Contractual savings or stock market development: which leads?

Catalan, Mario; Impavido, Gregorio; Musalem, Alberto R. Article Contractual savings or stock market development: which leads? Schmollers Jahrbuch – Zeitschrift für Wirtschaftsund Sozialwissenschaften. Journal of Applied Social Science Studies Provided in Cooperation with: Duncker & Humblot, Berlin Suggested Citation: Catalan, Mario; Impavido, Gregorio; Musalem, Alberto R. (2000) : Contractual savings or stock market development: which leads?, Schmollers Jahrbuch – Zeitschrift für Wirtschaftsund Sozialwissenschaften. Journal of Applied Social Science Studies, ISSN 1865-5742, Duncker & Humblot, Berlin, Vol. 120, Iss. 3, pp. 445-487, https://doi.org/10.3790/schm.120.3.445 This Version is available at: https://hdl.handle.net/10419/291965 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. https://creativecommons.org/licenses/by/4.0/ Schmollers Jahrbuch 120 (2000), 445-487 Duncker & Humblot, Berlin Contractual savings or stock market development: which leads? By Mario Catalan, Gregorio Impavido and Alberto R. Musalem Abstract This paper studies the relationship between the development of contractual savings (assets of pension funds and life insurance companies) and capital markets. The focus is on the macroeconomic and financial effects of contractual savings' development. New theoretical ideas and empirical results are presented. At the theoretical level, we explain how the growth of the contractual savings sector promotes financial development and economic growth through different channels. We argue that among institutional investors, contractual savings institutions are the most effective at developing capital markets. What is different about contractual savings is that their liabilities are long-term and illiquid ones in asset holders' portfolios. At the empirical level, we analyze Granger causality between contractual savings and both market capitalization and value traded in stock markets for some OECD and other countries. The evidence suggests that the growth of contractual savings cause the development of capital markets. Zusammenfassung Der Aufsatz untersucht den Zusammenhang zwischen planmäßigem Sparen (in Pensionsfonds und Lebensversicherungen) und dem Kapitalmarkt sowohl theoretisch wie empirisch. Auf der theoretischen Ebene wird gezeigt wie planmäßiges Sparen die Entwicklung des Kapitalmarkts und des Wachstums über verschiedene Kanäle beeinflußt. Es wird argumentiert, dass im Bereich institutioneller Anleger planmäßiges Sparen besonders effektiv ist, um den Kapitalmarkt weiter zu entwickeln, da es sich um langfristige Anlagen und illiquides Vermögen im Portfolio der Sparer handelt. Auf der empirischen Ebene wird untersucht, ob die Weiterentwicklung der Kapitalmärkte zu mehr planmäßigem Sparen führt oder ob planmäßiges Sparen die Weiterentwicklung der Kapitalmärkte anstößt. Die empirische Evidenz weist darauf hin, dass die Kausalrichtung vom planmäßigem Sparen hin zur Weiterentwicklung der Kapitalmärkte weist. JEL-Classification: G 23, G 10 Schmollers Jahrbuch 120 (2000) 3 OPEN ACCESS | Licensed under CC BY 4.0 | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/schm.120.3.445 | Generated on 2023-04-04 12:27:53 446 Mario Catalan, Gregorio Impavido and Alberto R. Musalem 1. Introduction In the last two decades, there has been a dramatic growth in the assets managed by contractual saving institutions (pension funds and life insurance companies) in developed countries as well as in some developing countries as shown in Table 1. In most countries in the sample, contractual savings share to GDP (deepening) increased several fold during the period. Furthermore, Netherlands, United Kingdom, Switzerland, and South Africa had contractual savings in excess of 100 percent of GDP in 1996. The only country in the sample that experienced a decline in the participation of contractual savings in GDP is Singapore. Pension reform favoring funding is considered to be one of the policy options that policy-makers face when attempting to develop the contractual savings sector, especially in developing countries. As evidence of the general interest on contractual savings development and its potential effects in the economy, extensive literature on the macroeconomic role of pension funds has been developed and the debate on the benefits of pension reforms has been enriched and intensified in recent years.1 Many studies focused on the effect of pension reforms on household saving rate and results are not conclusive. On the one hand, pension reform that relies on voluntary contributions based on expenditure tax treatment as opposed to income tax treatment is expected to have a negligible effect on saving as indicated by the extensive literature available on the inelasticity of saving to the real interest rate.2 On the other hand, either myopia or liquidity constraints explain why pension reforms based on mandatory contributions could increase the household saving rate. The liquidity constraints are assumed to affect young or low-income individuals who cannot borrow to consume and offset the compulsory saving.3 However, the effect on national saving will also depend on the government and firms response to pension reform. Even if the effect of a pension reform on the national savings rate were not significant, other effects could be important. In particular, capital markets development is indicated as one of the main potential consequences of contractual savings development.4 1 See, for example, Holzmann (1997), Arrau and Schmidt-Hebbel (1993), Feldstein (1974, 1996), Mackenzie, Gerson and Cuevas (1997), Schmidt-Hebbel (1998). 2 See for example, Whitehouse (1999). 3 See, for example, Feldstein (1978), Munnell (1976), Loayza, Schmidt-Hebbel and Serven (2000), Samwick (2000), Smith (1990), Bailliu and Reisen (1997), SchmidtHebbel and Serven, eds. (1999). 4 See, for example, Bodie (1990), Davis (1995), Vittas and Skully (1991), Vittas (1998a, 1998b, 1999). Schmollers Jahrbuch 120 (2000) 3 OPEN ACCESS | Licensed under CC BY 4.0 | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/schm.120.3.445 | Generated on 2023-04-04 12:27:53 Contractual savings or stock market development 447 This study is part of a larger research project that encompasses various contractual savings and financial sector issues. The purpose of this paper is to analyze the causality between contractual savings and stock markets development. We emphasize the role of pension funds and life insurance companies as financial intermediaries, and we compare results when different institutions, like non-life insurance companies are considered. The literature is not clear on its assumption regarding causality between contractual savings and capital market development. A one-way or a two-way relationship is assumed, usually interchangeably. In this paper, we address the question of which relationship leads empirically. The evidence, including descriptive statistics .as well as Granger causality tests is presented for OECD countries and some other countries such as Chile, Malaysia, Singapore, South Africa, and Thailand. The paper does not present a theoretical framework but explains with clear statements and intuitive examples the way in which we think the growth of the contractual savings sector promotes financial development. Table 1 Contractual savings ratio to GDP (percent) Countries 1980 1985 1990 1996 Netherlands 66.90 93.65 108.11 148.19 United Kingdom 38.81 74.77 86.90 141.72 Switzerland 70.00 88.5 131.38 United States 43.01 59.33 69.20 94.80 Canada 30.29 38.08 47.80 64.59 Australia 33.49 57.52 Sweden 23.92 28.63 47.96 Norway 13.15 17.29 25.80 30.02 Belgium 16.42 20.55 27.20 Korea, Rep. 4.06 10.48 19.24 24.36 Germany 12.73 17.63 20.68 23.82 Austria 13.28 21.35 Spain 3.21 9.87 18.78 South Africa 39.27 55.93 78.13 126.01 Singapore 153.36 115.13 93.50 Chile 1.00 29.28 50.61 Malaysia 20.08 35.65 47.18 51.02 Thailand 2.10 4.80 Source: 1998 OECD Institutional Investors Statistical Yearbook and WB institutional investors database. The paper is organized as follows. Section 2 presents the key propositions on the links between contractual savings and capital markets development. Schmollers Jahrbuch 120 (2000) 3 OPEN ACCESS | Licensed under CC BY 4.0 | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/schm.120.3.445 | Generated on 2023-04-04 12:27:53 448 Mario Catalan, Gregorio Impavido and Alberto R. Musalem Their effects and implications for the economy as a whole are analyzed in terms of growth, term structure of interest rates, capital structure, regulation, and comparative impact on developing versus developed economies. Section 3 discusses the role of contractual savings in the structure of the financial sector. In particular, it distinguishes between the effects of contractual savings, mutual funds and non-life insurance development. Section 4 presents a descriptive analysis of the data, which confirms that there is a positive relation between contractual savings and capital markets development. Section 5 analyzes the causality between contractual savings and non-life insurance companies and market capitalization or value traded in stock markets.5 Finally, section 6 summarizes the results and the main conclusions. 2. What is different about contractual savings? The key point to understanding the macroeconomic role of contractual savings and more specifically, their role as financial intermediaries, is to observe that they have a distinctive characteristic. While banks and open-end mutual funds have mainly short-term liabilities,6 some contractual savings institutions such as life insurance and close-end pension funds (i.e., employers sponsored pension plans) have long-term liabilities on their balance sheets. Although, an open-end pension fund system (i.e., individual accounts in defined contribution schemes) operate like open-end mutual funds, however, their funds are more stable because they are captive to the industry as a whole. Hence, open-end pension funds are less exposed to systemic risks than are open-end mutual funds. This distinction has important implications. It means that the depositors or investors cannot "run" (withdraw their deposits suddenly and in a large scale) against the assets of the contractual savings institutions where they have claims. In contrast, banks, open-end mutual funds, and to a lesser extent, open-end pension plans face the risk of an unexpected run against their assets that could generate a liquidity problem, and potentially trigger their bankruptcy As a consequence, the investment and lending strategies of banks, open-end mutual funds, and to a certain extent, open-end pension funds differ from those of other contractual savings institutions. Contractual savings institutions have a natural advantage over banks in financing long-term investment projects and their investment strategies will be more biased towards long-term 5 Market capitalization (also known as market value) is the share price times the number of shares outstanding. Stocks traded refers to the total value of shares traded during a given period. 6 Strictu sensu, mutual funds and open-end pension funds do not have liabilities since funds belong to plan members and not to the fund managers. Schmollers Jahrbuch 120 (2000) 3 OPEN ACCESS | Licensed under CC BY 4.0 | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/schm.120.3.445 | Generated on 2023-04-04 12:27:53 Contractual savings or stock market development 449 bonds and the equity markets. Needless to say, banks may still be able to finance long term projects while minimizing term transformation risks by financing such portfolios through the sale of long term bonds to contractual saving institutions. A dynamic hedging principle is at work, in the sense that financial institutions try to match the maturity structure of their assets and liabilities. Hedged positions help to reduce the risks they face; conversely the lack of hedged positions imply that either reinvestment (short-term assets and long-term liabilities) or refinancing (long-term assets and short-term liabilities) decisions will have to be taken. The ensuing maturity mismatch implies risk taking and can generate cash flow problems in volatile environments. As will become clear, for a given amount of total savings in the economy, contractual savings growth (for example, a pension reform from a pay-asyou-go to a funded system, a reform that transforms corporate pensions that are based on book reserves to funded schemes outside the firm, or reforms that improve the regulatory and tax environment) are expected to stimulate financial development. This is because from the point of view of household and corporate sectors, there is an important liquidity effect at work. The accounts held in the contractual savings sector are completely illiquid from the depositor's point of view. They can only be liquidated in the long-run upon retirement of the beneficiary (either as a lump sum and / or annuity) or upon the occurrence of a particular event (e.g., death, disability); firms have no access to them. Thus, if large deposits are made in contractual savings, this will change the actual portfolio composition of both households and corporations between liquid and illiquid assets to a level below their desired ratio. Therefore, to restore equilibrium, households' and corporations' demand for liquidity has to be satisfied with additional holdings of liquid assets. This could be achieved by a reshuffling of portfolios; for instance, by increasing holdings of deposits in the banking sector, open-end mutual funds, and traded securities, at the expense of some other non-liquid assets that households or corporations could have held (e.g., real estate, nontraded financial instruments). Thus, households' and corporations' behavior will reinforce financial market development, which is associated with contractual savings growth. However, the illiquidity effect of • contractual saving instruments on wealth holders' behavior will be weakened to the extent that plan members can either borrow from the plan, sometimes using accumulated funds as collateral, or simply they can withdraw funds for specific purposes. It is important to remark that these and next propositions hold even when the total saving of the household and corporate sectors remain constant. Schmollers Jahrbuch 120 (2000) 3 OPEN ACCESS | Licensed under CC BY 4.0 | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/schm.120.3.445 | Generated on 2023-04-04 12:27:53 450 Mario Catalan, Gregorio Impávido and Alberto R. Musalem Total saving proved to be very insensitive to the variables that are supposed to affect it, so the fact that the propositions do not depend on the change in saving in the economy is remarkable. Our analysis, although different, is consistent with previous work. Davis (1995) finds that pension fund portfolios have a greater proportion of uncertain capital and long-term assets than the household sector. He also finds that the personal sector tends to hold a much larger proportion of liquid assets. "The implication is that a switch to funding would increase the supply of long-term funds to capital markets and reduce bank deposits, even if savings and wealth do not increase, so long as households do not increase the liquidity of the remainder of their portfolios fully to offset growth of pension funds". This, he explains, is the impact of a pension reform on capital markets and the existence of the liquidity effect. Davis also suggests that there is some evidence that such offsetting to restore liquidity exists. Furthermore, the growth of contractual savings implies a reallocation of savings from intermediaries with a high probability of facing a run against their assets (banks and open-end mutual funds) towards intermediaries with a low probability of facing a run (pension funds and life insurance companies). This reallocation means that funds are moved towards institutions that invest more heavily in long-term bonds and equity. In addition, of course, there could be an independent effect of the reform on total savings that would cause further financial development. As an application of the previous statements to the case of pension and life insurance reforms, it is apparent that only an increase in the amount of assets accumulated in the contractual savings sector is necessary to develop the capital markets and that an increase in total savings is not necessary at all. Therefore, pension reforms, which increase the level of funding, will imply a large increase in assets managed by pension funds and thus, a higher degree of capital market development. Of course, our hypothesis also implies that if a pay-as-you-go system were to be transformed into a partially funded scheme that would be able to accumulate assets at a sustainable pace it would also produce the same financial deepening effect. This would be the case provided reserves are invested in market instruments and are not used as captive sources of finance by governments.7 Accordingly, contractual savings development would imply a movement towards completing financial market development. Although funding generates positive externalities through capital market development, this does not mean that forcing a given level of funding 7 There is some evidence however, that governments do use partially-funded public pension schemes as sources of captive finance. For a discussion see Iglesias and Palacios (2000). Schmollers Jahrbuch 120 (2000) 3 OPEN ACCESS | Licensed under CC BY 4.0 | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/schm.120.3.445 | Generated on 2023-04-04 12:27:53 Contractual savings or stock market development 451 through mandatory retirement schemes coincides with the social optimum. In other words, there is an argument for a minimum level of mandated funding to provide a minimum level of benefits, leaving the provision of additional benefits to voluntary arrangements. This minimum funding would be sufficient to address the market failures existing in a fully voluntary scheme. These failures derive from myopia of individuals, who do not necessarily save enough for retirement needs or other contingencies (e.g., death, disability); from the moral hazard of individuals relying on Government retirement income guarantee schemes; and from the adverse selection implicit in the different life expectancy of individuals. Hence, a fully funded mandatory pension system that ensures a minimum level of benefits would maximize social welfare, whilst a mandatory PAYG system that precludes the development of stock markets would not. The design of pension reform is likely to affect social welfare through this and other channels. For instance, regulations imposed on the portfolio composition of pension funds can severely affect the quantitative impact of contractual savings development on capital markets. As an extreme example, if pension funds were restricted to hold only government bonds, the development of contractual savings should have a minimum or no effect on stock markets and social welfare would be lower. In order to understand the mechanics of capital market development and its relation to contractual savings and the economy as a whole, let us summarize the most important propositions concerning the macroeconomic role of contractual savings. Conceptually, let us think of an economy with banks as the unique financial intermediaries that is subsequently transformed into an economy with both banks and a large contractual savings sector. The main micro / macroeconomic effects are the following. 2.1 Specialization in the financial sector, the term structure of interest rates, and growth The development of the contractual savings sector will initially have a static effect where the banking sector will tend to specialize in financing investment projects with short maturity and the contractual savings institutions funding those investment projects with long maturity. Of course, portfolios will be diversified and a complete specialization will not be observed, in the sense that only the shortest-maturity projects are financed by banks and only those with the longest maturity are financed by contractual savings institutions. We would rather observe that the diversified portfolios of banks are more biased towards short-term loans and those of the contractual savings institutions are more biased towards long-term and risky assets Schmollers Jahrbuch 120 (2000) 3 OPEN ACCESS | Licensed under CC BY 4.0 | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/schm.120.3.445 | Generated on 2023-04-04 12:27:53 452 Mario Catalan, Gregorio Impavido and Alberto R. Musalem but all institutions will have all kinds of assets. Of course, the development of contractual savings will allow banks to become intermediaries between their long term borrowers and contractual saving institutions while undertaking minimum term transformation risks through the issuance of asset backed securities. Again, regulations could introduce significant distortions. If pension funds and life insurance companies are restricted to holding primarily securities, there could be an important cost associated to the contraction of the banking system. In the last two decades, some academic economists made important contributions to the understanding of the special role that banks play in the financial system.8 Banks play an important microeconomic role of monitoring. Among other peculiarities, banks finance "difficult" projects requiring intensive monitoring. These "difficult" projects cannot be financed by the issuance of securities because large numbers of small security holders have no incentive to monitor individually. Bank loans and securities are not perfect substitutes and the expansion of contractual savings can have a very important distributional impact on the economy For instance, if small firms require more monitoring, the contraction of the banking system will make the financing of those firms very expensive and there will be incentives to create corporations. This effect is exacerbated if contractual saving institutions cannot hold loans, but it could exist even if there is no constraint on portfolio holdings because the issuance of demand deposits and loans are complementary activities.9 These conclusions are sensitive to the condition of the banking sector in an economy. The introduction of a funded pension scheme in an economy where the probability of bank runs is relatively high (i.e., many emerging economies) will have more important effects than in an economy with a relatively low probability of bank runs (i.e., most developed economies). This is because in the latter case, banks would already be allocating a significant proportion of their portfolio in long-term loans. The development of contractual savings also implies that the long-term interest rate should fall relative to the short-term rate and thus, more longterm projects will be financed. Given the fact that the expected return of long-term investment projects is higher than the returns on short-term investments (a technologically reasonable assumption), a higher growth rate will be observed. 8 See Fama (1985), James and Wier (1988), and Diamond (1984). 9 See for instance, Kashyap, Rajan and Stein (1998). Schmollers Jahrbuch 120 (2000) 3 OPEN ACCESS | Licensed under CC BY 4.0 | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/schm.120.3.445 | Generated on 2023-04-04 12:27:53 Contractual savings or stock market development 459 After the investment decision is taken and before date 1, there is a run against the assets of the institution that occurs with probability P and there is no run with probability 1 - P. If the investment decision of the institution was to finance the long-term project and there is a run, then the institution will be in an illiquid position and will default on its debt, thus it will go bankrupt and will lose its reputation with a loss equal to — C = -2.12 If the long-term project was financed and there is no run, it will get (1 + iLf- (1 + id)2= 0.3375 at date 2. If the investment decision of the firm was to finance the short-term project and there is a run, the institution will be liquid and able to pay the depositor, the profit will be (l + zs)2-(l + id) = 0.05. If there is no run, the institution will reinvest for one period and at the end it will get (l + zs)2-(l + zd)2= 0.1075. The strategy to be chosen will be the one that maximizes expected profits. The institution will choose to finance the long-term project if and only if the expected profit of that strategy is greater than the expected profit of the alternative one. In our example, the following condition must be satisfied: —2P + (1 - P)0.3375 > 0.05P + (1 - P)0.1075 iff P < 0.1 Thus, the inequality holds for a value of P that is lower or equal to 0.10. In other words, the long-term project will be financed by the institution only if the probability of a run is low enough. This example is instructive in several directions. We can think of this institution as being a pension fund if P = 0 (you cannot run against the pension fund) and a bank for P greater than 0. Suppose an economy where P in the banking sector is greater than 0.1, that means that the banks will either finance the long-term project at very high interest rates or not finance it at all, while a pension fund will do it, thus, the introduction of pension funds will have a very important real effect in promoting long-term investment and growth. Now, suppose other economy where P in the banking sector is lower than 0.1, that means that the banks will choose to finance the longterm project, thus the development of the pension fund sector will not generate this type of effect. Think of the first type of economy as one without a very resilient banking sector where the probability of a bank run is not negligible, and think of the second economy as one with a strong banking sector. We can conclude that 12 This is an arbitrary number that is supposed to represent all the costs of shutting down the institution, including the cost in reputation and the present value of future profits foregone. The message of our story is insensitive to the particular number used. Schmollers Jahrbuch 120 (2000) 3 OPEN ACCESS | Licensed under CC BY 4.0 | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/schm.120.3.445 | Generated on 2023-04-04 12:27:53 460 Mario Catalan, Gregorio Impavido and Alberto R. Musalem the potential benefits of developing the contractual savings sector are greater in economies without very strong banks, at least in terms of financial deepening, the term structure of investment and growth.13 Figure 2: Payoff tree 4. Descriptive evidence Figure 3 shows how contractual savings have become the dominant financial asset in several countries. In 1996, they represented 50 percent or more of financial assets (defined as the aggregation of money, quasi-money and contractual savings assets) in 9 out of 29 countries.14 Furthermore, the same 13 Of course, in this very simple example, the institution is constrained to hold a completely specialized portfolio, but a rigorous model with portfolio diversification can be constructed and a similar parable can be told. Similarly, the basic structure can also be extended to include risky assets. 14 Money and quasi money are liabilities of the consolidated banking system (including the Central Bank), which are liquid financial assets held by the household sector. Clearly, the assets of contractual savings institutions belong to the household sector. Of course, there is some double counting since assets of contractual savings institutions include cash and bank deposits. Schmollers Jahrbuch 120 (2000) 3 OPEN ACCESS | Licensed under CC BY 4.0 | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/schm.120.3.445 | Generated on 2023-04-04 12:27:53 Contractual savings or stock market development 461 figure shows that non-OECD countries such as South Africa, Chile, Singapore, and Malaysia have a dominant or a very important contractual savings sector. Source: 1998 OECD Institutional Investors Statistical Yearbook and WB institutional investors database. Figure 3: Contractual savings in system financial assets (%, 1996) Figure 4 shows the positive correlation between the financial assets of contractual savings institutions and market capitalization as a fraction of GDP for a cross section of OECD and non-OECD countries in 1996 (the positive relation is very stable for different years). Those countries with a more developed contractual savings sector are also countries with more developed stock markets. Furthermore, Figure 5 indicates a positive relationship between contractual savings development and the liquidity of the capital markets (measured by value traded over GDP). Figure 6 explores the relationship between changes in contractual savings as a fraction of GDP and changes in market capitalization over GDP for the same countries between 1990 and 1996. Figure 7 presents a similar relationship between changes in contractual savings and changes in value traded as a fraction of GDP It is clear that those countries that were able to develop their contractual savings sector also show a higher growth in their stock markets in terms of capitalization and value traded in the same period. The Schmollers Jahrbuch 120 (2000) 3 OPEN ACCESS | Licensed under CC BY 4.0 | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/schm.120.3.445 | Generated on 2023-04-04 12:27:53 462 Mario Catalan, Gregorio Impavido and Alberto R. Musalem MC/GDP CS/GDP Notes: The fitted line is given by yt = 0.177 + 0.958xt with a t statistic of 7.314 for the slope. See Table in Appendix 2 for the list of countries. Source: 1998 OECD Institutional Investors Statistical Yearbook and WB institutional investors database. Figure 4: Contractual savings and market capitalization, 1996 MC/GDP CS/GDP Notes: The fitted line is given by yt = 0.085 + 0.480xt with a t statistic of 4.650 for the slope. See Table in Appendix 2 for the list of countries. Source: 1998 OECD Institutional Investors Statistical Yearbook and WB institutional investors database. Figure 5: Contractual savings and value traded, 1996 OPEN ACCESS | Licensed under CC BY 4.0 | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/schm.120.3.445 | Generated on 2023-04-04 12:27:53 Contractual savings or stock market development 463 same conclusions are reached with estimates using panel data for 26 countries and with about 300 observations.15 Now, let us see whether the data show that contractual savings institutions are more willing to hold risky and long-term assets than other institutional investors and banks. Figure 8 compares the portfolios of US contractual savings institutions with those of the banking sector. Among the remarkable facts in Figure 8 are the high weight of securities in the portfolios of pension funds (84 percent), life insurance companies (79 percent) and open-end investment companies (90 percent) relative to banks (23 percent), and the low weight of short-term loans and cash in the portfolios of those institutions (4 percent, 7 percent, and 8 percent respectively) relative to banks (59 percent). Pension funds, life insurance companies and open-end investment companies are also heavily invested in long-term bonds. Clearly, US contractual savings institutions hold larger fractions of their total assets invested in traded securities such as stocks and long-term bonds while the assets of the banking sector are invested more heavily in private financial instruments (loans) of short-term maturity. Finally, Figure 9 shows the average portfolio composition of different institutional investors of some other selected OECD countries. In the United Kingdom, shares and long-term bonds account for 80 percent or more of the portfolios of contractual savings institutions. There is a very high fraction of loans in the portfolios in the Netherlands, but it is also striking that they are almost completely long-term loans. We could be tempted to say that the role of contractual savings institutions in the Netherlands is similar to those of banks in terms of lending strategy, but the financial services provided are absolutely different in terms of maturity structure. In Norway, even when we do not have the maturity structure of loans, the presumption is that a similar story can be told. Sweden and Norway are also examples of our hypothesis that if there are binding restrictions to invest in shares, then longterm bonds and / or loans will be in high demand. Finally, in Australia, contractual savings institutions invest more than 50 percent of their portfolios in shares and long-term bonds; while they represent about 40 percent in other institutional investors' portfolios. Thus, according to the evidence, if there were a reallocation of assets from the banking sector to the contractual savings sector, there would be a shift in the relative demands for financial instruments. There would be a reduction in the demand for non-traded financial instruments, or in other words, we would observe a reduction in the supply of funds to be lent to firms in the non-corporate sector (i.e., firms that do not issue publicly traded stock is See Impavido and Musalem (2000). Schmollers Jahrbuch 120 (2000) 3 OPEN ACCESS | Licensed under CC BY 4.0 | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/schm.120.3.445 | Generated on 2023-04-04 12:27:53 464 Mario Catalan, Gregorio Impavido and Alberto R. Musalem D(MC/GDP) D(CS/GDP) Notes: The fitted line is given by yt = 0.161 + 0.849xt with a t statistic of 2.593 for the slope. See Table in Appendix 2 for the list of countries. Source: 1998 OECD Institutional Investors Statistical Yearbook and WB institutional investors database. Figure 6: Changes in contractual savings and market capitalization, 1990-1996 D(VT/GDP) D(CS/GDP) Notes: The fitted line is given by yt = 0.044 + 0.968xf with a t statistic of 3.289 for the slope. See Table in Appendix 2 for the list of countries. Source: 1998 OECD Institutional Investors Statistical Yearbook and WB institutional investors database. Figure 7: Changes in contractual savings and value traded, 1990-1996 Schmollers Jahrbuch 120 (2000) 3 OPEN ACCESS | Licensed under CC BY 4.0 | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/schm.120.3.445 | Generated on 2023-04-04 12:27:53 Contractual savings or stock market development 465 and debt), and there would be an increase in the demand for publicly traded financial instruments such as stocks and bonds. Moreover, the fact that the portfolio weight of long-term bonds is high for contractual savings institutions means that the corporate sector will have additional long-term funds to finance their long-term production plans. As a consequence, the profit opportunities in the corporate sector will induce the entry of new firms that will issue both equity and debt, increasing the market capitalization of the economy, and thus, the market will become more liquid and the value traded in stocks will increase. Finally, the increased volume of transactions will imply a higher demand for money (transaction motive) and overall financial deepening in the economy. ST Loans and Cash Other Other ST Loans and Cash LT Loans 1%- ST Bills2%~ and Bonds , LT Loans 9% ST Bills . and Bonds 25%* LT Bonds Shares 57% LT Bonds US-Portfolios of Open-end Investment Companies (1996) US-Portfolios of Banks (1996) Other ST Loans and Cash Other Securities LT Loans ST Bills and Bonds LT Loans 35% ^ LT Bonds ST Loans and Cash US-Portfolios of Pension Funds (1996) US-Portfolios of Life Insurance Companies (1996) Source: OECD, Institutional Investors Yearbook, 1997, and Federal Reserve, Monthly Bulletins. Figure 8: United Nations Financial Institutions portfolios Schmollers Jahrbuch 120 (2000) 3 OPEN ACCESS | Licensed under CC BY 4.0 | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/schm.120.3.445 | Generated on 2023-04-04 12:27:53 466 Mario Catalan, Gregorio Impavido and Alberto R. Musalem UK-Portfolios of Life Insurance Companies (1980-1995) Other ST Loans and Cash 7*n LT Loans 2%\ ST Bills and Bonds tj V ?-'••{)* ' I / £ LT Bonds ^^BppP^ UK-Portfolios of Pension Funds (1980-1995) ST Loans 4y ^ LTÜ3ans 0%- 9% ST Bills U^ÄSfelk. and Bonds 1 | > LT Bonds1 I I / 70% ^¡M^^ Shares UK-Portfolios of Open-end Investment Companies (198^1995) Other ST Loans and Cash 5%| ST Bills and Bonds 1 %\ | LT Bonds ijg^w CD Netherlands-Portfolios of Life Insurance Companies (1980-1995) Shares 10% Other 16% ST Loans 11 o/0LT Bonds and Cash ST Bills and Bonds © Netherlands-Portfolios of Pension Funds (1980-1995) Other 10%i Shares 12% ST Loans 1%^jjlfl8H -(i LT Bonds STBills / — LTLoans Netherlands-Portfolios of Investment Companies (1980-1995) a her iilgWiJ^ii. ov STLoans and Cash LTLoans 3i STBills and Bonds 24% LT Bonds Source: OECD, Institutional Investors Yearbook, 1997. Figure 9: Institutional investors' portfolio Schmollers Jahrbuch 120 (2000) 3 OPEN ACCESS | Licensed under CC BY 4.0 | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/schm.120.3.445 | Generated on 2023-04-04 12:27:53 Contractual savings or stock market development 467 Australia -Life Insurance Companies (1988-1995) Other Total Loans 2% and Cash 14%, /< I \ 14 J X \ •MBKMrmx Shares STBills 1 ^FESTEN and Bonds \ / \y 26% - ¿V LTBonds Australia-Portfolios of Pension Funds (1988-1995) Other 9% V 25%Shares 21Y Total Loans / and Cash I 12% 21Y Total Loans / and Cash I 12% A ST Bills ^^ 33% and Bonds LT Bonds Australia-Open-end Investment Companies (1988-1995) ST Bills 37% and Bonds Sweden-Portfolios of Life Insurance Companies (199 0-1995) Total Loans Other and Ca»tj1% 0% STBills 6%X^' ""Bfe^ 29% Shares and Bonds Ilk i Lj 54% W LTBonds Sweden-Portfolios of Pension Funds (1990-1995) Total Loans other shares andCash^ 7% t% Q% ST Bills and Bonds Sweden-Portfolios of Open-end Investment Companies (1990-1995) Total Loans and Cash 6%^ STBills and Bonds Other Shares 78% LT Bonds Source: OECD, Institutional Investors Yearbook, 1997. Continued Figure 9: Institutional investors' portfolio Schmollers Jahrbuch 120 (2000) 3 OPEN ACCESS | Licensed under CC BY 4.0 | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/schm.120.3.445 | Generated on 2023-04-04 12:27:53 468 Mario Catalan, Gregorio Impavido and Alberto R. Musalem Norway-Portfolios of Life Insurance Companies (1980-1995) Other Shares 4% /"l 6% Total Loans 1 1 m and Cash 46°/J V ^^^m LT Bonds W^ 3% ST Bills and Bonds Norway-Portfolios of Pension Funds (1980-1995) Shares ®tlier3% 5% 46€ Total Loans \ ^H^^F LT Bonds and Cash VJ^ \—LJP^ 1% ST Bills and Bonds Norway-Portfolios of Open-end Investment Companies (1980-1995) a her 36% LT Bonds Source: OECD, Institutional Investors Yearbook, 1997. Continued Figure 9: Institutional investors' portfolio The international evidence suggests some stylized facts about contractual savings institutions. The fraction of investment in either shares or long-term assets (either bonds or loans) tends to be very high. In all the cases, the weight of short-term loans is very low. Obviously, regulations, relative yields, risk and liquidity preferences, and tax treatment could explain the differences in portfolios across these countries. The evidence also suggests that if binding constraints are imposed on the fraction invested in shares, they will try to invest their funds in the closer substitutes such as long-term bonds and long-term loans. The result will be a differential impact on the productive sector of the economy and on the structure of the financial sector. 5. Econometric evidence on contractual savings and capital markets development: which leads? This paper has emphasized the direction of causality from contractual savings to market capitalization. In Sections 2 and 3, we argued that if contractual savings are developed then market capitalization would follow. In Schmollers Jahrbuch 120 (2000) 3 OPEN ACCESS | Licensed under CC BY 4.0 | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/schm.120.3.445 | Generated on 2023-04-04 12:27:53 Contractual savings or stock market development 475 institutions (contractual savings, pension funds, life insurance, non-life insurance) to one of the market indicators (market capitalization or value traded); the second column reports the number of countries for which causality runs only from one of the markets to one of the institutions (<—); the third column reports the number of countries for which causality runs both ways (<->); and the fourth column reports the number of countries for which no causality was found in either direction (<>). Table 3 Granger causality tests: summary MC VT i—» <> <-<-<> CS 7 0 2 5 6 0 2 6 OECD PF 7 1 1 5 6 0 3 5 OECD 0 LI 9 0 0 5 6 1 2 5 NL 6 2 1 5 4 1 3 6 CS 2 0 1 2 2 2 0 1 Non-PF 2 0 1 2 0 0 1 4 OECD LI 1 0 1 3 3 1 1 0 NL 0 2 1 2 3 1 1 0 There is significant evidence in these data that either causality between institutions and markets does not exist, or if it exists, it is predominantly from institutions to markets only. To a lesser extent, causality simultaneously exists in the two directions between institutions and markets. Furthermore, there is very limited evidence that causality runs from markets to institutions only (the only exception seems to be for non-life insurance in developing countries). Results seem to support the idea that the development of institutional investors is likely to promote the development of market capitalization more than value traded. For developing countries, pension funds seem not to Granger cause value traded development while life and non-life insurance do. Thus, in developing countries pension funds predominantly buy and hold shares. The following tables allow us to analyze other causality patterns among the countries in our sample. Table 4 lists, by institution, the countries for which we find one way Granger causality from institutions to market capitalization or value traded only; these are indicated with a "1". Table 5 lists, by institution, the countries for which we find a two way Granger causality between institutions and markets. Table 6 lists, by institution, the countries for which we could not find Granger causality between institutions and market on either direction. Schmollers Jahrbuch 120 (2000) 3 OPEN ACCESS | Licensed under CC BY 4.0 | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/schm.120.3.445 | Generated on 2023-04-04 12:27:53 476 Mario Catalan, Gregorio Impavido and Alberto R. Musalem When causality exists only from institutions to markets this seems to take place in countries where financial markets are not yet completely developed. In countries with complete and sophisticated financial markets like the United States, no causality is found in either direction. Notice though that results are ambiguous for some countries. For example, in Korea, pension funds and non-life insurance seem to Granger-cause market capitalization while life insurance and in general contractual savings seem not to cause market capitalization. For this country causality is stronger among institutions with respect to value traded. In the United Kingdom, all institutions seem to Granger-cause value traded and only contractual savings and life insurance companies, market capitalization. Table 4 Granger causality (one way) from institutions to markets only CS PF MC LI NL TOT CS PF VT LI NL TOT NLD 1 1 1 4 FIN 1 1 1 1 4 BEL 1 1 3 GBR 1 1 1 1 4 CAN 1 1 3 NOR 1 1 1 1 4 DEU 1 1 3 CHL 1 1 1 3 FIN 1 1 3 KOR 1 1 1 3 THA 1 1 3 SWE 1 1 1 3 AUT 1 1 2 MYS 1 1 2 ESP 1 2 AUT 1 1 GBR 1 2 BEL 1 1 KOR 1 1 2 NLD 1 1 NOR 1 1 2 PRT 1 1 SWE 1 1 2 SGP 1 1 ZAF 1 1 2 THA 1 1 PRT 1 1 ZAF 1 1 AUS 0 AUS 0 CHL 0 CAN 0 MYS 0 DEU 0 SGP 0 ESP 0 USA 0 • USA 0 TOT 9 9 10 6 TOT 8 6 9 7 Schmollers Jahrbuch 120 (2000) 3 OPEN ACCESS | Licensed under CC BY 4.0 | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/schm.120.3.445 | Generated on 2023-04-04 12:27:53 Contractual savings or stock market development 477 Table 5 Granger causality (two ways) between institutions and markets CS MC PF LI NL TOT CS PF VT LI NL TOT PRT 1 1 1 3 DEU 1 1 1 3 CHL 1 1 2 NLD 1 1 1 3 NOR 1 1 2 THA 1 1 2 THA 1 1 AUS 1 1 ZAF 1 1 AUT 1 1 AUS 0 CHL 1 1 AUT 0 PRT 1 1 BEL 0 SWE 1 1 CAN 0 BEL 0 DEU 0 CAN 0 ESP 0 ESP 0 FIN 0 FIN 0 GBR 0 GBR 0 KOR 0 KOR 0 MYS 0 MYS 0 NLD 0 NOR 0 SGP 0 SGP 0 SWE 0 USA 0 USA 0 ZAF 0 TOT 3 2 1 3 TOT 2 4 3 4 There are other facts that help interpret some of our results. For example, the absence of causality in either direction in Malaysia and Singapore could be explained by the contractual savings regime in these countries as well as financial sector policies. Singapore and Malaysia have centrally managed provident funds, which are not geared at investing in shares. In Malaysia, contractual savings institutions invested in shares from 4 to 7 percent of their financial assets during 1987-93. Singapore only recently has allowed some members to pick private managers and to determine how a portion of their Central Provident Fund balance will be invested.30 Therefore, there should be no surprise that there is no causality in any direction between contractual savings and stock markets in these countries. 30 See Asher (1999). Schmollers Jahrbuch 120 (2000) 3 OPEN ACCESS | Licensed under CC BY 4.0 | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/schm.120.3.445 | Generated on 2023-04-04 12:27:53 478 Mario Catalan, Gregorio Impavido and Alberto R. Musalem Table 6 No Granger causality between institutions and markets CS PF MQ LI NL TOT CS VT PF LI NL TOT USA 1 1 1 1 4 ESP 1 1 1 1 4 AUS 1 1 1 3 USA 1 1 1 1 4 MYS 1 1 1 3 BEL 1 1 1 3 SGP 1 1 1 3 CAN 1 1 1 3 AUT 1 1 2 AUS 1 1 2 CHL 1 1 2 MYS 1 1 2 ESP 1 1 2 PRT 1 1 2 GBR 1 1 2 AUT 1 1 KOR 1 1 2 CHL 1 1 SWE 1 1 2 DEU 1 1 CAN 1 1 KOR 1 1 DEU 1 1 SGP 1 1 FIN 1 1 THA 1 1 ZAF 1 1 ZAF 1 1 BEL 0 FIN 0 NLD 0 GBR 0 NOR 0 NLD 0 PRT 0 NOR 0 THA 0 SWE 0 TOT 7 7 8 7 TOT 7 9 5 6 Table 7 Shares of stocks in investment portfolios: selected countries Country Year Contractual Life Pension Funds Country Savings Malaysia 1993 7.01 17.86 5.17 Singapore 1996 5.67 33.50 0.00 Source: WB institutional investors database. Another particular case is Chile, where causality for pension funds runs in both directions. This could be explained, in great part, by their investment regulations. When the system was introduced, they were quite draconian, at that time; the Government was mainly interested in preserving assets, hence, pension funds were not allowed to invest in shares.31 In addi31 At the beginning of Chile's pension reform, the investment regulations allowed up to 100 percent in government securities, up to 60 percent in corporate bonds, and up to 70 percent in each of the following categories: mortgage-backed securities, letSchmollers Jahrbuch 120 (2000) 3 OPEN ACCESS | Licensed under CC BY 4.0 | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/schm.120.3.445 | Generated on 2023-04-04 12:27:53 Contractual savings or stock market development 479 tion, real interest rates on bonds and bills were very high, hence pension fund portfolios were heavily weighted on government securities. As the system and the market developed, the regulations allowed increasing participation of shares in pension fund portfolios. At the same time, real interest rates were declining thus demand for shares increased fueled by both effects. Obviously, regulation of investment policies of these institutions and after tax rates of return on financial instruments matters.32 The cautiousness and reactive approach followed by the Chilean authorities resulted in a two-way causality. The evidence is consistent with the direction of causality emphasized in this paper. Contractual savings promote capital market development in countries where capital markets are relatively small. Of course, in countries where capital markets are already developed, the effect is not as strong and the direction of causality is not as clear. In those countries, we expect reciprocal and weaker effects between both variables. The latter would be, in part, due to the fact that the illiquidity effect of contractual savings, as discussed above, would be diluted in countries with well-developed financial markets.33 6. Summary, conclusions and recommendations Contractual savings are powerful enough to increase the supply of longterm funds and develop the capital markets in an economy. This is because contractual savings institutions have long-term and illiquid liabilities on their balance sheets. We argued that contractual savings development, in addition to its primary purpose of providing protection to the insured, produces the following effects: a) specialization in the financial sector where the banking system adjusts towards its comparative advantage as contractual savings grow, thus reducing banks exposure to term transformation risks (which may imply that banks could still lend long term but now they could better fund this activity by mobilizing resources from the contractual savings institutions); b) improvement in the financial structure of firms by reducing their leverage and refinancing risks; c) impact on the term structure of interest rates, the stock market and growth; d) reduce the implicit debt from unfunded liabilters of credit or fixed term deposits. As the market developed, regulations were relaxed to allow investments in shares, mutual funds, real estate funds, venture capital funds, securitised credit funds, foreign securities and hedging instruments. 32 See Srinivas, Whitehouse and Yermo (1999). 33 The direction of causality from contractual savings to capital markets was also accepted in Impavido and Musalem (2000). Schmollers Jahrbuch 120 (2000) 3 OPEN ACCESS | Licensed under CC BY 4.0 | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/schm.120.3.445 | Generated on 2023-04-04 12:27:53 480 Mario Catalan, Gregorio Impávido and Alberto R. Musalem ities of defined-benefit plans; and e) develop the market for long-term government bonds and increase possibilities of public debt management. We also argued that these effects must be stronger in developing countries than in developed ones, due to the instability of banks in developing countries. Therefore, contractual savings mitigate social and financial risks, thus improving the resilience of the economy to shocks, reducing the country risk premium, the level of interest rates, and the cost of capital, thereby promoting growth. In addition, the growth of contractual savings or either mutual funds or non-life insurance should produce different effects on capital markets. Contractual savings should be more powerful in developing capital markets because of the additional effect on the liquidity of households' and firms' assets. In the empirical analysis we showed that those countries with more developed contractual savings sectors are also the countries with more developed stock markets, both in terms of market capitalization and value traded. In addition, those countries where the contractual savings sector grew the most are also the countries that experienced the highest growth in market capitalization and value traded. In the analysis of causality between contractual savings and both market capitalization and value traded, the evidence strongly favors causality from contractual savings to market capitalization, particularly, in countries where capital markets are relatively small and have an enabling regulatory and policy environment. These results are confirmed by differentiating, with contractual savings institutions, between pension funds and life insurance companies. Causality between other institutional investors, like non-life insurance companies, and markets appear to be much weaker. For OECD countries, the direction of causality from contractual savings to stock markets and liquidity predominates. The small sample of developing countries results are mixed with Chile exhibiting causality in both directions, while Malaysia and Singapore exhibit little if any form of causality between institutions and markets. In these two countries, the fact that management is public and the governments have severely restricted investments in domestic capital markets is probably responsible for this result. Countries interested in developing contractual savings are usually confronted with the issue of having underdeveloped capital markets. Hence, sequencing of reforms is important. Our analysis suggests that significant benefits will be derived from developing contractual savings even if capital markets have not reached their appropriate level of development. Initially, contractual savings institutions could invest primarily in government securities, corporate bonds and long-term loans, and to the extent possible, in Schmollers Jahrbuch 120 (2000) 3 OPEN ACCESS | Licensed under CC BY 4.0 | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/schm.120.3.445 | Generated on 2023-04-04 12:27:53 Contractual savings or stock market development 481 shares and foreign securities.34 This would be equivalent to a strategy combining Chile and the Netherlands. The difference is that Chile, at the beginning of its pension reform, did not allow investments in shares, loans or foreign securities while it allowed investments in bank deposits. Such a strategy could work in an environment of fiscal discipline and sound banking supervision. This is why we believe that long-term loans to the private sector offer better prospects as evidenced by the Netherlands. Simultaneously, the authorities should start improving the regulatory framework for capital markets development (bond and stock markets), including regulations on asset-backed securities (e.g., mortgage bonds), futures and derivatives. As the market develops, investment regulations covering contractual savings institutions could become more flexible while moving from non-market based instruments (e.g., loans) to market based securities and ultimately adopting the prudent person rule. Thus, the strategy advocates a comprehensive approach to contractual savings and capital market development. We believe that it will provide greater benefits than first pursuing capital market development and only then promoting contractual savings. Both should be pursued simultaneously. Obviously, a successful reform requires an enabling macroeconomic environment, a sound banking system as well as reliable financial sector regulation and supervision, and an appropriate tax treatment. 34 Investment in foreign securities provides the potential for risk diversification to the insured (if investments are made in markets which have low or negative correlation with the local market) and could have a direct effect of preventing development of domestic capital markets. However, it signals that the government is committed to having an open capital account which may induce higher capital inflows and an indirect positive effect on capital markets. Hence, the net result could be positive. Schmollers Jahrbuch 120 (2000) 3 OPEN ACCESS | Licensed under CC BY 4.0 | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/schm.120.3.445 | Generated on 2023-04-04 12:27:53 482 Mario Catalan, Gregorio Impavido and Alberto R. Musalem o o co o o iH c» co 00 Ol co rH CO c-IO Ol IO iH IH Ol r> Ol 00 lO co 00 ^ co co co t> co o o CO IO rH IO o IO co 00 o IO 00 co 00 co oq o co ^ CSJ ^ CTI IO IO t-rH © IO CSI ^ t>-^ Ol o O o o d o o o © d d d d d d d d d d d d d d d o 00 o 1—1 co Ol o Ol co o o o co o o 00 O o o o o co o co CO co o IO ^ co Ol 00 ^ co i-i o o D-IO O CSI IO CSI Ol co Ol © CSI o co £> co co IO CSI IO IO IO CSj co CSI IO CO co co 00 IO IO ^ rH T-H O csi © O © IO o d T—i csi i-L d T-H T—i d ^ LO 1-1 csi T—i d i-I d ci ci e? ci ci ci ci ci ci ci ci ci CSI ci ci ci ci ci C^ ci ci ci ci ci ci ci ci ci ci ci ci ci ci ci ci ci ci ci ci ci ci ci ci ci ci ci ci £ •rH 2 'rS 3 2 2 S •»H 'rS 'rB 2 O u o O U U U U U o O O U u o O U U O O U O O O PQ •-5 tuo tì a SH O rQ O a o U t> Ol o IO C-00 00 T—1 o co IO i-i r> tr-00 iH Ol o o co IO co io CSI iH CSI Ol IO iH CSI i> ^ CSI csi io co i-l iH iH co CSI 00 co ^ IO o I> o co o Ol rH co CS] co i—i CSI Ol IO o LO q CO co d d d d d d d d d d d d d d d d d d d d d d d d co Ol o t> o iH o ^ co O 00 o co o o oo LO o CSI o LO i> i—i ^ IO o IO o CO IO IO Ìo TH 00 co i> co 00 o rH co iH LO co LO o 00 co co o iH co iH o TH o co IH IH o co CO ^ LO o 00 CO d d d d IO d ^ d CO d csi d rH d csi i-I d d ^ d IO d d d rH rH iH TH T—1 IH TH T—1 IH IH 1—1 IH iH IH IH TH i—I T—1 IH IH IH TH ci ci ci ci ci ci ci ci ci ci ci ci ci ci ci ci ci ci ci ci esf ci ci ci 2 2 2 X 2 2 2 S ^ ^ ^ ^ ^ ^ £ ^ ^ le! O U u o o u o U O U u O u U U o u o O O o O O O ^ co CSJ T> IO r-co CSI IH t> 00 LO Ol LO co I-I o [> i—i T—1 LO IO IH Ol Ol co ^ co CSI IH CSI IH ^ T> T> LO co o 00 co co ^ Ol co IO o o T-H Ol CSj 00 co R> CSI co Ol co q LO q CO ^ TtH d d d d d d d d d d d d d d d d d d d d d d d d IO C-^ o 00 o i> o o o Ol T> o ^ T> ^ o co o CSJ T—1 co O Ol Ol CSI o o CSl [> IH IO co o IH 00 o 00 LO LO co ^ ^ o o I> CSI rH IH © IH co o t> o Ol IH Ol q CSI LO co CSJ q q LO d d d d d d csi d i-I d d d i-I d d d co d d d d ^ ^ ^ ^ rH i—R rH T—T i—I i-T 1—1 1—R rH rH T-H T-T iH i—T rH rH ci rH rH ci 1—1 T—r ci T-H r—T ci TH T—r co T-T CO i—r co i—r CO i—T iH T—r iH r—T T—1 T-T TH T—r o i—r O T—r O i—r ST ST ET ST ST ST ST ET ST ST ST ST MC co o VT cs a co S o VT CS MC CS VT cs MC cs VT CS MC cs VT CS MC CS VT T T T T T T T T T T T T T T T T T T T T T T T co o MC co U VT co u o S CS VT CS MC CS VT CS MC CS VT CS MC CS VT co o y co S o I> T> T> c-c— CO LO CO LO c— T> T> co co co 1—lOlOlOlOli—IT—IT—ITHTHT—IT—IrH O T3 a» <D •a ¡3 tuo I—I <D PQ 3 < (-1 S ctf a co Schmollers Jahrbuch 120 (2000) 3 OPEN ACCESS | Licensed under CC BY 4.0 | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/schm.120.3.445 | Generated on 2023-04-04 12:27:53 Contractual savings or stock market development 483 CD CD 00 00 OS O O O ^N^t0l>00ONHONONNOO>0005(0^NC0m50 D-NOMhtDCOH«m>ODMM!0010^NONOOOin T^oìooooooi>i>i>cD05oooooooooocoTtHooooooot>i> oooooooooooooooooooooooooooo o tr-CD o O CSJ c— CSI 00 tr-o IH CSI co LO CSI ir-CSI co o o o o LO OS co o o os CSI OS OS CD co CD csj os 00 CS] o LO co CD o rH OS o o CD co 00 CD CD ^ I> ^ 1—1 ^ CSI CSI LO ^ CD Oí 1—1 CSJ CSI [> co co CS] t> o co OS co LO i-i O O r-i iH o o o o o o o o o o o o © o o 1—¡ CD o o o o o s s cq s s s s c^ s s s s s s s s s s s s s s s s s s fri cí ci ci cí ci ci ci ci ci ci ci ci ci CSI ci ci csf CSI CSI CSI ci ci ci ci ci ci CSI 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 £ 2 2 2 2 2 2 2 o o o o o o o o o o o o o o o o o u u o u u u u o o o o co t> co CSI co OS CSI CD 00 00 LO LO 00 1-1 00 o 1-1 o 00 os CD 1—1 CSI 1—1 CD o CD o CD CSI iH co CD 1—1 os CSI 00 CSI OS OS OS rH I> o rH o o co o 1—1 CD o co o LO o o o OS 1—1 co o o o os 1—1 00 o CD o 00 o o co o o t> o IO o o o o o o o o o o o o o o o o o o o o o o o o o o o o o [> o o o CSI o OS o o o o o 1-1 CSI o 00 o co 1-1 LO 1—1 ir-o 00 CD co CSI co LO o LO cr-00 OS es o CD 00 o IO CD o csi o csi LO o 1—i c-©co©i>ocooco rHTjHI>OlOOOO^ CDOCOOOOCDO O O CO 00 o co O CD O) 00 O 00 H M Ifl O ^ N O CD O O ^ O 1—1 1—1 1—1 1—1 rH rH rH IH rH rH 1—1 1—1 1—1 l-l IH 1—1 1-1 1—1 1—1 rH rH 1-1 1—1 rH rH 1—I rH 1—1 (N CS] ci CSI CS] ci ci CS] ci ci ci ci ci ci ci ci ci ci ci ci CSI ci ci ci CSI CSI CS] ci 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 o o o o o o o o o o o o o o o o o o u o o o o o o o o o os (NI 00 00 CSI 1—1 CSI co co CD 1-1 tr-1-1 LO o CS] CS] 00 tr-LO y—i 1-1 1-1 LO co tr-r-t> LO r-LO CD os os 1-1 LO CSI LO CD 1—1 CS] co LO 1—1 co CS] c-os o CD o o o os 1-1 o 1-1 o os CS] os o ir-o 00 o co o 00 o o CD 00 o CD o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o 1-1 o CSI o o o o o CSI o o o co o os o LO o o t> o CS] os tr-00 CD ^ o 00 os 00 co CSI o CD 1-1 1-1 co o co o LO 00 co co Tt< os os os o CSJ CS] CSI ir-o o LO 00 co 1—1 o CSI o os 1—1 CD o CS] o CD o co CS] cq o 1—( rH o LO co o csi o co csi Tji o I-i o o t> o [> 1-1 o LO o 00 LO o o o rH rH 1-1 rH 1—1 1-1 rH 1—1 cÑ? iH CSI rH co 1—1 co 1—1 1-1 1-1 rH 1-1 1-1 1—1 1—1 1—1 1-1 rH 1-1 1-1 i? fa sr sr fa fa ST sr ST sr S o VT es MC es VT es MC es VT CS î Î î T î î î î T î î î w a u S co o VT co o MC co o VT co u MC co o VT rrc-c-c-r-LO LO CD CD oT i—r os i—r oT i-T oT i—r i—r i—r i—r i—r iH i—r iH i—r iH TÍH iH î—r co i—r co î—r co i—r co i—T ST ST ST ST ST fa fa fa fa fa fa fa fa fa MC CS VT CS MC CS VT CS ö CO S o VT CS MC CS VT CS î T î î T î î T î î î î T î î î co o MC CO o VT CO o MC CO o VT co o MC co u VT co u MC co o VT CS] 1-H CS] rH CS] rH CSI rH c-t> D-rH DrH tH iH CD CD CD CD Schmollers Jahrbuch 120 (2000) 3 OPEN ACCESS | Licensed under CC BY 4.0 | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/schm.120.3.445 | Generated on 2023-04-04 12:27:53 484 Mario Catalan, Gregorio Impavido and Alberto R. Musalem PQ »-5 q; tuo ti c« O rO O ti o U 00 CTS t> CSI CSI co CSI CSI CSI O IO Oì 00 ai I> CO 00 Oì O o t> 00 IO CSI 00 00 CSI o CS] I> O LO o Oì CO i-H co 1-H CSI 1—1 co 00 O o CO CO CTS I> CO q co t> ^ IO i-H LO q co LO q ^ q o CSI q q q ^ [> ^ O © © d d d d d d d d d d d d d d d d d d d d d ^ IO IO o CSI i> o o o o o 00 o o o o o o o co o o Oì o Oì co CO i> Ci o i—i CSI co Oì o CSI ^ i—i LO co o co 00 CSI o co o (M © co CSI 00 co £> i> i-H IO 1-H q [> i-H q q ai CSI q CO q q co q LO © d © ^ d d i—i i—i ^ 1-1 csi d i—l i> 1-H t> 00 LO d csi 1—1 d rH CSI CSI S 0 s e? § e? c? ci c? § c? § § c? s s s s s S ci Q § S ci ci ci ci ci ci ci ci ci ci ci ci ci ci ci ci ci ci ci ci ci ci ci ci 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 •iH 2 O U u O o L> u o O O o o o O u u U U O o o U O u o ai TJH CD ^ co 00 co co co o o co LO 00 co Oì ^ co ^ r> o 1-1 CSI CSI 00 o o LO co co i-H Oì co ai LO CSI LO co t> o D-i-H 00 q q 1-H I> q q q CSj rH q Oì o CSI LO ^ ai o 1-H o co q q CSI q © © d d d d d d d d d d d d d d d d d d d d d d o o o CS] o o o o o 00 1-H o o LO co co o o o CSI o o o o L-c-CSI 00 co Oì co Oì CSI o o cq Oì co o co Oì co o 00 o LO CSI . TtH co i-l Oi ^ co co Oì CSI q co i-H co ^ q q q LO iH CSI CS] LO ai co 1-1 IO csi d csi co i-I i-I d d LO i-l d d d LO i-I ^ d oo co 1-1 csi 1-1 1—1 TH 1-H i—i i—I i—i 1-H 1-H i-H i—i 1—1 i-H 1-H 1-H i—I i-H rH iH iH rH rH 1-H ci cg ci ci ci ci ci ci ci ci ci ci ci ci ci ci ci ci ci ci ci ci CSI CSI 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 U O U u U U o U U u U O O O u O o o o o o o U u ^ 00 co oo IO ^ 00 rH i> CSI CSI LO D-o i-H co t> co 00 o iH co CO CSI t> Oì o IO co I> CO t> LO co oo co ai co o co iH CS] C-co q i—i CSI l> CSI q q 00 CSj q q q co LO LO q q CSj iH q iH CSj iH © ci d d d d d d d d d d d d d d d d d d d d d d o o o i> o o o IO o CO i-H o t> co 1-H 00 o o o LO o o o o CSI CSI 00 r-Oì CSI o ai 00 o ^ Oì 1-H 00 o 00 LO o CSI 00 o o co co q q q i-H ai IO i-H q co co co co q q co iH q t> co ^ 00 co i—I d i-i IO LO d i-I d d ^ d d d d co i-I co d co csi i-I csi io io io io co co CO co csT 1-H csT i—i ci i-H csT i-I c? i—i cT i—i ^ i-H i-I 00 00 co co co iH co iH co iH co iH 1—r i-T i-T i-T i—r 1—r 1-H i-T 1-H 1-H r-T iH rH i-T r-T 1-H i—r i—r i—r i—r r—r rH rH" 1-H sr ST ST ST ST ST ST ST SÌ ST ST ST £7 Pn fe ST ST sr ST fe ST o CO EH CO o CO EH CO U CO EH CO U CO Eh co u CO Eh co u CO EH CO s O > u S u > U S u > u s o > U a o > u S u > o T T T T T T T T T T T T T T T T T T T T T T T T co u co Eh co u co Eh co u co H co u co Eh co o co Eh co u co Eh U S u > u S u > o S u > u S u > u ^ o > u S u > LO LO LO IO LO LO t> 1-H iH iH rH Oì Oì oì Ol 00 00 00 00 Oì Oì Oì Ol i-I i-I i-H i-H i-I i-I 1-H i-H i-H iH iH rH iH i-H rH iH ed tuo t o PL, U s o & g 00 cd s T5 tì ed i—i 'ed o co Schmollers Jahrbuch 120 (2000) 3 OPEN ACCESS | Licensed under CC BY 4.0 | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/schm.120.3.445 | Generated on 2023-04-04 12:27:53