Corporate finance in developing countries: New evidence for India

WorldDevelopmentVol. 26, No. 6, pp. 1033-1047, 1998 Pergamon 0 1998 Elsevier Science Ltd All rights reserved. Printed in Great Britain 0305-750x/98 $19.00+0.00 PII: SO305750X(98)00025-4 Corporate Finance in Developing Countries: New Evidence for India* COBHAM zyxwvutsrqponmlkjihgfedcbaZYXWVUTSR DAVID University of St Andrews, St Andrews, UK and RAMESH SUBRAMANIAM Asian Development Bank, M anila, Philippines Summary. Recent work by Singh and Hamid (1992, Corporate Financial Structures in Developing Countties) and Singh (1995, Corporate Financial Patterns in Industrializing Economies: A Comparative International Study) has suggested that large firms in a number of developing countries, including India, use much more external finance in general, and equity finance in particular, than those in developed countries. However, the contrast is in part a product of methodological differences, and India is much less different from countries such as France and Italy. The Singh results are not due merely to bias arising from the focus on the largest companies, since the rest of the Indian corporate sector also issues large amounts of equity, via informal networks rather than organized stock exchanges. The importance of such issues suggests the need for further research before policy conclusions can be drawn. 0 1998 Elsevier Science Ltd. All rights reserved Key words - Asia, India, corporate financing, firm-size effects, equity markets 1. INTRODUCTION There is a considerable general literature on finance and economic growth, and much recent work has emphasized the importance of the linkages between the real and financial sectors.’ The capital structure of firms in developing countries had attracted little systematic attention, however, before the pioneering study for the International Finance Corporation by Singh and Hamid (1992), which examined the financial structure of large firms in a sample of nine developing countries, found that external finance in general and equity finance in particular were much more important for these firms than had been previously thought. Singh and Hamid (1992) analyzed data on the 50 largest manufacturing companies listed on stock exchanges in India, South Korea, Pakistan, Jordan, Thailand, Mexico, Malaysia, Turkey and Zimbabwe over 1980-88.2 Their basic method is to calculate the “net assets” of firms, equal to total assets minus current liabilities, and then to consider the sources of finance internal finance (retained profits, net of depreciation), equity and long-term debt (bond issues plus long-term bank loans) - as a proportion of the growth of net assets. They found, first, that these companies seemed to use much more external finance than companies in developed countries and, second, that they also made much more use of equity finance than companies in developed countries. More detailed work was undertaken by Singh (1995) in a study designed partly to check the robustness of these results. The new data set covered 100 companies where possible, a slightly longer time period (19SO-90), and an *We would like to thank seminar participants at the University of St Andrews and the Royal Economic Society Conference (Swansea, 1996) for comments, Ganga Darbha, Subir Gokarn, K. J. Joseph, R. Nagaraj, Manoj Panda and M. H. Suryanarayana for help with data sources, and Gustav Ranis, Cherian Samuel, Ajit Singh and anonymous referees for helpful suggestions and comments. Subramaniam would like to gratefully acknowledge financial support from the May and Stanley Smith Trust at the University of St Andrews and the hospitality of Yale University Economic Growth Center where a previous version of this paper was written. The views expressed here are those of the authors, and not of the organizations with which they are affiliated. Final revision accepted: January 12, 1998. 1033 WORLD DEVELOPMENT 1034 zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA major independent contribution to economic Table 1. Top listed companies in manufacturing mean propotiion of internal and external finance of corporate growth (% ), Singhs (1995) results Source/country Internal finance Equity Long-term debt Korea Pakistan Jordan Thailand Mexico India Turkey Malaysia Zimbabwe Brazil 19.5 74.0 66.3 27.7 24.4 40.5 15.3 35.6 58.0 56.4 49.6 1.7 22.1 N.A. 66.6 19.6 65.1 46.6 38.8 36.0 30.9 24.3 11.6 N.A. 9.0 39.9 19.6 17.8 3.2 7.7 ALL 38.8 39.3 20.8 Source: Singh (1995), extracted from Table C-5. extra country (Brazil). The key results, as exemplified by Table 1, confirmed the earlier findings: although there is considerable (and statistically significant) variation across countries, the mean proportion of internal finance (retained profits after tax as a percentage of the growth of “net assets”) was 38.8%, while issues of equity finance provided 39.3% and long-term debt provided 20.8%. For some countries (notably Korea, Turkey and Mexico) the use of internal finance was much lower and that of equity finance much higher. These results provide a sharp contrast with recent research on the net sources of finance for the non-financial corporate sector in developed countries, e.g. by Mayer (1988, 1990, 1994) and Corbett and Jenkinson (1994) which has shown that issues of securities (equity and bonds) do not provide large amounts of finance in any country considered, and have often been a negative source of finance in net terms in the United States and the United Kingdom. Moreover, the “pecking-order” theory of finance 1984) that firms prefer internal (Myers, financing to external and within the external category prefer debt to equity, which has been used to explain the latter findings, should, if anything, hold more strongly for developing countries where information imperfections are often more significant, companies are typically younger and less well-established, and capital markets are too small to benefit from the economies of scale found in the larger developed country securities markets. The Singh-Hamid results are also at odds with development economists’ conventional view that stock markets cannot be expected to make a growth (see, for example, Fry, 1988, p. 291; Fry, 199.5, p. 344; see also Nagaraj, 1996; Samuel, 1996) although in recent years some observers (e.g. Atje and Jovanovic, 1993; Atkin and Glen, 1992) have taken a more positive view of stock market development. Singh (1995) himself has rejected the suggestion that developing countries are simply repeating the historical experience of the United States (in which equity finance was originally important but has become much less so since the 1930s) and has put forward an explanation of his findings which relates to the specific circumstances of the period and countries concerned. He stresses the active role taken by governments in pushing the development of their stock markets through activities which include privatization, financial liberalization, and a range of specific measures affecting both the supply of and the demand for securities. He emphasizes the drop in the relative cost of equity capital over the 1980s which resulted from a large rise in share prices (and the associated rise in priceearnings ratios) and a rise in real interest rates, the former related internally to financial liberalization and governmental support for stock markets and externally to the awakening interest of developed country institutional investors in emerging markets, and the latter resulting largely from internal financial liberalization, together perhaps with some international influence. The main implication of this explanation is that the high level of external equity financing in developing countries over the 1980s is likely to turn out to be a passing phenomenon resulting on the one hand from the ending of conventional financial repression in these countries (together perhaps with the introduction of some new distortions intended to produce structural changes that are unlikely to become selfsustaining in the absence of governmental support), and on the other hand from periodspecific developments in the international environment facing developing countries. This explanation is attractive, if not yet wholly persuasive. Before it can be firmly accepted more research will be needed in a number of areas; time alone will generate relevant data to test whether the phenomenon is a transitional one or a continuing one. In the meantime a number of other useful points can be made. In this paper, we take one particular country, India, from the Singh-Hamid sample and examine it in more detail. In Table 1 India is shown as having had a near-average level of internal finance but less equity and more debt finance than the sample as a whole. However, in the early 1990s CORPORATE FINANCE IN DEVELOPING equity issues on the Indian stock exchanges continued to increase sharply, so that for the later period to which most of our results refer, India is likely to have been near to or above the mean of Singh’s sample. We first examine the net sources and Singh-Hamid methodologies and show that the contrast is at least in part the product of methodological differences. We provide some net sources of finance data for India, which suggest that, although markedly different from that in the Anglo-Saxon countries, the structure of corporate finance in India was not that far from that found in “Mediterranean” countries such as France and Italy. We then investigate, again with respect to India, the possibility that the Singh-Hamid work is samplespecific in that the large firms on which their analysis focuses are not representative of the private corporate sector as a whole. Our analysis of firm-level data from two samples of Indian firms suggests that small firms there make as much or more use of equity finance than large firms, while comparable firm-level data for the United Kingdom produce the opposite result. We rationalize these results by arguing that the nature of equity financing is different when firms are not listed on stock exchanges and are issuing equity through channels which involve less information asymmetry and less separation of ownership and control. Section 2 examines the differences in methodology between Singh-Hamid and the studies of developed countries by Mayer and his associates. Section 3 presents aggregate data for India compiled according to the net sources of finance methodology and compares India on that basis with the developed countries that have been studied in this way. Section 4 analyzes firm-level cumulative data in order to investigate whether in the case of India the Singh-Hamid results suffer from a large-firm bias, and Section 5 concludes and draws out the implications for policy. COUNTRIES 1035 2. METHODOLOGICAL ISSUES The net sources methodology can be understood most easily in terms of a simplified sources and uses of funds table, as in Corbett and Jenkinson (1994). The gross sources are presented as follows in Table 2. The gross of finance involves sources methodology considering each of the individual sources of funds in relation to the total sources. Partly because the total sources of finance are used to acquire financial as well as physical assets, the preferred net sources methodology subtracts Items 7 to 10 from the corresponding sources (see Table 3). The various net sources are then considered in relation to total net sources, which is equal to physical investment (gross fixed capital formation plus increase in stocks). The Singh-Hamid methodology uses balance sheet (stock) data rather than sources and uses (flow) data, but takes changes in the stocks so that in principle there should be no difference.3 It then subtracts depreciation (Item 14) from both internal finance (retained profits) and capital formation, and derives an aggregate (Item 15) for the change in net assets equal to the change in total assets minus current liabilities, as in Table 4. The sources of finance are then regrouped as (“net”) internal, external (equity) and external/long-term debt ( = longterm bank loans plus bond issues), and each of these is considered in relation to the total of net assets. To be precise, their ratios are given by: (a) internal finance of growth; (b) external finance of growth (long term debt); and (c) external finance of growth (equity): 1 -internal finance of growth-external finance of growth (long-term debt), where RP stands for Retained Profits, NA for Net Assets and LTD for Long Term Debt. p and m are the first and last years, respectively, in the period covered. Thus, the net sources methodology treats depreciation as a part of internal financing, looks at the net contribution from each source, and Table 2. Gross sources methodology zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQP Gross uses Gross sources 1. Internal 2. 3. 4. 5. Bank loans and advances New equity issues Bond issues Trade credit received 6. Total sources Identity: 1+2+3+4+5 = 6 = 7+8+9+10+11+12 7. 8. 9. 10. 11. 12. 13. = 13 Cash and bank deposits Equity purchases Bond purchases Trade credit given Gross fixed capital formation Increase in stocks Total uses WORLD 1036 DEVELOPMENT Table 3. zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJ Net sources methodology Net uses Net sources 1. 2-7. 3-8. 4-9. 5-10. Internal Net bank finance Net equity finance Net bond finance Net Trade Credit Total net sources ( = 6 - 7 - 8 -9 - 10) Identity: 1+(2-7)+(3-8)+(4-9)+(5-lo)= expresses the sources of finance as proportions of physical investment (including variations in stocks), while the Singh-Hamid methodology nets depreciation out from both the sources (in the form of internal finance) and the uses of finance, takes the main contributions gross, and expresses the sources of finance as proportions of the change in fixed assets net of current liabilities. Therefore both the numerators and the denominators differ between these three presentations of the sources of corporate finance. In particular, inspection shows that the total (gross) sources of finance, Item 6, must be.greater than total net sources or physical investment, Items 11+ 12 (hence the larger figure for the internal finance ratio in the net sources presentation). Similarly, total gross sources must be greater than total net assets, Item 15. The latter aggregate can in principle be greater or smaller than total net sources/physical investment, depending on the respective magnitudes of (i) firms’ acquisitions of financial assets, (ii) the change in firms’ current liabilities, and (iii) depreciation; but in practice the depreciation item is likely to be much larger than the net change in financial assets minus current liabilities,4 so that Item 15 is generally smaller than Items 11+12. This, together with the fact that equity is taken gross rather than net, explains why equity looks more important in the Singh-Hamid methodology. Similarly, internal finance appears smaller in the Physical investment 11+12 Singh-Hamid methodology partly because the denominator is larger and partly because it is considered net of depreciation. Singh (1995) provides some data for a sample of UK manufacturing companies using his own methodology, which show equity finance as considerably more important for UK companies than appears from the net sources methodology: for the average in his sample internal finance company accounted for 56.7% of the growth of net assets, equity finance for 16.6% and long-term debt for 26.6%. There is still a marked contrast, however, between these figures and the Singh-Hamid results for developing countries. The differences between these methodologies are not arbitrary. Singh-Hamid used a methodology which dates back at least to Singh and Whittington (1968) and is designed to answer questions about the growth of individual firms, while the net sources methodology is designed to investigate how the corporate sector finances its physical investment. The latter is less subject to international differences in the treatment of depreciation (Mayer, 1990; Corbett and Jenkinson, 1994). It is also less vulnerable to differences in the reporting of data that affect the gross sources methodology (Corbett and Jenkinson, 1994; Corbett and Jenkinson, 1997).’ In addition, a wider range of countries have been examined on this basis, so that more comparators are available: in particular, studies on this Table 4. Singh- Hamid methodology Sources 1-14. 2a. 3. 4. of finance Changes Internal finance (net of depreciation) Bank loans, long term Equity issues Bond issues Identity: in net assets 7-2b. Cash and bank deposits - short term bank loans 8. Equity purchases 9. Bond purchases 10-5. Net trade credit liabilities 11-14. Net fixed capital formation 12. Increase in stocks 15. Change in net assets (l-14)+2a+3+4=(7-2b)+8+9+(10-5)+(11-14)+12= 15 CORPORATE FINANCE IN DEVELOPING basis now exist for a range of countries other than the United States and the United Kingdom. For these reasons, and because our interest is in the relationships between investment, growth and the financial system as a whole, we choose to use the net sources methodology where possible (although for the samples of firms we examine in Section 4 data limitations mean that we can consider only the gross sources of finance). 3. AGGREGATE DATA FOR INDIA Table 5 provides data, at the aggregate level, on the gross (panel (a)) and net (panel (b)) sources of finance for India. Data are also presented separately for two periods: 1980-81 to 1986-87 corresponds to the pre-liberalization period in financial markets, and 1987-88 to 1992-93 approximately pertains to the liberalization period. It should also be noted that the figures in the table are period ratios (as in Corbett and Jenkinson, 1994) in that they show the contribution of each source of finance over the entire period. The figures presented in Table 5 suggest that market finance in India was rather less important on this basis than on Singh’s figures. For the whole period market finance in Table 5 is given as 17-18% in either gross or net terms. This figure can be contrasted with Singh’s (1995) figures of 19.6% of net asset growth financed by equity and 39.9% financed by long-term debt (which includes long-term bank loans as well as bonds) in 1980-90 (a period which excludes the two years 1991 and 1992 when equity issues were exceptionally high), as given in Table 1. The share of internal finance for the Indian corporate Table 5. Gross and net sources of finance Gross sources of finance as 5% of total identified Internal Market finance (equity issues+bonds) Bank lending Other (b) Net sources of finance as % of physical investment Internal Market finance Bank lending Other Notes: See Appendix (equity issues+bonds) 1037 sector is about 37.8% of gross finance and 42.0% of physical investment. Bank lending accounts for 31.2% of gross sources and (in net terms) for 28.8% of physical investment. The gross finance ratios decline marginally for retentions and “Other” sources between the two sub-periods, whereas for market finance and bank lending the ratios increase. All the net finance ratios, with the exception of the “Other” category, increase between the sub-periods (which may reflect mainly an improvement in the coverage of the data); in proportional terms the rise in market finance is particularly large. Examination of the data on a year-by-year basis shows that the pattern of important internal finance, substantial bank borrowing and moderate recourse to the capital markets is a consistent one over time, although there is some increase in securities issues in the early 1990s. The Indian figures can also be compared with the figures for other developed countries produced by Mayer (1990) and Corbett and Jenkinson (1994), as in Table 6. This table distinguishes between three sets of countries: those which Corbett and Jenkinson (1994) call “high internal finance” countries - the United States, the United Kingdom and Germany, designated as Group A; the others for which data are available Japan, France, Italy, Canada and Finland, designated Group B; and a subset of the latter - France, Italy and Finland - which arc designated as Group C. Group C thus excludes Japan, which may be regarded as zyxwvutsrqponmlkjihg sui generis, and Canada which is in some ways closer to the high internal finance countries; Group C is therefore perhaps more representative of those developed countries which lie between the two polar groups to which most attention has so far been paid. Type/source (4 COUNTRIES for the Indian corporate sector 1980-81 to 1992-93 1980-81 to 1986-87 1987-88 to 1992-93 37.8 18.4 31.2 12.6 38.5 15.3 29.3 16.9 37.4 20.2 32.3 10.1 42.0 16.9 28.8 12.3 33.1 10.4 22.6 33.9 49.7 22.5 34.0 -6.2 sources A for a general description of the Indian Flow-of-Funds table. Sources: National Income Bank of India’s Annual Reports and Reports on Currency and Finance. “Other” is the sum of trade credit and reported and unreported residual categories (see the text and Appendix A). Statistics and Reserve WORLD DEVELOPMENT 1038 Table 6. International comparisons zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQ Source Gross sources of finance Internal Equity Market finance Bank loans Net sources of finance Internal Equity Market finance Bank loans Group A Group B Group C India 60.4 to -4.9 to 3.1 to 14.7 to 62.7 7.0 9.3 23.3 38.5 to 3.9 to 7.4 to 12.8 to 54.2 11.9 18.0 41.5 38.5 to 5.6 to 7.4 to 27.2 to 44.1 10.8 13.2 41.5 37.4 to 38.5 N.A. 15.3 to 20.2 29.3 to 32.3 80.6 to - 10.4 to - 6.9 to 11.0 to 97.3 0.9 8.3 19.5 51.9 to -0.1 to 2.7 to 15.2 to 76.4 8.2 11.0 37.3 51.9 to -0.1 to 2.7 to 27.7 to 64.4 8.2 9.8 37.3 33.1 to 49.7 N.A. 10.4 to 22.5 22.6 to 34.0 Group A - high internal finance countries: US, UK and Germany; Group B - low internal finance countries: Canada, Finland, France, Italy and Japan; Group C - Finland, France and Italy. Notes: Data for US, UK, Germany and Japan pertain to 1970-89 and are taken from Corbett and Jenkinson (1994, Table 3). These are period ratios. Data for Canada, Finland, France and Italy pertain to 1970-85 and are taken from Mayer (1990, Table 12.3). These are averages of annual ratios. Data for India pertain to 1980-92 and are as explained in the notes to Table 2. The table gives the range (minimum to maximum) for each group of countries or, for India, across the two subperiods. Examination of Table 6 reveals that, although India is clearly very different from the high internal finance countries of Group A, it is not so different from the Group C countries. The gross finance ratios for India are within or close to the ranges for Group C. The Indian net ratios for internal finance and market finance are lower and higher respectively than those for Group C. Both of these differences are, however, consistent with the fact that Indian firms use a smaller proportion of their gross sources of finance to acquire financial (rather than physical) assets. On this basis it seems reasonable to conclude that the Indian private corporate sector is not financed in a way that is very different from the non-polar developed countries; in particular, India has much in common with France and Italy.” The macro evidence presented in Table 5 is based on the whole non-financial corporate sector. Thus it includes firms in all size-groups, ranging from the small to the very large. In order to test for any size effects, to see if the behavior of large firms is much different from that of smaller firms, we need disaggregated data. We shall therefore analyze cumulative balance sheet and income and expenditure data for a sample of firms in India from the Reserve Bank of India (RBI) and Industrial Credit and Investments Corporation of India (ICICI) data sets. We also examine some samples of United Kingdom firms from the Business Monitor surveys as a comparison. Appendix A presents a discussion of the nature of these data sets. We provide our results in terms of finance and investment ratios (both annual and period) and tests for size effects in the next section. 4. FIRM-LEVEL RESULTS FOR INDIA AND THE UNITED KINGDOM (a) Results zyxwvutsrqponmlkjihgfedcbaZYXWVU for firms in the RBI sample We begin our analysis with the Reserve Bank of India cumulative firm-level data set. Table 7 presents the averages of the annual ratios of sources to total finance and of total finance to physical investment for these firms. In terms of size-wise comparisons between large and small firms, a number of differences are found. The most striking difference between large and small firms is that internal finance is higher and bank loans are lower for large firms than for small. In addition, the annual ratios (not presented here) show that the ratios are much more variable for the small than for the large firms, and in some years bank finance is even more important than internal finance for the small firms. For the large firms, on the other hand, internal finance is consistently above 40%. The denominator (total identified sources of financing) in the ratios presented in Table 7 is the sum of the sources in the first three rows (internal, equity and total borrowings). The next two rows present the two most important elements of total borrowings: debentures and loans. Large firms make more use of bond finance than small firms, but small firms make slightly more use of equity finance. CORPORATE FINANCE IN DEVELOPING COUNTRIES 103Y Table 7. zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA Averages of annual ratios of (gross) sources of finance to total identified sources of finance and of total finance to investment for Reserve Bank of India data, 1981- 90 Source All firms Large firms Small firms Internal Equity Total borrowings” Debentures Bank loans 0.437 0.071 0.491 0.123 0.167 0.469 0.072 0.459 0.137 0.139 0.348 0.078 0.590 0.070 0.260 Total finance 1.111 1.149 1.004 to investment t-test for large vs small firms comparison” 4.60 0.13 -3.89 2.91 - 4.55 (14) p (18)~ (17)~ (18) p (16) p = = = = = 0.0004 0.8987 0.0012 0.0093 0.0003 2.63 (18) p = 0.0168 “Debentures, bank loans and other borrowings (from financial institutions) add up to total borrowings. The RBI tables provide only debentures and bank loans, but not the residual category. hDegrees of freedom are given in parentheses; a negative value for the t-ratio implies that the small firms group has a higher mean for the concerned finance ratio. Our preference is for the MCJ net sources methodology, but as stated above the RBI data do not allow us to calculate the net figures. However, we include in the last row of Table 4 the ratio of total identified (gross) sources of finance to net physical investment.’ These figures show that the large firms invest considerably more in financial assets than the smaller firms. Table 7 also presents the results of a set of f-tests of the hypotheses that the means of the ratios of each source of financing (over the 10 years, from 1981 to 1990) are equal between large and small firms. The variables (internal ratios for large and small firms, for example) are assumed to have unequal variances in calculating the t-ratios.8 Ideally we would have liked to use a cross-section, preferably a panel, of firm-level data in order to test whether the distributions of the finance and investment ratios are different between the various categories of firms, but such individual-level data are not available. Instead, we carry out a more limited test which is the only thing possible given the data: we examine whether the average ratios for each source of finance are different between the large and small groups of firms over the period under consideration. The degrees of freedom for the t-tests are provided in parenthesis in each cell. It can be seen from Table 7 that this hypothesis is rejected for each source of finance except for equity.’ Thus, there is a significant (probability values are also provided in Table 7) size difference between large and small firms in terms of internal, debentures and bank loans as sources of financing. The negative t-test values, for total borrowings and bank loans, imply that small firms (the latter group in the framework of the tests) have higher ratios in these categories than large firms. In other words, this test suggests that large firms use more internal finance and bond finance, but smaller firms use more bank loans. Smaller firms also use marginally more equity finance, but the difference is not significant. Below we provide some plausible explanations for this observed size difference in the behavior of firms. The next section presents the results for the ICICI sample. (b) zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPO Results for firmsin the ICKY sample Table 8 presents the averages of the annual finance ratios for the whole sample of firms as well as the different size subsamples in the ICICI data set. In column 1, for the whole sample, internal financing is the dominant source at 38%, followed by trade credits at 17%. Equity financing contributes a small 4% to total finances, while debentures and bank loans contribute 10% each. Thus, the pooled ICICI sample does not look much different from the pooled RBI sample in terms of the relative importance of the various sources, though both internal and equity financing are higher in the RBI than zyxwvutsrqpo in the ICICI sample. The ratio of total finance to investment presented in the last row is well above unity; this must reflect some combination of incompleteness in the data and firms’ acquisitions of financial rather than physical assets. The remaining columns in the table provide the averages of annual finance ratios for firms in the four different size groups. While the average ratios presented in Table 8 for the ICICI data show similar size-source patterns as in the RBI data, with larger firms using more internal finance and bond finance and smaller firms using more equity finance, the r-test results presented in Table 9 show that not all the size-source differences are statistically significant. In addition, as can be seen from Table 8, there is no consistent size pattern for bank financing (with firms in S2 and S4 reporting WORLD DEVELOPMENT 1040 Table 8. zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA Averages of annual ratios of (gross) sources of finance to total sources of finance for industrial credit and investment corporation of India data, 1982- 90, by size groups (Sl, S2, S3 and S4)” All firmsh Sl s2 s3 s4 Internal Equity Bonds Bank borrowings Trade credit 0.378 0.040 0.099 0.099 0.166 0.217 0.110 0.023 0.052 0.214 0.273 0.063 0.057 0.095 0.240 0.392 0.045 0.088 0.058 0.169 0.363 0.040 0.156 0.093 0.136 Total finance to investment 1.462 1.609 1.729 1.582 1.459 Source “Sl - firms with gross hxed assets (end of a financial year) below Rs. 5 Crores; S2 - firms with GFA between Rs. 5 Crores and Rs. 20 Crores; S3 - firms with GFA between Rs. 20 Crores and Rs. 50 Crores; S4 - firms with GFA above Rs. 50 Crores; 1 Crore = 10 million. bData pertain to 1980-92. The sample has 417 public limited companies for 1980-88 and 620 public limited companies for 1989-92. The other sources are: bank borrowings for working capital (O.OSl), unsecured loans and deposits (O.OSS), deferred credit (O.OOS), other borrowings (0.03) and other current liabilities (0.046), approximately. a higher ratio for bank borrowings than firms in the other two groups). The column for internal finance in Table 9 shows that there is no significant difference between Sl firms and those in S2 and S4. Firms in S3 use more internal finance than those in Sl. Firms in S3 and S4 use more internal sources than firms in S2 and there is no significant difference between S3 and S4 firms.‘” For equity financing (Table 9, column 2) there is no significant size effect in any of the pairwise comparisons, with the exception of Sl versus S4 where Sl firms report a significantly higher ratio for this source than the largest (S4) firms. On the other hand, there is a significant positive relationship between size and bond financing as can be seen from column 3 of Table 9. In line with the bank finance ratios in the last column of Table 8, only in the last pair is there a significant Table 9. Type Sl vs s2 Sl vs s4 s2 vs s3 s2 vs s4 s3 vs s4 (c) Results for firms Internal Equity - 0.64 (7) 0.5440 - 1.98 (7) p = 0.0877 - 1.66 (7) p = 0.1399 -4.25 (13) p = 0.0009 -3.52 (14) p = 0.0034 1.01 (14) p = 0.3298 1.28 (9) 0.2334 1.82 (7) p=o.1122 2.00 (7) p = 0.0855 1.00 (12) p = 0.3364 1.43 (9) p = 0.1876 0.47 (12) p = 0.6497 p = in the UK business monitor survey The period ratioI presented in Table 10 suggest that large UK firms use more internal and equity finance, with smaller firms using more bonds, bank borrowings and trade credit. The t-ratios presented in the same table show, however, that only the differences in equity and trade credit are statistically significant. The net equity-investment t-ratio indicates that large firms use significantly more equity than smaller firms, but the gross equity-finance t-ratio is only T-test results for comparisons of annual finance ratios across firms S4), ICICI data set, 1982-9p p = Sl vs s3 difference. Finally, firms in S2 use more trade credits than larger firms, but there is no significant difference in any of the other comparisons.” -2.27 (12) 0.0423 -4.05 (11) p = 0.0019 -4.65 (7) p = 0.0023 - 1.62 (14) p = 0.1285 -3.26 (9) p = p = 0.0099 -2.23 (10) p = 0.0499 in different size groups (Sl, S2, S3 and Bank borrowings Trade credit -1.21 (11) 0.2523 -0.24 (10) p = 0.8121 - 1.79 (11) p = 0.1004 1.16 (8) p = 0.2789 0.06 (9) p = 0.9534 - 2.05 (14) p = 0.0601 -0.47 (10) p = 0.6492 0.86 (8) p = 0.4141 1.48 (9) p = 0.1732 2.28 (13) p = 0.0399 3.18 (14) p = 0.0066 1.19 (14) p = 0.2543 p = “Degrees of freedom are given in parentheses. A negative value for the c-ratio implies that the latter group of firms in a pair (such as Sl vs S3) has a higher mean for the concerned ratio. Table 10. zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA Period ratios of (gross) .sources to total sources of finance and (net) sources to phy sical investment for business monitor - company finance - data, United 5 Kingdom, 1982- 90 Large Source Finance firms Small firms Investment Finance All firms Investment Finance t-tests for large vs small firms comparison” 0.49 0.61 0.34 0.64 0.44 0.62 1.09 (13) p = 0.2951 Equity 0.12 0.10 0.06 - 0.02 0.10 21 Investment Finance Internal [ 0.07 Bonds 0.08 0.12 0.11 0.22 0.09 0.14 Bank borrowings 0.11 0.07 0.13 0.11 0.11 0.08 Trade credit 0.20 0.10 0.36 0.01 0.25 0.07 1.61 (14) p = 0.1286 -0.23 (16) p = 0.8195 -0.64 (14) p = 0.5308 - 1.82 (16) p = 0.0876 Investment 0.48 (18) p = 0.6359 1.98 (11) p = 0.0730 -0.91 (11) p = 0.3810 -0.90 (12) p = 0.3401 1.13 (10) p = 0.2854 f G 2 g 2 Q 2 8 2 Sample: 2000 largest industrial and commercial “Degrees of freedom are given in parentheses. concerned ratio. companies; A negative 1 in 300 other registered companies in the United Kingdom, randomly chosen (see Appendix value for the t-ratio implies that the latter group of firms (small firms here) has a higher A). mean for the $ R v1 1042 WORLD DEVELOPMENT zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA tions set out in Section 2 of the paper. These results can be rationalized only by recognizing that the type of equity financing undertaken by the smaller Indian firms must be of a different nature from that which implicitly underlies the orthodox economic literature on corporate financing in developed countries. That literature deals essentially with the case of an already-listed firm choosing between issuing equity or bonds or obtaining funds in some other way. But all of the smaller, and many of the larger, firms in our samples are likely to be unlisted firms.lJ In the RBI sample, for example, the mean paid-up capital of a small firm is about Rs. 0.16 crores and that of a large firm is about Rs. 0.84 crores, which are well below the Rs. 3 crores requirement for listing. (d) zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA Discussion of the micro-data results For these firms, most of the equity issued is likely to be sold to existing owner/directors of The evidence presented in the last three the firms or to members of their family networks, subsections suggests that in the United Kingdom rather than to unknown members of the public.15 large firms use equity finance more than small In that case, the agency costs (increased scope firms, while the Indian samples of data show the for managerial discretion) associated with equity opposite. To summarize the results of the issues are minimal, and issuing equity does not (admittedly limited) test we carried out in give an adverse signal: equity finance is therefore Sections 4a and 4b: (i) large firms in the RBI more attractive to both the firms and their sample seem to use more internal finance and financiers. bonds than smaller firms, while the latter report At this point it is worth drawing attention to higher bank loans and total borrowings than the the only comparable work on a developing former; (ii) there seems to be no significant country, by Cho (1995) who collected data on difference between the equity ratios; (iii) large gross and net sources of finance for Korea. Cho firms in the ICICI seem to use more internal used three sets of data, covering respectively the finance and bonds than smaller firms (three of whole of the corporate sector (including governthe pair-wise comparisons for internal ratios are ment-owned firms), the private manufacturing statistically significant), while the latter make use sector, and a sample of around 400 companies of more trade credit; (iv) the smallest group of which were listed on the stock exchange by the firms seems to have a significantly higher ratio end of the 1980s. Korea is a more extreme case for equity financing than the largest group of in Singh’s (1995) analysis, which reports 19.5% firms. of net asset growth as coming from internal The UK findings and results (i) and (iii) for finance, 30.9% from long term debt, and 49.6% the Indian firms are, in broad terms at least, from equity finance. Cho found, however, that on consistent with theoretical expectations: larger the Mayer-type methodology equity finance was firms have a higher average age and thus have much less important: 12.8% in gross terms and longer and better reputations than smaller firms, 16.4% in net terms for the whole corporate which enables them to finance growth from sector: 7.7% gross for the manufacturing sector greater retentions and to access the bond market and 9.8% gross for the firm sample (for both of more easily, at lower cost; smaller firms, on the the latter a separate net term for equity is not other hand, have a lower average age (and/or available). Moreover, if Cho’s figures for the they are growing faster) which reduces their latter two sets of data are compared with the low ability to access the stock market for long term internal finance countries referred to in Section 3 funds or to use retained profits (which are lower above, Korea looks broadly similar to them relative to their investment). (though the similarity is less marked for the Results (ii) and (iv) for the Indian firms, which whole corporate sector). In addition, Cho is able show either no significant difference between the equity-finance ratios (for the RBI firms) or a to divide zyxwvutsrqponmlkjihgfedcbaZYXWVUTSR his second and third sets of data so as negative relationship between size and equityto compare large and small firms, domestic and finance ratios (for the ICICI firms), are not export firms, heavy and light industry, and other consistent, however. with the theoretical expectapairs. The main finding there is that large firms marginally significant. The gross trade creditfinance t-ratio is significant, implying that small firms use more of this source than large firms. For bank borrowings and bonds, the differences are not statistically significant.” The trade credit difference disappears in net terms (last row in Table lo), implying that smaller firms give, as well as receive, more trade credit than larger firms. The next section compares the evidence from this data set with the results from the Indian micro data sets and attempts to explore the possible reasons for the difference between the behavior of Indian firms and that of their UK counterparts in terms of the relative importance they assign to the various sources of financing. CORPORATE FINANCE IN DEVELOPING use more internal finance and more bond finance than small firms, but there is also some tendency for small firms to use more equity finance. Thus in a number of respects Cho’s findings for Korea parallel our findings for India.16 5. CONCLUSIONS In this paper we have presented sector and firm-level data on the sources of finance for non financial Indian firms, in an attempt to put the Singh and Hamid (1992) and Singh (1995) findings in context. Singh himself has argued that his results reflect period and country specific circumstances, rather than generic characteristics of developing countries. While we find his explanation attractive at the level of the firms concerned, we think it is difficult to assess at this point in time. Instead we have produced data for India directly comparable to that most widely used to analyze the financial systems of developed countries. We have shown that Singh’s methodology naturally finds a greater role for equity finance than the Mayer type methodology, and that the structure of corporate finance in India is not that much different from what economists might expect, once the specific characteristics of the economy are taken into account. Indeed, we would suggest that, when allowance is made for the lower rate of acquisition of financial assets by firms, India is broadly comparable to those developed countries such as France and Italy which have relatively small stock markets (with little or no market for corporate control), large sectors of medium and small-sized companies, and a banking system which lends substantial amounts to companies but does not have very close ties with firms and cannot exert the influence and control over them typical of Japanese banks.” COUNTRIES 1043 At the same time we have shown that Singh’s results cannot be ascribed merely to sample selection bias arising from the focus on the largest firms, since the rest of the Indian corporate sector issues large amounts of equity. It is clear, however, that these equity issues are undertaken in large part by non-listed firms and in that case they do not have the same characteristics as the equity issues by large listed firms which are the typical focus of analysis in the corporate financing literature. Thus, the question of the social gains from investment in the establishment and development of formal stock markets is rather more complex than some writers have suggested. For large firms where the ownership of capital is already dispersed, at least in part (even if there are some large stable shareholdings), so that information asymmetries and agency costs are already and inevitably substantial, and where at the same time the capital needs of the firms are significant, then the development and improvement of stock markets may be beneficial in improving the supply of finance. But for the mass of small and mediumsized firms which, because their ownership is not dispersed, benefit from lower information asymmetries and lower agency costs, it may be better to encourage equity finance of the existing form, that is, through informal networks of family, friends and business contacts. If incentives are to be given for firms to obtain or investors to provide equity finance, these incentives should apply to both sorts of firms. At the very least, there should be no extra incentive to firms to move from the second category to the first, other than the increased need for capital as firms expand. Before more definite policy conclusions are drawn, however, there is clearly a need for a rigorous firm-level analysis of the financing behavior of a representative sample of firms. Such an attempt will be made in our future work. NOTES 1. For example, Gertler (1988) reviews the literature on the linkages between the financial system and the real economy, including the impact of macroeconomic policies. Sundararajan (1987) finds that the high debtequity ratios in Korean firms in the 1960s and 1970s weakened the effectiveness of interest rate policies. 2. For some countries, data on less than 50 such companies were available, and in some countries a shorter time period had to be used. 3. In practice it may increases in own capital be difficult which arise to distinguish from valuation changes equity. 4. from Particularly those which in countries arise from new issues where takeovers of are rare. 5. The net sources methodology is also less vulnerable to the valuation problems that affect international comparisons of leverage (see, for example, Rajan and Zingales, 1995). 6. More reliable data for these two countries can be found in Cobham and Serre (1996) and Cobham et al. (1996) who put forward the idea of a “Meditteranean” model to characterize these two countries. but the conclusions are unaltered. 1044 WORLD DEVELOPMENT 7. We can also calculate ratios of the gross sources to physical investment. For cumulative ratios such as the gross average of the annual ratios (or period ratios), the ratio of Total Gross Sources of Finance to Physical Investment is all that is needed to calculate the Gross Sources of Finance to Investment ratios, since the latter are scaled up versions of the former, where the scaling factor is the Total Finance to Physical Investment ratio. We have not therefore presented these additional ratios here. 8. Let x and y be two variables. The (t-test) statistic for testing the equality of the means of these two variables, p1 = A, when the variances of x and y, err and gY, are unknown and cr, # o,, is given by: s, st -+r rl1 41 The result is distributed as Student’s t with v degrees of freedom, where v is given by the closest integer to (&I,)’ -+1<+1 (&I$ 9,+1 See Hoe1 (1984, pp. 140-161) for an introduction explanation of the calculation of these tests. and 9. The t-ratio in the last cell of the last column of Table 7 suggests that there is a significant difference in the ratio of total sources of finance to investment between small and large firms. We can also test for differences in the ratios of each source (gross, in this data set) to investment, and this would yield different results insofar as the year-to-year movements of total sources and investment are different. The results for these tests (not reported here) are broadly similar to those in Table 7, except for the ratios of total borrowings to investment, which are not significantly different between small and large firms. 10. These results are the opposite of Singh and Hamid’s (1992) finding of a negative relationship between size and internal financing of growth. They report that the smallest 25% of the firms financed about half their growth through savings while for the largest 25%’ of the firms the ratio was less than a third. 11. While the patterns revealed in the t-tests for investment ratios are broadly similar to those for the gross finance ratios discussed so far, there are some differences: in terms of internal finance ratios, the largest firms do not necessarily have a higher internal finance to investment ratio than the smaller ones, and in terms of equity financing, there is no significant size difference across the firms (whereas for finance ratios, the smallest firms had a higher mean ratio than the largest firms). Also, the investment t-ratios do not reveal any consistent pattern in the proportion of total finances invested by firms in the different groups. These investment ratios are not presented here since they are not based on net sources of financing. 12. See Appendix A for a precise definition of period ratios. Period ratios are presented here since the sample size has remained the same in the UK data, whereas in the RBI and ICICI data sets, there have been changes in the sample size across the years. 13. Singh and Whittingtons (1968) study of UK companies, covering 1948-60, did not find any significant relationship between size and other corporate characteristics. A later study by Kumar (1984) on large UK quoted companies, covering 1972-76, found a positive relationship between size and long-term debt. On differences in equity financing, see also Hughes (1994) who examined the Business Monitor data used here in more detail for 1987-89. 14. The rules for listing of securities of a company on stock exchanges laid down by the Department of Economic Affairs, Ministry of Finance of the Government of India, stipulate that the following criteria will apply: (i) the minimum issued capital of a company shall be Rs. 3 crores. and (ii) \ , the minimum oublic offer of equity capital shall not be less than Rs. 1.8 crores. In addition, the Securities Contracts (Regulation) Rules, 1957, have additional restrictions on the percentage of issued capital apportioned for public offer, based on whether a company is an established or a new one. On this and other laws pertaining to the regulation of capital issues, see “Capital Issues, SEBI and Listing” by Chandratre et ul. (1992). 15. Very broad estimates on the stock-holding pattern are as follows: controlling management/family: 25-30%; institutions (public sector inst&tion, mutual funds etc): 35-45s: uublic (free tloat): 35-400/o. See George (1994). In the unlisted small’ tirm case, the family/management fraction is likely to be substantially higher, because there are no outside (public) shareholders. 16. Findings reported by Sak (1995) on the various sources of external finance for Turkish firms also suggest that bank credit was a much more important source of finance than equity. 17. India has had “lead bank” systems since the 1960s but these institutions and the context in which they operate are very different from those of the Japanese main bank system (see Bhatt, 1994). 1045 zyxwvutsr CORPORATE FINANCE IN DEVELOPING COUNTRIES REFERENCES Atje, R. and Jovanovic, B. (1993) Stock markets and Kumar, M. S. (1984) Growth, Acquisition and Investdevelopment. zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA European Economic Review 37, ment. Cambridge University Press, Cambridge. 632- 640. Mayer, C. (1988) New issues in corporate finance. Atkin, M. and Glen, J. (1992) Comparing corporate European Economic Review 32, 1167- 1189. capital structures around the globe. The International Mayer, C. (1990) Financial systems, corporate finance Executive 34(5), 369- 387. and economic development. In Asy mmetric InformaBhatt, V. (1994) The lead bank system in India. In The tion, Corporate Finance and Investment, ed. R. Japanese M ain Bank System: Its Relevance for Hubbard. National Bureau of Economic Research, Developing and Transfoming Economies, ed. M. Abki Cambridge, MA. and H. Patrick. Oxford University Press, Oxford. Mayer, C. (1994) Stock markets, financial institutions Chandratre, K. R., Israni, S. D., Acharya, B. S. and and corporate performance. In Capital M arkets and Sethuraman, K. (1992) Capital Issues, SEBI and Corporate Governance, ed. N. Dimsdale and M. Prevezer. Clarendon Press, Oxford. Listing with Guidelines, Clarifications, Circulars, Rules etc. Bharat Publishing House. New Delhi. Myers, S. (1984) The capital structure puzzle. Journal of Finance 39, 575- 592. Cho, Y.-D. (1995) FLancial ‘factors and corporate Nagaraj, R. (1996) India’s capital market growth: investment: A microeconomic analysis of Korean Trends, explanations and evidence. Mimeo, Indira manufacturing firms. D.Phil. thesis, University of Gandhi Institute of Development Research, Oxford. Bombay. Cobham, D., Cosci, S. and Mattesini, F. (1996) The Rajan, R. and Zingales, L. (1995) What do we know Italian financial system: Neither bank-based nor about capital structure? Some evidence from intermarket-based. Mimeo, University of St Andrews, St national data. Journal of Finance 50, 1421- 1460. Andrews, UK. Reserve Bank of India (various years) Annual Reports. Cobham, D. and Serre, J.-M. (1996) A characterization Reserve Bank of India, New Dehli. of the French financial system. Mimeo, University of Reserve Bank of India (various years) Report on St Andrews, St Andrews, UK. Currency and Finance. Reserve Bank of India, New Corbett, J. and Jenkinson, T. (1994) The financing of Dehli. industry, 1970-89: An international comparison. Sak, G. (1995) Istanbul stock exchange and the CEPR Discussion Paper no 958, CEPR, Oxford. corporate sector: Reflections on the Turkish experiCorbett, J. and Jenkinson, T. (1997) How is investment ence in financial liberalization. In Economic financed? A study of Germany, Japan, UK and US. Research Forum for the Arab Countries, Iran and Manchester School. Turkey, Development of Financial M arkets in the Fry, M. (1988) M oney, Interest and Bunking in Arab Countries, Iran and Turkey, Cairo. Economic Development. Johns Hopkins University Samuel, C. (1996) The stockmarket as a source of Press, Baltimore, MD. finance: A comparison of U.S. and Indian firms. Fry, M. (1995) M oney, Interest and Banking in Policy Research Working Paper 1592, The World Economic Development, 2nd edn. Johns Hopkins Bank, Washington, DC. University Press, Baltimore, UK. Singh, A. (1995) Corporate financial patterns in indusGeorge, P. L. (1994) The Handbook of Emerging trializing economies: A comparative international M arkets: A County-by-Country Guide to the W orld’s study. International Finance Corporation Technical Fastest Growing Economies. Probus Publishing, Paper No. 2, IFC Washington, DC. London. Singh, A. and Hamid, J. (1992) Corporate financial Gertler, M. (1988) Financial structure and aggregate structures in developing countries. International economic activity. Joumul of M oney, Credit and Finance Corporation Technical Paper No.1, IFC, Banking. Washington, DC. Hoel, P. G. (1984) Introduction to M athematical StatisSingh, A. and Whittington, G. (1968) Growth, Profitatics, 5th edn. John Wiley and Sons, New York. bility and Valuation. Cambridge University Press, Hughes, A. (1994) The “problems” of finance for Cambridge, UK. smaller businesses. In Capital M arkets and Corporate Sundararaian. V. (1987) The debt-eauitv ratio of firms Governance, ed. N. Dimsdale and M. Prevezer. and the effectiveness of interest rate policy: Analysis Clarendon Press, Oxford. with a dynamic model of saving, investment, and Industrial Credit and Investments Corporation, India growth in Korea. International M onetary Fund Staff (various years) Financial Performance, Companies - ICICI portfolio. ICICI, New Dehli. Papers 34(2), 261- 310. II I I * , 1046 WORLD DEVELOPMENT APPEN ‘DIX A MACRO-LEVEL DATA FOR INDIA AND does not add up to total physical investment, as it should, most likely as a result of incomplete FIRM-LEVEL SAMPLE DATA FOR INDIA data. Therefore we defined a residual equal to AND THE UNITED KINGDOM zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIH investment - (internalfnet bank physical finance+net market finance). The “other” category in panel (b) of Table 5 refers to this (a) Flow-of-funds data for India residual source of finance. Of this ratio, 12.3% is accounted for by the data reported in the Report Data on internal finance and investment for India are taken from National Accounts Statistics on Currency and Finance, and only 0.1% is not reported for the whole period. For 1980-86, and the data on the other sources of finance are derived from a financial flows table in the however, only 12.3% is reported and 21.4% is not reported, and the corresponding figures for Reserve Bank of Indias Report on Currency and 1987-92 are 12% reported and -18.2% not Finance. The flow of funds table for India does reported. not provide data for equity and bond issues The period ratios presented are given by: separately, hence the gross figures reported in the paper (Table 5) are for corporate securities (equity+bonds = “market finance”) and the net figures are for issues of corporate securities minus acquisitions of all (corporate and public where I : denotes the amount of finance of type sector) securities. The broad categories of financial claims j, say internal or equity, in year t (measured in presented in the flow-of-funds tables are: (i) current prices of year t), Z, denotes total finance in period t (that is, the sum of the different types Currency; (ii) Deposits bank and other deposits; (iii) Investments in all forms of securiof finance in each year) and P, denotes the price ties; (iv) Loans and advances; (v) Small savings; index in year t - here, the GDP deflator, with (vi) Provident funds; (vii) Life funds; (viii) 1990 as the base year. For physical investment Compulsory deposits; (ix) Trade credit/debt; (x) ratios, It was replaced by physical investment in Foreign claims not elsewhere classified; and (xi) year t. Other items not elsewhere classified. It should be noted that for the Indian non-financial private sector, both the sources and uses column in the (b) Cumulative firm- level data flow-of-funds tables were nil for categories (v), (vi), (vii) and (viii). This could be a reason for We use three different (published) sources of the presence of large residuals in the data (see cumulative firm-level data. The Report on Currency and Finance published by the Reserve below). Paid-up capital does not include forfeited Bank of India (RBI) provides a database, with the sample size varying over the years (upward shares which are added to reserves and surplus. Borrowings of the private non-financial sector of 1650 firms), for two size-groups (large and are from the organized as well as from the small, based on the amount of paid-up capital). unorganized sectors. Trade dues and other The cumulative balance sheet data presented in the RBI reports cover a sample of varying sizes current liabilities include: sundry credit, liabilities to subsidiary and holding companies, interest on from 1975 to 1976 until the latest available year, loans and unclaimed dividends, deposits from 1990-91. The most recent Report on Currency customers, agents etc., and others. Loans and and Finance (1993- 94) actually provides cumulaadvances include loans to subsidiary companies, tive balance sheet data for 1991-93, but only for companies under the same group and holding large companies. 1990-91 is the latest year for companies. which data for both small and large firms are Net sources of finance are defined as follows: available. The average number of large firms bank finance = (sources of loans and advances (each with a paid-up capital greater than Rs. 1 -uses of loans and advances)+(sources of crore) is 580, and the average number of small currency and deposits -uses of currency and firms (each with a paid-up capital greater than deposits. Market finance and the “other” Rs. 5 lakhs but less than Rs. 1 crore) is 1,300. categories are obtained by netting out the uses New firms get added to the sample each year. from the corresponding sources. The sum of These data cover the main sources of finance for these three and the internal sources, however, firms, but do not cover the firms’ acquisitions of CORPORATE FINANCE IN DEVELOPING COUNTRIES 1047 financial assets, so that only an approximation of size varies in each of these groups across the the MCI gross sources presentation is possible. years. The average numbers of firms in the The Industrial Credit and Investments groups are: 85 in Sl, 175 in S2, 90 in S3 and Corporation of India (ICICI) maintains a about 65 in S4. We utilize these data to study the database on a sample of firms of various sizes relationship between size and the relative (based on gross fixed assets). We use these data importance of each source of financing. (taken from various issues of the report zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHG Financial In the United Kingdom, the Business M onitor Pet$ormance of Companies, ICKY Portfolio) for surveys (MA3) on Company Finance provide a 1980-92. This database provides all the balance consistent source of data from our point of view sheet information, in aggregative form, for a for 1982-90. The data are divided into two sample of 417 firms for 1980-88 and 620 firms strata: the large firms group is comprised of the 2,000 largest firms, and the small firms group is a for 1988-92. The firms in this sample accounted sample (just over 1,000 firms) consisting of 1 in for about 45% of the total paid-up capital and 300 of all the other Great Britain registered about 53% of the gross value added in the companies which file accounts selected at Indian private corporate sector as a whole for random. For large firms, size is measured using 1990-91. Like the RBI data set, however, the capital employed, the cut-off being three million ICICI data set does not allow us to calculate the GB pounds or more. Capital employed is defined net sources of financing. The cumulative balance as the sum of all items which finance net assets sheets of the ICICI data are also presented for (such as shareholders’ interest, minority sharefour different size groups, based on gross fixed holders’ interest), deferred taxation, long-term assets at the end of each financial year, and for loans (including debentures and mortgages) plus profit and loss making firms separately. Firms are classified into: Size 1 (Sl) - gross fixed bank loans and overdrafts, short-term loans and indebtedness to directors and group members, assets below Rs. 5 crores, Size 2 (S2) - gross fixed assets between Rs. 5 crores and Rs. 20 less amounts due from group members. This data set provides detailed information on both crores, Size 3 (S3) - gross fixed assets between sources and uses of funds, enabling us to Rs. 20 crores and Rs. 50 crores and Size 4 (S4) calculate net sources of funds for our analysis. - gross fixed assets above Rs. 50 crores. The average firm size in S4 is about 200 times as We present ratios for both gross and net sources. large as the average firm size in Sl. The sample