سنجش واسطه گری مالی در ژاپن

In this paper, we examine the evolution of the Japanese financial structure, in order to challenge the expected incidences of the financial liberalization. We compute financial intermediation ratios for Japan (1979–2004) on a book value basis. According to our results, the intermediation ratio has remained quite stable, at around 85%. This stability is the result of two opposite trends: a decrease in credits and an increase in financial securities owned by financial (mostly, non-banking) institutions. These two trends are partly the consequence of the heavier weight of the Government in domestic external financing, which is traditionally less financed by credits than companies are. Besides, these two trends would not have appeared if we had used intermediation ratios in market value or other traditional indicators (Deposits/GDP, Loans to private sector/GDP, stock market capitalization/GDP, etc.). Our results provide evidence for a very close relationship between intermediate financings and market financings and tend to reject the hypothesis of the Japanese financial system's convergence toward a capital market-based system.

During the 1980s and the beginning of the 1990s, the Japanese banking sector underwent a partial financial liberalization. Manufacturing firms gained in independence vis-à-vis banks more quickly than did savers. Consequently, banks stayed large, but they had to find new opportunities for loans as the banking regulation prevented them from redeploying their activity. Some of the Japanese banks were even reluctant to write off bad loans and enter into forbearance lending (Kobayashi et al., 2002). For Hoshi and Kashyap (1999), the asymmetry in the liberalization process played a key role in the banking crisis and needed to be corrected by a deeper deregulation. The Japanese Big Bang initiated in 1996 was aimed at radically transforming the financial system toward a market-based system. Hoshi and Kashyap (1999) expected the Big Bang would lead “more firms to migrate to capital market financing”, and would induce “a massive contraction in the size of the Japanese banking sector”. Anderson and Makhija (1999) underlined, as a result of this financial liberalization, “a disintermediation evident in Japanese balance sheets”. Such an analysis of the financial mutation incidences is not specific to Japan, but is often applied to many other countries (see, for instance, Allen and Santomero, 2001 and Rajan and Zingales, 2003a). Three points are generally considered. The first one deals with the incidence of financial mutation on the relative importance of financial intermediaries compared to capital markets. A current view is that financial development is prejudicial to banks. Capital markets compete with banks on both sides of the bank's balance sheet. On the assets side, the increase in claims undermines credit. On the liabilities side, the development of collective saving (which occurred relatively late in Japan) increases the cost of resources collected by banks. As a consequence, traditional banking firms should decline and there should be a necessary redeployment towards other activities such as financial engineering, risk management, etc. (Boot and Thakor, 2000). In the specific case of Japan, the decline of traditional banking would also be accompanied by a weakening of the main banking relationships. The second point relates to the global evolution of financial systems. Nowadays, the majority opinion puts forward the idea of a standardization of financial systems toward a capital market-based system, in opposition to a bank-based system. Interestingly, the only empirical study specifically dedicated to this phenomenon (Schmidt et al., 1999) finds neither a global trend toward disintermediation nor a convergence toward capital market-based financial systems in the major European economies (France, Germany or the United Kingdom) at a general level. The third question concerns the relative merits of bank-based versus capital market-based systems (for a comprehensive discussion on comparative financial systems, see Allen and Gale, 2000 and Levine, 2002). For a long time, the literature seemed to conclude in favor of capital market-based systems, at least for developed countries (Boyd and Smith, 1998). However, several recent studies using cross-country comparisons challenge this idea: although global financial development is a significant determinant of economic growth, there is no support for either the bank-based or market-based view (see Levine, 2005). One major problem with this traditional approach is that financing structure is always a challenge to quantify (see, for a discussion, Beck et al., 2001). Consider, for instance, the Japanese case. While Japan is still referred to as an archetype of a bank-based system, Tokyo is one of the leading financial centers in the world; in 2003, Tokyo was ranked first in market capitalization of newly listed domestic shares. Moreover, the Japanese corporate bond market is small, but the Japanese Government Bond (JGB) market is the largest in the world. At first sight, one can interpret this as evidence of the convergence of the Japanese financial system toward a capital market-based system. But it seems more interesting to see it as evidence of a close connection between banks and capital markets in the financial system. Beyond the empirical problem, a more fundamental issue challenges the ground of the traditional approach. The latter considers markets and intermediaries as two substitutable and opposite modalities. But one should not ignore the important interactions between markets and intermediaries: the services provided by each may overlap, and it is very likely that there is cross-fertilization between markets and intermediaries. In accordance with Levine (2005), we suggest that the debate should not focus on bank-based versus market-based systems. By nature, the traditional approach cannot cover the whole variety of financial systems. Moreover, this dichotomous vision of financial systems is gradually being replaced by theoretical analyses that underline the complementary quality of financial services. For instance, Merton and Bodie, 1995 and Merton and Bodie, 2004 argue in favor of a functional approach instead of an institutional one; in this case, banks and capital markets are not opposed, but assume largely identical functions (financing, portfolio management, risk management, liquidity insurance), although in different ways. Moreover, numerous studies challenge the convergence hypothesis by highlighting the role of historical or institutional characteristics such as the legal system, the political context, the cultural and religious legacies, the geographical endowments or the social capital that may shape national financial systems (see, again, Levine, 2005). In any case, these conflicting analyses highlight the need for better empirical measures of financing structure. The aim of this paper is to empirically assess the financial structure in Japan in order to challenge the expected incidences of the financial liberalization. There is little empirical evidence concerning the evolution of the Japanese financial structure and our paper attempts to fill this gap. Traditional measures of financial structure1 focus more on the level of financial activity, give a narrow vision of financial intermediation and do not capture its changing nature, nor do they correct for changes in security prices. In this paper, we compute financial intermediation ratios, which do not suffer from these drawbacks. In our breakdown of external financing, we do not consider only direct and indirect financing, but we also break down the total financial intermediation ratio into the sum of the credit intermediation ratio and the market intermediation ratio. This finer breakdown allows us to appreciate the modalities and the changing nature of the financial intermediation. Moreover, as flows are too erratic for long-term studies, we use stock series. But stock series are not fully suitable since they are usually expressed in market value, which amplifies financial cycles. Therefore, we correct market value to obtain book value. That is particularly significant in the case of Japan, whose stock exchange market was affected these last decades by episodes of very strong volatility and bull and bear markets. Our study offers new stylized facts concerning the Japanese financial system: the rise of the capital markets, which started at the beginning of the 1980s, did not involve a disintermediation and, symmetrically, the decline in stock market prices, at the beginning of the 1990s and at the beginning of the 2000s, did not support a re-intermediation in Japan. The financial intermediation ratio remained stable overall between 1979 and 2004. This stability of the financial intermediation ratio is the result of two opposite trends: a decrease in the share of credits in the domestic nonfinancial sectors’ external financing (i.e. the credit intermediation ratio) and an increase in the share of investments in claims carried out by banks and other financial intermediaries (i.e. the market intermediation ratio). The first trend results from the decrease in credits requested by corporations and from the reduced share of corporate fund raisings in domestic external financing relative to government, which is traditionally less financed by credits. Similarly, the second trend results from the increase in financial institutions’ investments in securities issued by the domestic nonfinancial sectors and from the growing share of government in domestic external financing. In other words, the global evolution of intermediation ratios is explained both by the change in the relative weight of government and corporations in domestic external financing and by significant changes in the financial behavior of domestic nonfinancial agents. The nature of financial intermediairies’ participation in the financing of the Japanese economy, therefore, has changed: the increase in claims investments has compensated for the relative fall in credits. However, our results suggest that this evolution in the supply of financing was not due to a proactive behavior of financial intermediaries, but rather mostly reflects the increase in the government need for financing. The paper is structured as follows. Section 2 presents the data and defines the intermediation scope. Section 3 provides results. Section 4 concludes.

Our results show that the intermediation ratio in Japan remained quite stable between 1979 and 2004, at around 85%. However, this stability is the result of two opposite trends: a contraction of credits and an increase in financial securities owned by financial (mostly public and non banking) institutions. We stressed also that the decrease in the credit intermediation ratio and the increase in the market intermediation ratio are not fully explained by a change in financial behaviors, but are also due to a composition effect of external financings: the decreasing share of corporations in the domestic non financial sector external financings makes the relative fall of loans stronger. Conversely, the increasing share of the general government sector accentuates the relative rise of securities in intermediate financings. At the methodological level, one of the general results of this study is that the measurement of intermediation ratios must be done with outstandings (better suited than flows to long period analyses) corrected for stock price changes. If we had used data on a market value basis, one would show in periods of a strong fall of the Japanese stock market (early 1990s and early 2000s) an increasing, but actually artificial, intermediation ratio. Similarly, a computation using data on a market value basis would have emphasized an artificial disintermediation during the sharp rise in the Japanese stock market from 1980 to 1990. In this respect, a generalization of accounting standards concerning market value would make the measurement of the financings structures even more difficult. In any case a treatment of the data would have to be systematic. Finally, on a theoretical level, by showing that intermediate financings increasingly consist in the purchasing of claims by financial intermediaries, our study confirms the close connection between market financings and intermediate financings and, in this respect, goes beyond the case of Japan. Concepts traditionally used to analyze the evolution of financial systems thus appear less and less operational. This is particularly true regarding the distinction between market (or direct) financing and intermediate (or indirect) financing, and between a bank-based system and a capital market-based system. Market financings are to a large extent intermediate and the orientation of the financial systems is basically mixed. Therefore, to characterize the structures of financing and their evolution over time, the relevant typology should be centered on intermediate financings between credit and claims investment by financial intermediaries.