 50.55 c 24 Germany 1.77 997.73 8.86 269.52 -1.04 1790.89 18.76 130.88 25 Growth F. Spain -15.50 55.09 b -15.50 55.09 c 26 Irish Investment -19.43 47.04 b -19.43 47.04 c 27 Italy -11.97 112.42 -8.64 165.08 -11.57 83.53 -5.71 303.20 28 Japan OTC Equity 0.93 2100.31 b 0.93 2100.31 c 29 New Germany -2.09 1074.94 b -2.09 1074.94 c 30 Spain 20.81 205.51 28.24 160.45 30.38 176.82 26.10 135.53 31 Swiss Helvetia -7.53 98.42 -5.36 138.87 -8.45 79.33 -2.27 303.76 32 United Kingdom -16.42 31.84 -14.77 24.45 -16.07 16.85 -13.47 29.30 Notes: (a) It fund is incepted before year 1989. (b) Because inception occurs during year 1989 or 1990, parameter is estimated for less than 104 weeks. (c Because inception occurs during year 1989 or 1990, parameter is estimated for less than 52 weeks. TABLE 15: Time Series Repression Resdts Paramdter E.mfatets Fund Sparnsn Substtuion Global P/ID Pi t(P) 0P2 t(P2) P 3 t(P3) A Emerging Markets Brazil -0.0558 -0.4405 -0.2806 -2.7226* 0.3945 1.6701 0.8610 Chile 0.0912 1.2076 -0.0487 -1.6865 0.6281 2.9520* 0.8636 First PhWippines -0.0017228 -0.0546 -0.0026662 -1.8884 0.9319 8.4187* 0.7742 India 0.0067448 0.0767 0.0099327 0.3198 0.3197 2.2125* 0.9209 Indonesia -49992 -0.0530 0.0256 0.5149 1.3220 4.0041* 0.7416 Jakarta -0.0725 -0.7977 -49413 -0.1469 1.3074 3.9889* 0.7237 Korea -0.0572 -0.5405 0.0352 0.7832 1.3639 8.4313* 0.9598 Malaysia 0.1046 1.1134 -0.0230 -0.8544 1.2385 8.0456* 0.8942 Mexico 0.1041 2.0773* -0.0001965 -0.0087619 0.4856 4.7911* 0.8109 Portugal -0.0028923 -0.0279 0.0090882 0.5350 1.2434 7.0208* 0.7932 ROC Taiwan 0.1503 1.5075 -0.0654 -1.2731 0.8757 6.0904* 0.8128 Taiwan 0.0752 0.5856 0.0961 2.0567* 1.0403 3.8131* 0.7059 Thai 0.1851 2.1963* 0.0487 2.0603* 1.7055 6.1021* 0.6922 Thai -0.0142 -0.1645 -0.0233 -1.1701 1.3889 8.3366* 0.8586 Turkish 0.0763 0.6248 0.0823 0.4404 0.5680 1.6167 0.7476 Developed Markets Austria 0.0698 0.9427 0.0242 1.1043 0.9581 4.2934* 0.5087 Emerg. -0.1297 -1.6814 0.0268 0.4173 0.7659 4.0287* 0.6315 Germany France -0.0542 -0.6035 0.0082915 1.3214 0.5286 3.5312* 0.2960 Fst. Austr. 0.0506 1.0262 0.0279 2.5850* 0.3691 4.1035* 0.8140 Future Germ. -0.0268 -0.3199 -0.0078742 -0.1260 1.0731 6.7805* 0,7252 Germany 0.0421 0.3004 -0.0353 -1.3793 2.1302 10.2547* 0.8646 Gr, Spain -0.2439 -.53164* -0.1057 -2.5332* 0.6194 4.5617* 0.8930 Irishlnv. 0.0199 0.4659 -0.1017 -9.2316* 0.2028 2.0641* 0.8658 Italy 0.0093571 0.1467 0.0124 0.7054 0.4375 3.5173* 0.8475 Japan OTC -0.1120 -0.7581 0.0337 0.6174 2.4663 6.7133* 0.9049 Ne Germ. 0.0568 1.2281 -0.0457 -4.5929* 0.7089 5.5338* 0.5951 Spain 0.2667 1.6036 0.0606 1.9284 1.8803 7.4391* 0.9397 Swiss Hel. 0.1205 3.3535* 0.0553 3.6178* 0.3561 6.0165* 0.8114 U.K. 0.0087840 0.2974 0.0381 4.1510* 0.1597 4.4488* 0.4839 46 ,figure1 LAUNCHING OF NEW () COUNTRY FUNDS IN HEW YORK (-) Now lsuoe of existng funds ar considemd as launchirng of naw funds 3,000.0 Yew ErnwgIng Doveiopod MuPti-Country Total 1081 109.1 109.1 2.500.0 1983 25.5 25.5 1984 80.0 00.0 1986 0.0 60.0 60.0 1986 67.0 151.0 60 0 270.0 2,000.0 1987 87.0 166.0 286.5 54.6 1988 349.8 120.0 235.0 704.6 1989 777.0 136.0 0.0 912.0 * MulftiCounty P 1 l990 _422.1 1,382.51 896.7 2.701.3 3 1.500.0 Total 1,897,5 2,016.5 1,477.2 5,391.2 1 Doveb 52 ~~~~~~~~~~~~~~~~~~~~~~~~~~~C Emergoin 1,000.0 500.0 1981 1983 1984 1985 1988 1987 1988 1989 1990 CAPITALIZATION OF FUNDS IN NEW YORK End of 1990 Multi-Country 27% Market Valuo 1990 _ Emergingi _ US$ Mil1lone 38% Emefgingo 1,861.9 _ _ ~~~~~~ ~~Develped 1.402.3 \ u_ti-Country 1,141.4 Total 4,295.8 Devbloped 35% 47 Figure 2 EFFICIENT FRONTIERS FOR PORTFOLIOS OF SELECTED LOCAL STOCK MARKETS 0.009 _ A 0.008 _ Period: 71/89 - 6/28/91 ') Emerging Market Indices (7) 0.007 +) Developed Market Indices (6) (A Market Inidices. Energing and Developed (13) ( Standard & Poors 500 Index 0.006 E 0.005 o 0.004 0.0./ ^ 00 00 0.001 0.000 0.014 0.016 0.018 0Q020 0.022 0.024 0.026 0.028 0.030 0.032 0.034 Portfolio Risk Note: 277 Observattons d mIssing values or were out ol range. sadxW63647& FIgue 3 EFFICIENT FRONTIERS FOR PORTFOLIOS OF SELECTED FUND (PRICES) - 0.011 _ 0.010 Peiod: 711/89 - 6 2&91 ) Emewgng atw Ftmds (8) (,) Oe ed Markot Funds (6) 0.009 A Er nefAya Devebad Stock Market Funds (14) IS~ ta r &Poofs 500 0.003 _ 0.007 _ E a: 0.006- 0- 0.005 -r- 0,003 0.002 - 0.001 _ 0.000 0.017 0.020 0.023 0.026 0.029 0.032 0.035 0.038 0.041 Portolo Risk Note: 306 Observalnle had nisshng vahis or were out of range. s5*dsV3647b Policy Research Working Paper Series Contact Ttle Author Date for paper WPS1180 The Financing and Taxation of U.S. Harry Hukinga September 1993 R. Vo Direct Investment Abroad 33722 WPS1181 Reforming Health Care: A Case for Zeljko Bogetic September 1993 F. Smhh Stay-Well Health Insurance Dennis Heff ley 36G72 WPS1 182 Corporate Governance In Central Cheryl W. Gray September 1993 M. Berg and Eastern Europe: Lessons from Rebecca J. Hanson S1450 Advanced Market Economies WPS1 183 Who Would Vote for Inflation In Cheikh Kane September 1993 T. Hollentelle Brazil? An Integrated Framework Jacques Morisett 30968 Approach to Inflation and Income Distribution WPS1 184 Providing Social Benefits in Russia: Simon Commander September 1993 0. del Cid Redefining the Roles of Firms and Richard Jackman 35195 and Government WPSI 185 Reforming Hungarian Agricultural Morris E. Morkre September 1993 N. Artis Trade Policy: A Quantitative David G. Tarr 38004 Evaluation WPS1 186 Recent Estimates of Capital Flight Stijn C aessens September 1993 R. Vo David Naud6 33722 WPS1 187 How Should Sovereign Debtors Andrew Warner September 1993 J. Queen Restructure Their Debts? Fixed 33740 Interest Rates, Flexible Interest Rates, or Inflation-indexed WPS1 188 Developmentalism, Socialism, and Mario Marcel September 1993 S. Florez Free Market Reform: Three Decades Andr6s Solimano 39075 of Income Distribution in Chile WPS1 189 Can Communist Economies Alan Gelb September 1993 PRDTM Transform Incrementally? China's Gary Jefferson 37471 Experience lnderjit Singh WPS1190 The Government's Role in Japanese Yoon Je Cho SepLember 1993 T. Ishibe and Korean Credit Markets: A New Thomas Hellmann 37665 Institutional Economics Perspectivs WPS1 191 Rent-Sharing in the Multi-Fibre Geoffrey J. Bannister September 1993 A. Daruwala Arrangement: The Case of Mexico 33713 WPS1 192 Effects of Tax Reform on Argentina's Jacques Morisset September 1993 G. Carter Revenues Alejandro lzquierdo 30603 WPS1 193 The Armenian Labor Market in Milan Vodopivec September 1993 S. Florez Transition: Issues and Options Wayne Vroman 39075 Policy Reuarch Wwrkng Paper Serles Contot Title Author Date for paper WPSI 194 How Fast Has Chinese Industry Tom RawskW September 1993 E. Khino Grown? 37471 WPS1195 The Enterprise Sector and Mark Schaffer September 1993 E. Khine Emergence of the Polish Fiscal 37471 Crisis, 1990-91 WPS1 196 Corporate Tax Structure and Jeffrey Bernstein September 1993 C. Jones Production Anwar Shah 37754 WPS1 197 Determinants of Inflation among Bruno Bocwar Sepmber 1993 0. Jon Franc Zone Countries in Africa Shantayanan Devaraian 37754 WPS1 198 Enterprise Reform in China: The Natalie Lichtenstein September 1993 M. Rangarajan Evolvng Legal Framework 81710 WPS1 199 Public Pension Govemance and Olivia Mitchell October 1993 D. Evans Performance: Lessons for Developing 37496 Countries WPS1 200 The Life-Cycle Distributional Jane Falkingham October 1993 D. Evans Consequences of Pay-As-You-Go Paul Johnson 37496 and Funded Pension Systems WPS1 201 Five Criteria for Choosing among Margaret E. Grosh October 1993 M. Quintero Povertv Programs 37792 WPS1202 Privatization and Foreign Investment Frank Sader October 1993 Rose Vo in the Developing Wold, 1988-92 33722 WPS12O3 Determinants of Value-Added Tax Zeljko Bogetic October 1993 F. Smith Revenue: A Cro-s-Section Analysis Fareed Hassan 36072 WPS1204 Structural Adjustment, Economic Nisha Agrawal October 1993 K. Rivera Peo formance, and Aid Dependency Zafar Ahmed 34141 in Tanzania Michael Mered Puger Nord WPS1205 Wagoa;nd Employment Decisions Simon Commander October 1993 0. del Cid in the Russian Economy: An Analysis Leonid Liberman 36303 of Developments in 1992 Ruslan Yemtsov WPS1206 Empirical Perspec1ves on National Ishac Diwan October 1993 A. Yideru Index Funds Vihang Errunza 36067 Lemma W. Senbet WPS1207 Characteristics and Performance Bill H. Kinsey October 1993 'inswanger of Settlement Programs: A Review Hans P. Binswanger P,4 r... ...