www.elsevier.com/locate/worlddev World Development Vol. 30, No. 2, pp. 227 253, 2002 2002 Elsevier Science Ltd. All rights reserved Printed in Great Britain 0305-750X/02/$ - see front matter PII: S0305-750X(01)00110-3 Creating an Environment for Venture Capital in India RAFIQ DOSSANI Stanford University, California, USA and MARTIN KENNEY * University of California, Davis and Berkeley Roundtable on the International Economy, USA Summary. The institution of venture capitalism is a di cult one to initiate through policy intervention, particularly in developing countries with unstable macroeconomic environments and histories of state involvement in the use of national capital and in the composition of production. India has all these constraints. The emergence of a thriving software services industry after 1985 created the raw material that venture capital could nance, thus achieving a critical precondition for venture capital s growth. It was followed by e orts to create a venture capital industry. After several setbacks, some success has been achieved largely due to a slow process of moulding the environment of rules and permissible institutions. The process was assisted by the role of overseas Indians in <a href="/keyword/silicon-valley/" >silicon valley</a> s success in the 1990s. Yet, in terms of what is needed, most of the work remains to be done. Inevitably, this will be the result of joint work by policymakers and practitioners. 2002 Elsevier Science Ltd. All rights reserved. Key words venture capital, India, entrepreneurship 1. INTRODUCTION In the last decade, one of the most admired institutions among industrialists and economic policymakers around the world has been the US venture capital industry. A recent OECD (2000) report identi ed venture capital as a critical component for the success of entrepreneurial high-technology rms and recommended that all countries consider strategies for encouraging the availability of venture capital. With such admiration and encouragement from prestigious international organizations has come various attempts to create an indigenous venture capital industry. This paper examines the e orts to create a venture capital industry in India. The possibility and ease of crossnational transference of institutions has been a subject of debate among scholars, policy makers, and industrialists during the entire 20th century, if not earlier (e.g., Kogut &amp; Parkinson, 1998). National economies have particular path-dependent trajectories, as do their national systems of innovation (NIS). 1 The forces arrayed 227 against transfer are numerous and include cultural factors, legal systems, entrenched institutions, and even lack of adequately trained personnel. Failure to transfer is probably the most frequent outcome, as institutional inertia is usually the default option. In the transfer process, there is a matrix of possible interactions between the transferred institution and the environment. There are four possible interactions: (a) The institution can be successfully transferred with no signi cant changes to either the institution or the environment. * Ra q Dossani is grateful to participants at the SEBI workshop on venture capital held in Mumbai, India, in August 1999 and to participants at Fair eld University s South Asia Conference held in November 1999 for their comments on an earlier version of this paper. Martin Kenney would like to thank Frank Mayadas of the Alfred P. Sloan Foundation for supporting his portion of this research. The authors thank Vijay Angadi, Kyonghee Han, Matthew C.J. Rudolph, and Lawrence David Saez for valuable comments. Final revision accepted: 27 September 2001. 228 WORLD DEVELOPMENT (b) The institutional transfer can fail. (c) The institution can be modi ed or hybridized so that it is able to integrate into the environment. (d) An interaction between the existing institutions and venture capital to modify the environment occurs. Though (a) and (b) are exclusionary, it is possible for transfer to yield a combination of (c) and (d). 2 The establishment of any institution in another environment can be a di cult trial-anderror learning process. Even in the United States, state and local government policies to encourage venture capital formation have been largely unsuccessful, i.e., Interaction (b) (Florida &amp; Smith, 1993). Similarly, e orts in the 1980s by a number of European governments to create national venture capital industries also failed. Probably the only other country to develop a fully <a href="/keyword/silicon-valley/" >silicon valley</a> -style venture capital industry is Israel. Taiwan, perhaps, is another country that appears to have developed a venture capital industry, though there has been little research on the dynamics of this process in Taiwan. Given the general di culties in more wealthy and developed countries, it would seem that India would have poor prospects for developing a viable venture capital community. India is a signi cant case study for a number of reasons. First, in contrast to the United States, India had a history of state-directed institutional development that is similar, in certain ways, to such development in Japan and Korea, with the exception that ideologically the Indian government was avowedly hostile to capitalism. Furthermore, the government s powerful bureaucracy tightly controlled the economy, and the bureaucracy had a reputation for corruption. Such an environment would be considered hostile to the development of an institution dependent upon a stable, transparent institutional environment. India did have a number of strengths. It had an enormous number of small businesses and a public equity market. Wages were low, not only for physical labor, but also for trained engineers and scientists, of which there was a surfeit. India also boasted a homegrown software industry that began in the 1980s, and became visible upon the world scene in the mid-1990s. Experiencing rapid growth, some Indian software rms became signi cant successes and were able to list on the US NASDAQ. Finally, beginning in the 1980s, but especially in the 1990s, a number of Indian engineers who had emigrated to the United States became entrepreneurs and began their own high-technology rms. They were extremely successful, making them multimillionaires or even billionaires, and some of them then became <a href="/keyword/venture-capitalists/" >venture capitalists</a> or angel investors. So there was a group of potential transfer agents. For any transfer process, there has to be some match between the environment and the institution. In addition, there must be agents who will mobilize resources to facilitate the process, though these agents can be in the public or private sector. Prior to 1985, the development of venture capital in India was very unlikely. But the environment began to change after 1985, and continues to change. Even in the United States, venture capital is only a small component of the much larger NIS, and as such is dependent on many other institutions. In the United States and in India the development of venture capital has been a coevolutionary process. This is particularly true in India, where it remains a small industry precariously dependent upon other institutions, particularly the government, and external actors such as international lending agencies, overseas investors, and successful Indian entrepreneurs in <a href="/keyword/silicon-valley/" >silicon valley</a> . The growth of Indian venture capital must be examined within the context of the larger political and economic system in India. As was true in other countries, the Indian venture capital industry is the result of an iterative learning process, and it is still in its infancy. If it is to be successful it will be necessary not only for it to grow, but also for its institutional context to evolve. The paper begins with a brief description of the development of US, Israeli, and Taiwanese venture capital industries. Particular attention is given to the role of the state. The second and third sections discuss the Indian economic and nancial environment. This is followed by a brief overview of the Indian information technology (IT) industry and a section on the role of nonresident Indians (NRIs). These three sections provide the environmental context within which the Indian venture capital industry was formed. The actual development of the Indian venture capital is described chronologically. The rst period is when government and multilateral lending agencies are the primary actors and investors in the Indian venture capital industry. In the second period there is a gradual process of liberalization of the venture capital market and the entrance of more private <a href="/keyword/venture-capitalists/" >venture capitalists</a> particularly from the United States. The nal sections re ect upon the progress of the Indian venture capital industry, VENTURE CAPITAL IN INDIA 229 while also highlighting the institutional barriers to continuing expansion. 2. VENTURE CAPITAL AS AN INSTITUTION IN OTHER COUNTRIES 3 Investing in a edgling start-up rm is extremely risky, because of the high rate of failure among new rms, something Stinchcombe (1965) termed the liability of newness. In general, private banks are unwilling to lend money to a newly established rm, because of the lack of collateral and high risk of losing the principal. The alternative of nancing growth from retained earnings is feasible, however, for a start-up this can be very slow. Moreover, technology-based start-ups often experience a period of nancial losses prior to the generation of surpluses for reinvestment. Prior to WW II, the source of risk capital for entrepreneurs everywhere was either the government, government-sponsored institutions meant to invest in such ventures, or informal investors that usually had some prior linkage to the entrepreneur. After WW II, a set of intermediaries, the <a href="/keyword/venture-capitalists/" >venture capitalists</a> , emerged whose sole activity was investing in edgling rms that they believed were capable of rapid growth with a concomitant capital appreciation. From these modest beginnings, venture capital gradually expanded and became increasingly formalized. Moreover, the locus of the venture capital industry shifted from the East Coast to the West Coast (Florida &amp; Kenney, 1988a,b). By the mid-1980s, the ideal-typical venture capital rm was based in <a href="/keyword/silicon-valley/" >silicon valley</a> and invested largely in the electronics sector, with lesser sums devoted to medical technology. 4 During this process, the US over-the-counter market was formalized and technically upgraded into what is now known as the NASDAQ, which specialized in listing technology-based rms. After an evolutionary learning process, the ideal-typical institutional form for venture capital became the venture capital rm operating a series of funds raised from wealthy individuals, pension funds, foundations, endowments and various other institutional sources. Moreover, this form has now di used globally. The <a href="/keyword/venture-capitalists/" >venture capitalists</a> were professionals, often with industry experience, and the investors were silent limited partners. At present a fund generally operates for a set number of years (usually between seven and 10) and then is terminated. Normally, each rm manages more than one partnership simultaneously. Even though the venture capital rm is the quintessential organizational format, there are other vehicles, the most persistent of which have been venture capital subsidiaries of major corporations, nancial and non nancial. The venture capitalist invests in recently established rms believed to have the potential to provide a return of ten times or more in less than ve years. This is highly risky, and many of the investments fail entirely; however, the large winners are expected to more than compensate for the failures. In return for investing, the <a href="/keyword/venture-capitalists/" >venture capitalists</a> not only receive a major equity stake in the rm, but they also demand seats on the board of directors. By active intervention and assistance, <a href="/keyword/venture-capitalists/" >venture capitalists</a> act to increase the chances of survival and rate of growth of the new rm. Their involvement extends to several functions, such as helping to recruit key personnel and providing strategic advice and introductions to potential customers, strategic partners, later-stage nanciers, investment bankers and various other contacts (Florida &amp; Kenney, 1988a; Gompers &amp; Lerner, 1999). The venture capitalist therefore provides more than money, and this is a crucial di erence between venture capital and other types of funding. The venture capital industry has, more recently, specialized even by stages of growth: there are early or seed funds, mature-stage funds, and bridge funds. The venture capital process is complete when the company is sold through either a listing on the stock market or the acquisition of the rm by another rm, or when the company fails. For this reason, the venture capitalist is a temporary investor and usually a member of the rm s board of directors only until their investment is liquidated. The rm is a product to be sold, not retained (Kenney &amp; von Burg, 1999). The venture capital process requires that investments be liquidated, so there must be the possibility of exiting the rm. Countries that erect impediments to any of the exit paths (including bankruptcy) are choosing to handicap the development of the institution of venture capital. Most recently, Jeng and Wells (2000, p. 241) found that the single strongest driver of venture capital investing is the number of IPOs. They also nd that government policies can have a strong impact, both by setting the regulatory stage and by galvanizing investment during downturns. This is not to say that countries not providing exit opportunities will be unable to foster entrepreneurship, only that 230 WORLD DEVELOPMENT it is unlikely that venture capital as an institution will thrive. In fact, Black and Gilson (1998) argue that venture capital industries will be more vital in countries with stock marketcentered capital markets than in nations with bank-centered capital markets. 5 There has been much debate about the preconditions for venture capital. One obvious conclusion is that entrepreneurship is the precondition for venture capital, not vice versa; however, this is a misleading statement in a number of dimensions. At some level, entrepreneurship occurs in nearly every society, but venture capital can only exist when there is a constant ow of opportunities that have great upside potential. Information technology has been the only business eld that has o ered such a long history of opportunities. So entrepreneurship is a precondition, but not any type of entrepreneurship will do. 6 Moreover, venture investing can encourage and increase the proper type of entrepreneurship, i.e., successful <a href="/keyword/venture-capitalists/" >venture capitalists</a> can positively a ect their environment. In the US the government played a role in the development of venture capital, though for the most part it was indirect. This indirect role, i.e., the general policies that bene ted the development of the venture capital industry, was probably the most signi cant. To list some of the most important, the US government generally followed sound monetary and scal policies ensuring relatively low in ation; as a result, the nancial environment and currency were stable. US tax policy, though it changed repeatedly, has been favorable to capital gains, and several decreases in capital gains taxes may have had some positive e ect on the availability of venture capital (Gompers, 1994). The stock market, which has been the exit strategy of choice for <a href="/keyword/venture-capitalists/" >venture capitalists</a> , has been strictly regulated and characterized by increasing openness. This has created a general macroeconomic environment of nancial stability and openness for investors, thereby reducing the external risks of investing in high-risk rms. Put di erently, an extra set of environmental risks stemming from government action was minimized a sharp contrast to most developing countries during the last 50 years. Another important policy has been a willingness to invest heavily in university-based research. This investment has funded generations of graduate students in the sciences and engineering, and from this research has come trained personnel and innovations, some of which have resulted in the formation of rms that have been funded by <a href="/keyword/venture-capitalists/" >venture capitalists</a> . US universities, such as MIT, Stanford, and UC Berkeley, played a particularly salient role (Kenney &amp; von Burg, 1999; Saxenian, 1998). The most important direct US government involvement in encouraging the growth of venture capital was the passage of the Small Business Investment Act of 1958 authorizing the formation of small business investment corporations (SBICs). The legislation was not aimed at encouraging venture capital per se. It meant to create a vehicle for funding small rms of all types. The legislation was complicated, but for the development of venture capital the following features were most signi cant: It permitted individuals to form SBICs with private funds as paid-in capital. They could then leverage their paid in capital to borrow money up to a limit that increased over time with government guarantees. There were also tax and other bene ts, such as income and a capital gains pass-through and the allowance of a carried interest for the investors. The SBIC program also provided a vehicle for banks to circumvent the Depression-era laws prohibiting commercial banks from owning more than 5% of any industrial rm. The banks SBIC subsidiaries allowed them to acquire equity in small rms. This made even more capital available to edgling rms, and was a signi cant source of capital in the 1960s and 1970s. The nal investment format permitted SBICs to raise money in the public market. For the most part, these public SBICs failed and/or were liquidated by the mid-1970s. After the mid1970s, with the exception of the bank SBICs, the program was no longer important to the venture capital industry. The SBIC program experienced serious problems almost immediately. Starting in 1965 Federal criminal prosecution was necessary to rectify SBIC corruption. By one estimate, nine out of ten SBICs had violated agency regulations and dozens of companies had committed criminal acts (Bean, 2001). Despite the corruption, something valuable also occurred. Particularly in <a href="/keyword/silicon-valley/" >silicon valley</a> , several individuals used their SBICs to leverage their personal capital, and they were so successful that they were able to reimburse the program and raise institutional money to become formal <a href="/keyword/venture-capitalists/" >venture capitalists</a> . The SBIC program accelerated their capital accumulation, and as important, government regulations made these new <a href="/keyword/venture-capitalists/" >venture capitalists</a> professionalize their investment VENTURE CAPITAL IN INDIA 231 activity, which had been informal prior to entering the program. Now-illustrious rms such as Sutter Hill Ventures, Institutional Venture Partners, Bank of America Ventures, and Menlo Ventures began as SBICs. The historical record also indicates that government action can harm venture capital. The most salient example came in 1973 when the US Congress, in response to widespread corruption in pension funds, changed federal pension fund regulations. In their haste to prohibit abuses, Congress passed the Employment Retirement Income Security Act (ERISA) making pension fund managers criminally liable for losses incurred in high-risk investments. This was interpreted to include venture capital funds. The result was that pension managers shunned venture capital, nearly destroying the entire industry. This was only reversed after active lobbying by the newly created National Venture Capital Association (NVCA) in 1973 (Pincus, 2000; Stultz, 2000). Only in 1977 did they succeed in starting a process of a gradual loosening that was completed in 1982 (Kenney, n.d.). The reinterpretation of these pension fund guidelines contributed to a ood of new money into venture capital funds. In a sense, since the mid-1980s the development of the US venture capital industry has become a process of formalizing the institutions, a process that was even manifested in the contracts between <a href="/keyword/venture-capitalists/" >venture capitalists</a> and entrepreneurs (Suchman, 2000). After the mid1980s, the institution of venture capital had become a part of the US NIS with its own industry association, practices, and routines. So, in the United States venture capital as an institution emerged from an organic trial-anderror experimental process prior to maturing into today s powerful institution for the encouragement of new rm formation. In this process, the government played a limited and contradictory role. Since the mid-1980s there has been much interest in the transference of institution of venture capital to other countries. Important proponents of this transfer were the International Finance Corporation and various bilateral and multilateral development organizations including the Asian Development Bank, the West German Deutsche Entwicklungs Gesellschaft (DEG), and the British Commonwealth Fund among others (Kenney, n.d.). Their modus operandi was to invest in venture capital rms dedicated to funding start-ups in various nations. Though these organizations were not signi cant for the establishment of venture capital in Taiwan and Israel, foreign investors did play an important role in those two countries. The two most successful adopters of US-style venture capital practice are Taiwan and Israel, in both cases the national governments played a signi cant role in encouraging the growth of venture capital (see, respectively, Autler, 2000; Kenney, Han, &amp; Tanaka, 2001). There are a number of similarities between the two countries. They are smaller countries closely allied with the United States, that have sizable numbers of their citizens in the United States. Both governments have relatively good economic records and operate with a minimum of corruption. Both are also national defense states. Importantly, in both countries the government developed policies that allowed its <a href="/keyword/venture-capitalists/" >venture capitalists</a> to bene t nancially from their successful investments. Yet, despite these similarities, there are also a number of di erences between the industries in the two. The Israeli venture capital industry is closely related to the United States. The rst venture capital investments in Israel were made by the well-known US venture capitalist, Fred Adler, who began investing in Israeli startups in the early 1970s (Adler, 2000; Autler, 2000, p. 40). In 1985, a former Commander of the Israeli Air Force Dan Tolkowsky, joined with partner Fred Adler to form the rst Israeli venture capital fund, and it was capitalized at $25 million with much of the investment coming from the United States. The US connection was built upon Israeli and American Jewish networks that were able to mobilize large sums of capital and political support in the US government. Israel had three important sources of entrepreneurs: the military research establishment, its universities, and already existing rms. Moreover, there was a constant ow of entrepreneurs between the United States and, especially, <a href="/keyword/silicon-valley/" >silicon valley</a> and Israel. In addition, Israeli rms had begun listing on the NASDAQ in the 1980s, so the Tel Aviv stock market was not the only exit route. The creation of a venture capital industry (rather than a few rms) would wait, however, until the 1990s, when the Israeli government created a government-funded organization, Yozma, meant to encourage venture capital in Israel. Yozma received $100 million from the Israeli government. It invested $8 million in 10 funds that were required to raise another $12 232 WORLD DEVELOPMENT million each from a signi cant foreign partner, presumably an overseas venture capital rm (Autler, 2000, p. 44). Yozma also retained $20 million to invest itself. These sibling funds were the backbone of an industry that in 1999 invested in excess of $1 billion during 2001 in Israel (PriceWaterhouse, 2001). So, initially venture capital investing in Israel was entirely undertaken by the private sector (though often the investments were in spin-o s of public research). But, the government Yozma program created a hybrid of private investment with government support that supercharged the growth of venture capital. The critical point is that private venture capital existed before government involvement, as was true in the case of the US SBIC program. In the late 1990s government support became less signi cant. The government was able to play such a powerful catalyzing role, because there were supportive environmental conditions, and privatesector venture capital had already emerged. In other words, government support did not create the industry, but had an e ect on its growth. In the case of Israel, the institution of venture capital was largely an unproblematic transfer from the United States. The creation of the Taiwanese industry differs signi cantly from the Israeli case in some important respects. Taiwan did not have the same level of high-technology research as Israel as it was a developing country. Nor did it have the same level of relationship to the nancial and academic elite in the United States that Israel did. There were, however, a large number of successful Taiwanese and other Chinese scientists and engineers in the United States that it could and did draw upon (Saxenian, 1999). The inception of the Taiwanese venture capital industry can be traced directly to governmental initiative. In 1983 the then Finance Minister, Li-Teh Hsu, and a former Finance Minister K.T. Li took a study trip to the United States and Japan to study how new high-technology rms were formed, after this they decided to create incentives to catalyze the creation of a venture capital industry in Taiwan (Saxenian, 1998; Shih, 1996, p. 282). In 1983 legislation was passed giving attractive tax incentives to individuals willing to invest in professional venture capital rms. The most important feature of the 1983 legislation was an up to 20% tax deduction for Taiwanese individuals provided they maintained their venture capital investment for at least two years. One interesting and important feature of this e ort was that the Taiwanese <a href="/keyword/venture-capitalists/" >venture capitalists</a> were not forbidden from investing abroad as long as a bene t to Taiwan could be shown. Despite the attractive bene ts, investment grew only gradually. The rst venture capital rm in Taiwan was the Acer subsidiary, Multiventure Investment Inc., which was formed in November 1984 and made its rst investment in a <a href="/keyword/silicon-valley/" >silicon valley</a> startup that year (Shih, 1996, p. 35). But the venture capital fund that received the most attention was formed by H&amp;Q. The key person was Ta-Lin Hsu, a former executive in IBM s San Jose Laboratories. H&amp;Q Taiwan recruited investment from major Taiwanese industrial groups and had a total capitalization of roughly $25 million (Sussner, 2001). Its rst investment was in the Taiwanese subsidiary of Data Corporation, a Santa Clara manufacturer of disk drive controllers and oppy disks (Kaufman, 1986, p. 7D). In 1987, the San Francisco-based Asian-American venture capital rm, the Walden Group became the other important early foreign venture capital fund to invest in Taiwan. Its rst two investments were in Northern California (Besher, 1988, p. C9). This fund evolved into the Walden International Investment Group. Notice that both H&amp;Q and Walden s rst investments were in <a href="/keyword/silicon-valley/" >silicon valley</a> and this set the binational investment pattern that characterizes the Taiwanese venture capital industry. There is an important di erence between Taiwan and Israel, in the that Taiwan received little outside capital, whereas Israel was much more successful in attracting overseas capital. The Taiwanese venture capital industry differed from that in Israel and in the United States in one very signi cant way. Namely, Taiwan, quite simply, did not have the worldclass research capability of these other two countries. But, it honed its manufacturing expertise and found that its rms could grow very rapidly providing contract manufacturing services to US rms. The Taiwanese strategy was to enter product markets that had not yet been commoditized and transform them into commodities as it recently has done in notebook computers. Success in these markets made it possible for the Taiwanese rms to grow rapidly and conduct initial public o erings on either the Taiwan Stock Exchange or on NASDAQ. In addition, Taiwanese <a href="/keyword/venture-capitalists/" >venture capitalists</a> invested between 20% and 30% of their capital in <a href="/keyword/silicon-valley/" >silicon valley</a> , especially in rms started by Chinese engineers, thereby diversifying their portfolio. VENTURE CAPITAL IN INDIA 233 In this brief discussion it is impossible to describe fully the di erences between the organization and operation of the Taiwanese venture capital industry and that of the United States and Israel. 7 Even the investments are di erent, however, because Taiwanese <a href="/keyword/venture-capitalists/" >venture capitalists</a> invest in manufacturing rms. In contrast to the other two countries, the Taiwanese government continues to closely regulate the venture capital industry. Taiwanese <a href="/keyword/venture-capitalists/" >venture capitalists</a> are organized as operating rms compensating investors through dividends, rather than operating as partnership managers giving investors distributions. Thus in the case of Taiwan it is possible to say that the institution of venture capital was hybridized to t into the environment. The Israeli environment resembled the United States far more than it did Taiwan, and thus there was less need to change the institutional parameters of the venture capital industry when it was transferred. But, in both countries there were strong connections to <a href="/keyword/silicon-valley/" >silicon valley</a> that made the transfer of the social technology of venture capital investing possible. Not coincidentally, <a href="/keyword/venture-capitalists/" >venture capitalists</a> in both countries also used these connections to invest in <a href="/keyword/silicon-valley/" >silicon valley</a> deals. The governments in both countries played a signi cant role in catalyzing the development of the venture capital industry. Curiously, in historical terms both countries had had bank-centered nancial systems, but this changed roughly along with the development of the venture capital industry. 3. THE GENERAL INDIAN ENVIRONMENT From its inception, the Indian venture capital industry has been a ected by international and domestic developments; its current situation is the result of the evolution of what initially appeared to be unrelated historical trajectories. The creation of a venture capital industry in India through transplantation required the existence of a minimal set of supportive conditions. They need not necessarily be optimal, because, if the industry survived, it would likely set in motion a positive feedback process that would foster the emergence of successful new rms, encourage investment of more venture capital, and support the growth of other types of expertise associated with the venture capital industry. In other words, if the venture capital industry experienced any success, it could entrain a process of shaping its environment. In contrast to Israel and Taiwan, India would have to experience a hybrid of Interactions (c) and (d) discussed in Section 1, if venture capital was to thrive. It would have to change its environment to allow the sustenance of a venture capital industry and the societal institutions would have to modify themselves. Venture capital could begin with a suboptimal though minimally sustainable set of conditions, the venture capital industry could take root and shape its environment to a more optimal situation, while the institutions themselves would also need to change. Small and medium-sized enterprises have a long history and great importance to India. The leaders of the Independence movement were supporters of small businesses as an alternative to exploitation by multinational rms. Yet, despite the emphasis upon and celebration of small enterprises, the Indian economy was dualistic. It was dominated by a few massive private sector conglomerates, such as the Tata and Birla groups, and various nationalized rms, even while there was an enormous mass of small shopkeepers and local industrial rms. As anywhere else, these small rms were in traditional industries and were not relevant to the emergence of venture capital, but they do indicate a culture of private enterprise. This entrepreneurial propensity also has been demonstrated by the willingness of Indians emigrating in other countries to establish shops, restaurants, hotels and enterprises of all sorts. Indians valued education very highly at least since the colonial era. After Independence, the Indian government invested heavily in education, and Indian universities attracted excellent students. In the 1960s, the Ford Foundation worked with the Indian government to establish the Indian Institutes of Technology (IIT), which adopted MIT s undergraduate curriculum. These Institutes and other top Indian educational institutions very quickly became the elite Indian engineering schools with extremely competitive entrance examinations and to which only the best students could gain entry. Excellent Indian students were in demanded by overseas university graduate programs, generally, and in engineering, particularly. After graduating from overseas programs, many of these Indian students did not return to India. Many other Indian graduates remained in India, working in the large family conglomerates, 234 WORLD DEVELOPMENT the many Indian universities, and various toplevel research institutes such as those for space research (Baskaran, 2000). This meant that there remained in India a large pool of capable engineers and scientists that were underpaid (by global standards), and potentially mobile. Despite these strengths, India had many cultural rigidities and barriers to entrepreneurship and change, beginning with an extremely intrusive bureaucracy and extensive regulations. Until recently the labor market was quite rigid. For well-educated Indians the ideal career path was to enter the government bureaucracy, a lifetime position; enter the family business, which was then a lifetime position; or join one of the large conglomerates such as Tata and Birla, which also e ectively guaranteed lifetime employment. The nal career path was to emigrate; not surprisingly, among the immigrants were many seeking better opportunities and release from the rigidities at home. In summation, the institutional context discouraged investment and entrepreneurship. The next sections examine the features of the Indian economy that would evolve to make the creation of the Indian venture capital industry possible. 4. THE INDIAN FINANCIAL SYSTEM India has a large, sophisticated nancial system including private and public, formal and informal actors. In addition to formal nancial institutions, informal institutions such as family and moneylenders are important sources of capital. India has substantial capital resources, but as Table 1 indicates, the bulk of this capital resides in the banking system. In the formal nancial system, lending is dominated by retail banks rather than the wholesale banks or the capital markets for debt. The primary method for rms to raise capital is through the public equity markets, rather than through private placements. (a) The banking system Prior to independence from Britain, the banking system was entirely private and largely family-operated. In the pre-war period, the family-run banks often invested in new ventures. After Independence, the Reserve Bank of India (RBI) and the State Bank of India were nationalized, with the State Bank of India continuing to play the role of banker to government agencies and companies. Then, in 1969, the next 14 largest banks were nationalized. With the State Bank of India, the state controlled 90% of all bank assets. The nationalized banking system became an instrument of social policy. During 1969 91, the nancial position of the banks progressively weakened, due to loss-making branch expansions, ever-strengthening unions, oversta ng, and politicized loans. Until 1991, depositors were reluctant to use banks because although their savings were safe, the government set deposit interest rates below the rate of in ation. By 1991, the entire bank system was unpro table and nearing collapse. 8 The socialized banking system had other perverse e ects. For example, although the bank managers were civil servants and very risk-averse, they could o er below-market interest rates. This created excessive demand for funds, but, quite naturally, bankers extended the loans to their safest customers. These were primarily the large rms owned by the government, which operated the largest steel, coal, electrical, and other manufacturing industries. The other large bank borrowers were the giant family conglomerates such as the Tata and Birla group. This increased the group s economic power, but did not lead to economically e cient decisions about how to deploy capital. Table 1. The disposition of indian capital resources and their availability for venture investing in 1996 97 Type of funds Currency and bank deposits Government securities Life insurance funds Pension funds Privately held shares and debentures (including mutual funds) Total Billions of rupees 634.90 116.36 156.36 262.48 96.34 1266.44 Percentage of total 50.1 9.2 12.3 20.7 7.6 100 Percentage permitted for venture capital investment Up to 5% of new funds, since April 1999 None None None Some None Source: Statistical Outline of India 1988 99 (1999). VENTURE CAPITAL IN INDIA 235 Small rms were starved for capital. Thus the Indian banks provided no resources for entrepreneurial rms. (b) Equity markets The rst Indian stock markets were established during the British Raj era in the 19th century. During the early part of the 20th century, Indian equity markets actively nanced not only banking, but also the cotton and jute trades (Schrader, 1997). In 1989 there were 14 stock markets in India, though Bombay was by far the largest (World Bank, 1989). The socialization of the economy and particularly banking after independence reinforced the strength of the stock markets as a source of capital, and by the 1960s, India had one of the most sophisticated stock markets in any developing country. There were several reasons for the growth of the Indian stock market. Motivated by its egalitarian principles, the government supported the stock markets as an instrument for reducing the concentration of ownership in the hands of a few industrialists (an outcome of the government policy of providing below-market interest loans to the large family conglomerates). Second, the government industrial licensing policy instituted in 1961 meant that businesses had to apply for government permission to establish new ventures. Permissions were given only in the context of the Sovietstyle national plans for each sector. There was a strong element of favoritism in who received permission. Most important, due to government central planning controls, shortages were endemic, and thus, permission to produce was a guarantee of pro ts. The distortion these policies created by encouraging concentration were meant to be o set by RBI stipulation that private sector borrowers could not own more than 40% of the rm s equity if they wished to receive bank nance. In 1973 the government required all foreign rms to decrease ownership in their Indian subsidiaries to 40%. Faced with a choice between selling stakes privately and listing on the stock exchanges, most rms chose the latter and issued new stock, which led to a large increase in public ownership of such companies. So, to raise money the private sector became reliant on stock markets. Investors, large and small, readily nanced ventures since the shortages induced by the planning system guaranteed a ready market for anything pro- duced. Curiously, the retention of 40% of the equity by the core investors meant that in reality they controlled the rm. The 40% regulation did not liberalize the markets as much as one might expect. Because loans were also necessary for rms and this required collateral in xed assets, new entrepreneurs were restricted to sectors with assetheavy projects. This disadvantaged the service sector, resulting in even greater concentration, and equity markets focused on nancing lowrisk projects. Moreover, the public s enthusiasm for rms operating within a licensed industry meant that it was di cult for other new rms to secure capital through listing on the stock exchanges. In 1991, as part of a large number of nancial reforms, the Securities and Exchange Board of India (SEBI) was created to regulate the stock market. At the time, there were 6,229 companies listed on all the stock exchanges in India (RBI, 1999, p. 16). The reforms and loosening of regulations resulted in an increase in the number of listed companies to 9,877 by March 1999, and daily turnover on the stock exchanges rose to 107.5 billion rupees (US$2.46 billion) by December 1999. One reform was the removal of a pro tability criterion as a requirement of listing. To replace the pro tability requirement, it was stipulated that a rm would be de-listed if it did not earn pro ts within three years of listing. This reform meant unpro table rms could be listed, providing an exit mechanism for investors. Not surprisingly, there was a dramatic increase in the listings of rms, many of which could be considered as high technology. In terms of experience, India contrasted favorably with most developing countries, which had small, ine cient stock markets listing only established rms. Even in Europe, until the creation of new stock markets in the mid-1990s, it was extremely di cult to list small hightechnology rms (Posner, 2000). But, although these stock markets provided an exit opportunity, they did not provide the capital for rm establishment. Put di erently, accessible stock markets did not create venture capital for startups; they merely provided an opportunity for raising follow-on capital or an exit opportunity. (c) Other institutional sources of funds India has a strong mutual fund sector that began in 1964 with the formation of the Unit 236 WORLD DEVELOPMENT Trust of India (UTI), an open-ended mutual fund, promoted by a group of public sector nancial institutions. Because UTI s investment portfolio was to consist of longer-term loans, it was meant to o er savers a return superior to bank rates. In keeping with the risk-averse Indian environment, initially UTI invested primarily in long-term corporate debt. But, UTI eventually became the country s largest public equity owner as well. This was because the government controlled interest rates in order to reduce the borrowing costs of the large manufacturing rms that it owned. These rates were usually set well below market rates, yet UTI and other institutional lenders were forced to lend at these rates. In response, rms started issuing debt that was partially convertible into equity in order to attract institutional funds. By 1985, the conversion of these securities led to UTI becoming the largest owner of publicly listed equity (UTI Annual Report, 1985). By 1991, the equity portion of the UTI portfolio had grown to 30% (UTI Annual Report, 1991). In 1992, in tandem with banking sector reform, permission to form privately-owned mutual funds (including foreign-owned funds) was granted, leading to a gradual erosion in UTI s then-dominant market share. Until April 1999, mutual funds were not allowed to invest in venture capital companies. Since then, the mutual funds have been allowed to commit up to 5% of their funds as venture capital, either through direct investments or through investment in venture capital rms. But, the mutual funds have not yet overcome their risk-averse nature and invested in venture capital, either directly or indirectly through investment in venture capital funds. Certainly, should the mutual funds decide to invest directly in rms, there would be operational issues regarding the capability of mutual funds to perform the venture capital function. The largest single source of funds for US venture capital funds since the 1980s has been public and private sector pension funds. In India, there are large pension funds but they are prohibited from investing in either equity or venture capital vehicles, thus closing o this source of capital. In summation, prior to the late 1980s, though India did have a vibrant stock market, the rigid and numerous regulations made it nearly impossible for the existing nancial institutions to invest in venture capital rms or in startups. Nearly all of these institutions were politicized, and the government bureaucrats operating them were risk-adverse. On the positive side, there was a stock market with investors amenable to purchasing the equity in fairly earlystage companies. It was also possible to bootstrap a rm and/or secure funds from friends and family if one was well connected. But, no nancial intermediaries comfortable with backing small technology-based rms existed prior to the mid-1980s. It is safe to say that little capital was available for any entrepreneurial initiatives. An entrepreneur aiming to create a rm would have to draw upon familial capital or bootstrap their rm. 5. THE INDIAN INFORMATION TECHNOLOGY INDUSTRY A viable venture capital industry depends upon a continuing ow of investment opportunities capable of growing su ciently rapidly to the point at which they can be sold yielding a signi cant annual return on investment. If such opportunities do not exist, then the emergence of venture capital is unlikely. In the United States and Israel such opportunities occurred most regularly in the information technologies. Moreover, in every country, with the possible exception of the United States, any serious new opportunity has to be oriented toward the global market, because few national markets are su ciently large to generate the growth capable of producing su cient capital gains. Since Independence, the Indian government strove to achieve autarky, and the protection of Indian markets and rms from multinational competition guided nearly every policy the information technology industries were no exceptions. For example, in the mid-1970s, the Indian government demanded that IBM reduce the percentage ownership of its Indian operations to 40%, but IBM refused and left India in 1978. After this the government gave the staterun Computer Maintenance Corporation a legal monopoly to service all foreign systems (Brunner, 1995 quoted in Lateef, 1997). IBM s retreat released 1,200 software personnel into the Indian market, and some of these established small software houses (Lateef, 1997). The protectionist policy had bene ts and costs. The bene t was that it contributed to the creation of an Indian IT industry; the cost was that the industry was backward despite the excellence of its personnel. After IBM s withdrawal, there was little further interest by multinational rms in the Indian market. Due to this lack of VENTURE CAPITAL IN INDIA 237 foreign investment and despite the presence of skilled Indian personnel, India was a technological backwater even while East Asia progressed rapidly. In 1984, in response to the failure of the government-run computer rms and the success of private rms such as Wipro, HCL, and Tata Consultancy Services, the Indian government began to liberalize the computer and software industry by encouraging exports (Evans, 1992). This was particularly timely because there was a worldwide shortage of software programmers. In 1986, Texas Instruments received government permission to establish a 100% foreign-owned software subsidiary in India. The capabilities of the Indian personnel, which was due, in large measure, to a wise decision in the 1960s by the government to increase its investments in education, soon attracted other foreign software engineering operations to India. Contemporaneously, entrepreneurs exploited the labor cost di erential to body-shop Indian programmers overseas. By 1989, this accounted for over 90% of software revenues (Schware, 1992 cited in Lateef, 1997). The body-shopping and the foreignowned engineering operations provided a conduit through which Indian engineers could learn about the cutting-edge software techniques and developments in the West and particularly <a href="/keyword/silicon-valley/" >silicon valley</a> . These activities created a network of contacts and an awareness of the state-of-the-art in global computing and software technology. The center of this activity was the South Indian city of Bangalore (Mitta, 1999). By the early 1990s, Indian rms also began the development of a market for o -site contract programming. This market grew impressively in the 1990s, as shown in Table 2. Moreover, the percentage of on-site contract programming revenues fell from 90% in 1988 to 45% in 1999 2000 (Nasscom, 1998). But, because an additional 35% of work is o -site contract programming, low value-added services remain dominant. High value-added next-stage businesses, including turnkey projects, consultancy and transformational outsourcing make up the balance, and branded product development for the export market is negligible. A few of these rms, particularly Infosys and Wipro, grew to be quite large both in terms of manpower and revenue. They were successful on the Indian stock market in the 1990s, and in the late 1990s they listed their stock on the US NASDAQ. Interestingly, in contrast to Taiwan, Korea, and, increasingly, China, the Indian hardware sector remains negligible, although there have been a few notable recent successes, such as Armedia and Ramp Networks. One small conglomerate, Wipro, established an IT spin-o business in Bangalore in 1980, and grew to become the most highly valued Indian IT company. Another leading Indian rm, Infosys, was founded 1981 and listed on the US NASDAQ. In August 2000 its valuation reached nearly $17 billion. These and other Table 2. Domestic and export sales of the Indian IT industry, 1994 99 in US$ millions 1994 95 Software Domestic Exports Total Hardware Domestic Exports Total Peripherals Domestic Exports Total Other Training Maintenance Networking and other Grand total 350 485 835 590 177 767 148 6 154 107 142 36 2,041 1995 96 490 734 1,224 1,037 35 1,072 196 6 202 145 172 710 2,886 1996 97 670 1,083 1,753 1,050 286 1,336 181 14 195 183 182 156 3,805 1997 98 950 1,750 2,700 1,205 201 1,406 229 19 248 263 221 193 5,031 1998 99 1,250 2,650 3,900 1,026 4 1,030 329 18 347 302 236 237 6,052 1999 2000 1,700 4,000 5,700 1,450 86 1,536 435 27 462 400 263 310 8,671 Source: http://www.nasscom.org 2001 238 WORLD DEVELOPMENT successes demonstrated that rms capable of rapidly increasing in value could be formed in India. This is despite the fact, as Arora and Arunachalam (2000) note, that the development of globally distributed packaged software programs, which is the source of the greatest pro ts, has not occurred. Indian software rms have an assured market for their contract programming work and, therefore, little incentive to take the risk of developing software packages (Naqvi, 1999). Finally, the largest rms, perhaps due to their visibility in overseas markets through their branch o ces, have leveraged their manpower resources to better advantage than small rms have, leading to an erosion in the share of small rms (Naqvi, 1999). By the early 1990s, India had an information technology industry that centered on software and drew upon the large number of relatively skilled Indian engineers. 9 The Indian IT industry was growing quickly, and newly formed rms had entered the industry; however, the opportunities had not yet led to the creation of any information technology-oriented venture capital rms. In addition, at the time there were not yet any knowledgeable information technology industry veterans in India who also had experienced the startup and venture capital process rst hand. In <a href="/keyword/silicon-valley/" >silicon valley</a> , however, an entire cadre of Indians was emerging who had such experiences. 6. NONRESIDENT INDIANS (NRIS) From the 1950s onward, bright, well-educated Indian engineers attended US universities and remained to work in high-technology rms. In the 1960s and 1970s, this was considered a brain drain. It seems likely, however, that even if these engineers had returned to India, there would have been few opportunities for them outside of government research organizations and the corporate research laboratories for rms such as Tata and Birla. For this and other reasons, many Indians remained in the United States and secured employment in universities and corporations, including <a href="/keyword/silicon-valley/" >silicon valley</a> electronics rms. Initially, these Indian engineers joined existing rms, but not surprisingly they were not immune to the attractions of entrepreneurship, especially in <a href="/keyword/silicon-valley/" >silicon valley</a> . The rst noteworthy group included Kanwal Rekhi, who cofounded Excelan, a data networking rm, with three other Indian engineers in 1981. Excelan later was purchased by Novell, leaving Rekhi and the other engineers with enormous capital gains. Another early entrepreneur was Vinod Khosla, who in 1982 had co-founded Daisy Systems, and was a driving force in the establishment of Sun Microsystems. After leaving Sun, he joined the prestigious venture capital rm, Kleiner, Perkins, Cau eld and Byers. Yet another Indian engineer, Dalal (2000), cofounded Metaphor in 1982, Claris in 1987, and in 1991 joined the top-tier venture capital fund, May eld Fund. Another successful early Indian entrepreneur was Suhas Patil, who cofounded Cirrus Logic in 1984. These NRIs not only were successful entrepreneurs, but they soon began investing in yet other startups. Quite naturally, the NRIs remained in contact with family, friends, and classmates in India. Moreover, by the late 1980s, the success of the NRIs came to the attention of Indian policymakers. Even while concern about the brain drain continued, policymakers recognized that it would be impossible to retain such highly skilled individuals if there were no opportunities for them in India. NRIs wished to assist India: they visited India and discussed their experiences in the United States, and expressed a willingness to invest in ventures their friends and classmates might launch in India. They quickly discovered, however, that it was not so simple to transform their willingness to invest capital into the reality of a reasonable investment. But this initiated a process through which the NRIs were re-conceptualized from being defectors to being a potential source of knowledge, connections, and even capital (Saxenian, 1999). From the perspective of creating venture capital, India had a stock market that with minimal e ort could handle public stock o erings from edgling high-technology rms. There was also a growing IT industry with some rms that experienced extremely fast growth. There was also a cadre of Indians familiar with the operation of the US <a href="/keyword/silicon-valley/" >silicon valley</a> , and there were su cient skilled engineers in India to sta startups. In other words, by 1990 the environmental preconditions for the successful establishment of a venture capital industry were in place. 7. VENTURE CAPITAL IN INDIA In the early 1980s, the idea that venture capital might be established in India would VENTURE CAPITAL IN INDIA 239 have seemed utopian. India s highly bureaucratized economy, avowed pursuit of socialism, still quite conservative social and business worlds, and a risk-averse nancial system provided little institutional space for the development of venture capital. With the high level of government involvement, it is not surprising that the rst formal venture capital organizations began in the public sector. As in the case of Israel and Taiwan, from its inception Indian venture capital was linked with exogenous actors, public and private. In India, one of the most autarchic economies in the world, both the development of venture capital and the information technology industry have been intimately linked with the international economy. The earliest discussion of venture capital in India came in 1973, when the government appointed a commission to examine strategies for fostering small and medium-sized enterprises (Nasscom, 1998). In 1983, a book entitled Risk Capital for Industry was published under the auspices of the Economic and Scienti c Research Foundation of India (Chitale, 1983). It showed how the Indian nancial systems operation made it di cult to raise risk capital for new ventures and proposed various measures to liberalize and deregulate the nancial market. Oddly, this book never mentioned venture capital per se. Though this and other books did not have immediate results, they were an indication of a political discussion of the di culties India had in encouraging entrepreneurship and the general malfunctioning of the Indian nancial system. Venture capital would become one small part of this larger discussion. (a) The rst period, 1986 95 Indian policy toward venture capital has to be seen in the larger picture of the government s interest in encouraging economic growth. The 1980s were marked by an increasing disillusionment with the trajectory of the economic system and a belief that liberalization was needed. A shift began under the government of Rajiv Gandhi, who had been elected in 1984. He recognized the failure of the old policy of self-reliance and bureaucratic control. The policy toward venture capital should be seen in this larger movement. Prior to 1988, the Indian government had no policy toward venture capital; in fact, there was no formal venture capital. In 1988, the Indian government issued its rst guidelines to legalize venture capital operations (Ministry of Finance, 1988). These regulations really were aimed at allowing state-controlled banks to establish venture capital subsidiaries, though it was also possible for other investors to create a venture capital rm. There was only minimal interest, however, in the private sector in establishing a venture capital rm (Ramesh &amp; Gupta, 1995). The government s awakening to the potential of venture capital occurred in conjunction with the World Bank s interest in encouraging economic liberalization in India. So, in November 1988, the Indian government announced an institutional structure for venture capital (Ministry of Finance, 1988). This structure had received substantial input from the World Bank, which had observed that the focus on lending rather than equity investment had led to institutional nance becoming increasingly inadequate for small and new Indian companies focusing on growth (World Bank, 1989, p. 6). In addition, the [capital] markets have not been receptive to young growth companies needing new capital, making them an unreliable source for growth capital (World Bank, 1989, p. 8). Though the exact sequence of events is di cult to discover, some measure of the impetus for a more serious consideration of venture capital came from the process that led to the 1989 World Bank study quoted above. Noting that the government s focus until then had been on direct involvement in research and development (R&amp;D) activities through its own research institutes, in technology selection on behalf of industry, and promoting of technological self-reliance within Indian industry, the 1989 World Bank report described approvingly a new trend in government thinking toward shifting decision making with respect to technology choice and R&amp;D to industry (both private and public) and a more open attitude toward the import of technology. The World Bank was keen to encourage this shift. The 1989 World Bank (1989, p. 2) report noted that Bank involvement . . . has already had an impact on the plans and strategies of selected research and the standards, institutes and, with support from the IFC, on the institutional structure of venture capital. Making the case for supporting the new venture capital guidelines with investments into Indian venture capital funds (a rst for the World Bank), the World Bank calculated that demand over the next 23 years would be around $67 133 million per annum, and it 240 WORLD DEVELOPMENT period of experimentation and thereby demonstrate the viability of venture capital in India. For this reason, during the initial phase, the Guidelines focus on promoting venture capital under the leadership of well-established nancial institutions (World Bank, 1989). proposed providing a total of $45 million to be divided among four public sector nancial institutions for the purpose of permitting them to establish venture capital operations under the November 1988 guidelines issued by the Government of India. (One of these operations, TDICI, slightly predated the guidelines, having been established in August 1988.) The venture capital guidelines o ered some liberalization, but not everything the World Bank wished. The most important feature of the 1988 rules was that venture capital funds received the bene t of a relatively low capital gains tax rate (but no pass through), i.e., a rate equivalent to the individual tax rate, which was lower than the corporate tax rate. They were also allowed to exit investments at prices not subject to the control of the Ministry of Finance s Controller of Capital Issues (which otherwise did not permit exit at a premium over par). A funds promoters had to be banks, large nancial institutions, or private investors. Private investors could own no more than 20% of the fund management companies (although a public listing could be used to raise the needed funds). The funds were restricted to investing in small amounts per rm (less than 100 million rupees); the recipient rms had to be involved in technology that was new, relatively untried, very closely held or being taken from pilot to commercial stage, or which incorporated some signi cant improvement over the existing ones in India. The government also speci ed that the recipient rm s founders should be relatively new, professionally or technically quali ed, and with inadequate resources or backing to nance the project. There were also other bureaucratic fetters including a list of approved investment areas. Two government-sponsored development banks, ICICI and IDBI, were required to vet every portfolio rm s application to a venture capital rm to ensure that it ful lled the right purposes. In addition, the Controller of Capital Issues of the Ministry of Finance had to approve every line of business in which a venture capital rm wished to invest. In other words, the <a href="/keyword/venture-capitalists/" >venture capitalists</a> were to be kept on a very short leash. Despite these constraints, the World Bank supported the venture capital project, noting that the Guidelines re ect a cautious approach designed to maximize the likelihood of venture capital nancing for technology-innovation ventures during the initial Interestingly, the US experience had shown that such highly constrained and bureaucratically controlled venture capital operations were the least likely to succeed. Nonetheless, four state-owned nancial institutions established venture capital subsidiaries under these restrictive guidelines and received a total of $45 million from the World Bank. The World Bank sought to ensure a level of professionalism in the four new venture capital funds, two of which were established by two well-managed state-level nancial organizations (Andhra Pradesh and Gujarat), one by a large nationalized bank (Canara Bank) and one by a development nance organization (ICICI). The World Bank loaned the money to the Indian government that would then on-lend it at commercial rates to these institutions for 16 years, including a seven-year moratorium on interest and principal repayments. The venture capital funds were expected to be investing principally in equity or quasi-equity. Some monies were allocated for training personnel through internships in overseas venture capital funds. This was an innovative project for the World Bank. The report noted (1989, p. 42) that this is the rst experience of this kind in India or in the Bank. To protect itself, the Bank agreed to engage in substantial supervision. The World Bank would (1989, p. 31) Review and approve initially nearly 30 investments (made by the <a href="/keyword/venture-capitalists/" >venture capitalists</a> ) . . . In addition, during supervision missions, Bank sta would review and comment on additional projects before they are approved. Each venture capitalist would have to submit semiannual progress reports detailing (1989, p. 33) prospects, risks, and reasons for choice of nancial instruments. The <a href="/keyword/venture-capitalists/" >venture capitalists</a> agreed with the World Bank that they would operate under the following operating guidelines: (i) the primary target groups for investment would be private industrial rms with above-average valueadded in sectors where India had a comparative advantage, and protected industries would be avoided; (ii) the quality and experience of the management was key, along with the prospects of the product; (iii) the portfolio return should VENTURE CAPITAL IN INDIA 241 be targeted to be at least a 20% annual return; and (iv) no single rm should receive more than 10% of the funds, and the venture capitalist would not own more than 49% of an investee s voting stock. It is interesting that, in its spirit, the World Bank s funding was similar to that of the US SBA s funding of SBICs. 10 In 1988, the rst organization to actually identify itself as a venture capital operation, Technology Development and Information Company of India (TDICI), was established in Bangalore as a subsidiary of the Industrial Credit and Investment Corporation of India (ICICI), India s second-largest development nancial institution (at the time, it was stateowned and managed). ICICI had already had an interest in venture capital investing beginning in 1984. It started a small investing division at its Mumbai headquarters in 1985 that focused upon unlisted, early-stage companies (Nadkarni, 2000; Pandey, 1998). This division was run by Kiran Nadkarni, who later went on to become TDICI s second president. As an ICICI division, its venture capital activities were circumscribed by the laws of the time. Learning from responses to similar restrictions in Korea, Nadkarni (2000) and ICICI Chairman Vaghul implemented a novel instrument for India, termed the conditional loan. It carried no interest but entitled the lender to receive a royalty on sales (ICICI charged between 2% and 10% as a royalty). They would typically invest three million rupees in a rm, of which one-third was used to buy equity at par for about 20 40% of the rm, while the rest was invested as a conditional loan, which was repayable in the form of a royalty (on sales) after the venture generated sales. There was no interest on the loan. The problem with this scheme is, however, that the loan did not provide capital gains (Pandey, 1998, p. 256). In its rst year, the division invested in seven such deals, of which three were in software services, one in e uent engineering, and one in food products (a bubble gum manufacturer). Being a novel institution in the Indian context, as Nadkarni (2000) noted, There was no obvious demand for this kind of funding and a lot of work went into creating such deals. Our objective was not purely monetary but to support entrepreneurship. We searched for deals that would earn us 2 3 percent above ICICI s lending rate. It was initially just a one-person operation [he] and then another person joined the team. Looking back, there was no sectoral focus, so it is remarkable that we invested so much in Information Technologies. Their primary task was identifying good rms and making sure they were properly nanced, although they provided little input on setting corporate strategy, business development ideas, or recruitment, as this was not part of the mandate. There were some missed opportunities. For example, Infosys, then a edgling startup, had approached ICICI, which was managing PACT, a USAID project meant to fund cooperative research between US and Indian rms, for funding a medical diagnosis project. ICICI rejected Infosys request, however, because the project was deemed too risky. In 1988, the ICICI division was merged into the newly formed TDICI in Bangalore, which was an equal joint venture between ICICI and the state-run mutual fund UTI. The primary reason for creating the joint venture with UTI was to use the tax pass-through, an advantage that was not available to any corporate rm at that time other than UTI (which had received this advantage through a special act of parliament). Hence, while the investment manager for the new funds was TDICI, it was a 50/50 partnership between ICICI and UTI, and the funds were registered as UTI s Venture Capital Unit Schemes (Vecaus). Vecaus I, established in 1988, had a paid-in capital of 300 million rupees. Founded in 1991, Vecaus II had a paid-in capital of one billion rupees. Another reason for forming TDICI was that venture capital investments were too small relative to ICICI s own portfolio to be worth managing. Furthermore, ICICI did not have the exibility or the ability to evaluate venture investments. For example, Praj Industries, an e uent engineering rm funded by ICICI, used its funding to build a demonstration unit for a sugar cooperative, but delays in project implementation led to a delay in the unit s completion until after the sugar harvesting season ended. This meant that the demonstration unit would not receive raw material for another year. The ICICI investment committee was not equipped to deal with such delays (Nadkarni, 2000). Rather than remaining in Bombay where ICICI was headquartered, TDICI decided to open its operations in Bangalore. The reason for this was that by 1988, when TDICI was prepared to begin serious investing, interest in technology had increased due to the success of multinationals such as Texas Instruments and Hewlett Packard that were operating in Bangalore. In addition, the TDICI managers 242 WORLD DEVELOPMENT wanted to escape from ICICI s large- rm culture. TDICI had to decide between Pune and Bangalore, both of which were emerging as technology centers. Bangalore was chosen because the Indian software rms such as Wipro, PSI Data, and Infosys were based in Bangalore (Nadkarni, 2000). Moreover, Bangalore was the bene ciary of an earlier decision by the Indian government to establish it as the national center for high technology. The research activities of state-owned rms such as Indian Telephone Industries, Hindustan Aeronautics Limited, the Indian Space Research Organization (ISRO), and the Defence Research Development Organization, along with the Indian Institute of Science (India s best research university), were centralized there. TDICI s Vecaus I assumed ICICI s venture capital investments, then valued at 25 million rupees, and invested in several successful information technology rms in Bangalore. Its rst president, P. Sudarshan, was an 18-year veteran of the Indian Scienti c Research Organization and had headed their technology transfer division. Nadkarni was the head of the venture capital division and became the TDICI president in March 1990. Several established software rms received funds from TDICI, including Wipro for developing a ruggedized computer for army use. There were several successes, including several rms which went public, such as VXL, Mastek Software Systems, Microland, and Sun Pharmaceuticals. From its inception through 1994, the fund had an in ation-adjusted internal rate of return of 28% (before the compensation of the <a href="/keyword/venture-capitalists/" >venture capitalists</a> ), despite several mistakes, such as initially funding several import-substitution products that were negatively impacted by a lowering of import tari s. 11 Moreover, initially TDICI saw itself as an organization funding technology ventures, and did not focus as directly upon commercial objectives, thus it made investments in interesting technology and not, perhaps, the best business opportunities (Pandey, 1998, p. 258). There also were organizational failures, the most important of which was allowing relatively junior recruits to lter the deals before the senior professionals saw them. As a result, several high-quality potential investments were missed. Vecaus I made 40 investments to be managed by seven professionals. In retrospect, Nadkarni (2000) believed that was too large a number. By 1994, the dif culties created by the institutional ownership in the management and funding of TDICI be- gan to tell on sta morale, and the performance of the fund s new investments was relatively poor. Despite its di culties, TDICI was the most successful of the early government-related venture capital operations. Moreover, TDICI personnel played an important role in the formalization of the Indian venture capital industry. Kiran Nadkarni established the Indian Venture Capital Association, and was the Indian partner for the rst US rm to begin operations in India, Draper International. In addition to Nadkarni, TDICI personnel left to join yet other rms. For example, Vijay Angadi joined ICF Ventures, a fund subscribed to by overseas investors (ICF Ventures, 2001). In addition, a number of TDICI alumni became managers in Indian technology rms. So, the legacy of TDICI includes not only evidence that venture capital could be successful in India, despite all of the constraints, but also a cadre of experienced personnel that would move into the private sector. There were other early funds. For example, in 1990, Gujarat Venture Finance Limited (GVFL) began operations with a 240 million rupee fund with investments from The World Bank, the UK Commonwealth Development Fund, the Gujarat Industrial Investment Corporation, Industrial Development Bank of India, various banks, state corporations, and private rms (GVFL, 2001). GVFL was established with a handicap, namely it was meant to invest in the State of Gujarat. Although its returns were not high, it was su ciently successful so that in 1995, it was able to raise 600 million rupees for a second fund. Then in 1997 it raised a third fund to target the information technology sector. As with the other rms, GVFL s investment targets shifted through time. As Table 3 indicates, in 1995 there were fewer food and agriculture-related rms and a greater emphasis on information technology than in 1990. The 1997 fund invested exclusively in information technology, discovering what US <a href="/keyword/venture-capitalists/" >venture capitalists</a> had learned 40 years earlier, namely that only fast-changing industries in which large returns are possible can justify venture investing (Kenney, n.d.). As with other rms, GVFL was restricted in the nancial instruments it could use and, as a state institution, they were under pressure to invest in the state of Gujarat. Even more than TDICI, GVFL lacked professional expertise. In 1998, the president of GVFL estimated that ultimately he would be able to VENTURE CAPITAL IN INDIA Table 3. Gujarat venture nance limited investments in 1990 and 1995 funds Firm name 1990 Premionics India Saraf Foods Systech Ajay Bio-tech Colortek Agrochem Lactochem Cals TIPCO Madhusudan Ceramics Industrial eld 1990 Filtration membranes Dried fruits Flight data acquisition Systems Bio-fertilizers Liquid colorants for plastics Tocopherol from fatty acid Lactic acid from molasses Wide area networking Thermoplastic Glazed tiles using new ring process Firm name 1995 Computerskill Srinisons Cables Lokesh Machines Mark Walker Opticals Apex Electricals Akshay Software Technologies Minda Instruments Indus Bo a Brakes Nexstor India 20 Microns Radiant Software Source: GVFL (2000). Industrial eld 1995 Specialized printing Auto wiring Machine tools Designer optical frames High rating transformers Turnkey software 243 Auto panel instrument clusters Disk brake pads Hardware system manageability product Micronized minerals IT training &amp; software achieve a only 15% annualized return (Radhakrishnan, 1998) a relatively weak return. The other two venture funds had only modest success. The Andhra Pradesh state government formed a venture fund subsidiary in its AP Industrial Development Corporation (APDIC). As in the case of GVFL, it was mandated to invest in its state. Though located in a relatively strong high-technology region around Hyderabad, APDIC su ered from all the di culties of state-operated venture capital funds and has also had a relatively low return. The nal venture fund was the only bank-operated venture capital fund, which was a subsidiary of nationalized Canara Bank, CanBank Venture Capital Fund (Canbank). Canbank, which was headquartered in Bangalore, also performed only modestly. Despite their rather weak performance, all of these rms raised new funds and were able to continue their operations. This rst stage of the venture capital industry in India was plagued by inexperienced management, mandates to invest in certain states and sectors, and general regulatory problems. The rms overall performance was very modest, and only TDICI could be considered a success. As mentioned earlier, such problems with government-sponsored venture capital are not unique. Venture capital investing is a di cult art. Government interference and limitations almost invariably increase the risks in an already risky enterprise, making failure more likely. Yet, from this disappointing rst stage, there came a realization that there actually were viable investment opportunities in India, and a number of <a href="/keyword/venture-capitalists/" >venture capitalists</a> had received training. (b) The second period, 1995 99 The success of Indian entrepreneurs in <a href="/keyword/silicon-valley/" >silicon valley</a> that began in the 1980s became far more visible in the 1990s. This attracted attention and encouraged the notion in the United States that India might have more possible entrepreneurs. Figure 1 shows that the amount of capital under management in India increased after 1995. 12 Moreover, it also indicates that the source of this increase was the entrance of foreign institutional investors. This included investment arms of foreign banks, but particularly important were venture capital funds raised abroad. Very often, NRIs were important investors. In quantitative terms, it is possible to see a dramatic change in the role of foreign investors. Notice also the comparative decrease in the role of the multilateral development agencies and the Indian government s nancial institutions. The overseas private sector investors became a dominant force in the Indian venture capital industry. In the United States, the venture capital industry is clustered in three speci c regions (in order of descending importance): <a href="/keyword/silicon-valley/" >silicon valley</a> (San Francisco Bay Area), New York, and Boston. It is signi cant to note that <a href="/keyword/silicon-valley/" >silicon valley</a> and Boston are what can be termed technology-related venture capital clusters, while New York is a nance-related cluster (Florida &amp; Kenney, 1988a,b). In India, as Table 4 indicates, there is also a clustering under way. During the last two years, Bombay and New Delhi dramatically increased their share of the venture capital o ces. This resembles the US historical record, in some ways, 244 WORLD DEVELOPMENT Figure 1. Capital under management by the Indian venture capital industry by year in US$ millions. (Source: IVCA, various years.) Table 4. The location of the headquarters of the members of Indian venture capital association 1993 98 and 2000 1993 Bombay New Delhi Hyderabad Bangalore Ahmedabad Calcutta Pune Chennai Lucknow Total 4 3 1 2 1 1 0 0 0 11 1994 3 3 1 2 1 1 0 0 0 11 1995 2 3 1 2 1 1 1 0 0 11 1996 4 2 1 5 1 1 1 1 1 17 1997 5 3 1 5 1 1 1 1 1 19 1998 6 4 1 5 1 1 1 1 1 21 2000 31 10 1 8 1 2 2 1 0 56 Where there was more than one o ce, only the headquarters was counted. Source: Indian Venture Capital Association (various years), Nasscom (1998). and also indicates the immaturity of the Indian industry. As late as 1998, Bombay, Bangalore, and New Delhi were comparable in terms of number of o ces. In contrast to the United States, where <a href="/keyword/silicon-valley/" >silicon valley</a> asserted its dominance as a technology center at the end of the 1970s, Bangalore had a smaller share of o ces when compared with both Bombay and New Delhi, which are the nancial and political capitals of India, respectively. This apparent weakness of Bangalore is overstated, because the largest venture capital rms in capital terms, i.e., TDICI, Draper, Walden-Nikko, JumpStartup and e4e, are headquartered in Bangalore. In fact, WaldenNikko closed its Bombay o ces in 2000 to consolidate in Bangalore. The other important private venture capital rms, such as Chrysalis (Bombay) and In nity (Delhi), have operations in Bangalore as well. In some cases, there were distinct reasons for locating in Bombay. For example, Chrysalis chose Bombay because of the internet boom, which was centered in Bombay. As in the case of New York s Silicon VENTURE CAPITAL IN INDIA 245 Alley, this proved to be a strategic error. In the case of In nity, it chose Delhi because the partners lived there. Another large venture capital rm, SIDBI, is based in Bombay because it is a subsidiary of a state-owned bank with corporate headquarters in Lucknow and Bombay. Moreover, Citibank s venture capital operation is headquartered in Bombay; but its portfolio consists mainly of private equity salvage-type deals and some internet startups. Many of the other Bombay venture capital operations are involved in what, in the United States would be considered nancial engineering. Thus Bombay has a venture capital community that resembles the New York nancial cluster. The involvement of the overseas private sector in the Indian venture capital industry was a path-dependent experimental process. For example, in 1993, in an attempt to develop investment opportunities businesses in India, Vinod Khosla (2000) spent three years commuting between India and <a href="/keyword/silicon-valley/" >silicon valley</a> . In 1996, he gave up, returning full-time to <a href="/keyword/silicon-valley/" >silicon valley</a> , as he concluded that the Indian environment for venture capital development was inadequate. Khosla was not alone. In 1993, Bill Draper, who had begun venture investing in 1959, returned to <a href="/keyword/silicon-valley/" >silicon valley</a> from a series of positions dealing with development in the federal government and then the United Nations. In conjunction with a Stanford second-year MBA student, Robin Richards, Draper decided to begin venture investing in a developing country. After reviewing a number of countries, he decided that India s strength in software and English capabilities made it an ideal candidate. So, in 1995, he formed Draper International and recruited investments from a number of successful <a href="/keyword/silicon-valley/" >silicon valley</a> entrepreneurs and investors of Indian descent (Draper, 2000). This would the rst major overseas rm to begin investing in India. To head their Indian o ce, they attracted Kiran Nadkarni from TDICI. Though Draper International would be the rst, others soon followed. In late 1996, the Walden Group s Walden International Investment Group (WIIG) initiated its India-focused venture capital operation. The rst fund, Walden-Nikko India Venture Co., was a joint venture between WIIG and Nikko Capital of Japan, investing in early and late-stage companies (WIIG.COM, 2000). In the late 1990s, many more venture capital funds established operations in India. The formalization of the Indian venture capital community began in 1993 with the establishment of the Indian Venture Capital Association (IVCA) headquartered in Bangalore. The prime mover for this was TDICI s Nadkarni, who became its rst president. There were nine members, the state-owned or managed ones being TDICI, GVFL, the Industrial Development Bank of India s venture capital division, RCTC, APIDC s venture capital division, and Canbank Ventures. The private members were Credit Capital Corporation, a joint venture with Commonwealth Development Corporation, headed by investment banker Udayan Bose, headquartered in Mumbai, Indus Ventures of Mumbai (started by T. Thomas, an ex-Unilever board member, and the Mahindra Group), Grindlays (later part of the ANZ banking group, subsequently purchased by Standard Chartered Bank), and the British venture rm 3i Corporation. Initially, the IVCA met only quarterly due to their geographical dispersion. Since the majority of the rms were subsidiaries of the Indian government agencies or banks and received funds from international development agencies, there was little real need to interact intensively. IVCA had little success in addressing the main problem of most of its members, the absence of tax pass-through. The reasons for lack of success were twofold: First, the government did not understand the bene ts of venture capital in economic development terms. This was a result, in large measure, of the lack of recognition of the potential of the Indian-owned portion of the software industry. Moreover, the venture capital industry was tiny with respect to the overall Indian economy, having committed funds of three billion rupees and disbursements of less than half that amount. Hence, they were unable to mobilize su cient political pressure to motivate any liberalization. Second, the largest player, TDICI, had few reasons to demand changes in the regulations, because it was una ected by them (and, perhaps, bene ted by being able to attract funding away from its taxed brethren). As late as 1999, the IVCA was still struggling to be an e ective lobbying force. In fact, much of the most powerful lobbying for the Indian venture capital industry comes from the Indian information technology industry association NASSCOM. The IVCA does not even have a web site. The most authoritative web-based 246 WORLD DEVELOPMENT information on the Indian venture capital industry is available from the NASSCOM site. There are internal divisions also. According to one source, in 1999 approximately 80% of the total venture capital investments were derived from overseas rms (Singhvi, 1999). These foreign rms registered in Mauritius as a strategy to avoid the onerous regulations and taxes imposed by the Indian government a mechanism that foreign securities rms seeking to invest in India had pioneered. 13 A Mauritius registry allowed tax pass-through, and since they did not have other issues, such as nding funds, they had little incentive to join IVCA or actively lobby the Indian government. IVCA thus was a vehicle for Indian venture capital funds seeking to obtain a level playing eld with the foreign funds. In fact, in the past, di erences within IVCA surfaced, with the overseas funds arguing for more regulation from the Foreign Investment Promotion Board, which has a liberal record, and less from the Ministry of Finance, which has a contradictory record. The domestic funds favored a single regulator, and they ultimately won this debate when the SEBI was formed. Still, the divisions in the IVCA prevented it from becoming an important player in the regulatory debates during the late 1990s. Only in 1996 did overseas and truly private domestic <a href="/keyword/venture-capitalists/" >venture capitalists</a> begin investing. This increase in investment was accelerated by SEBI s announcement of the rst guidelines for registration and investment by venture capital rms. Though these changes had a salutary e ect, the development of venture capital continued to be inhibited because overall the regulatory regime remained cumbersome. The inhibition is partly expressed in the fact that as of December 1999 nearly 50% of the o shore pool of funds had not yet been invested (IFC, 1999). The IVCA also has di culties; as of December 1998, only 21 venture capital rms were registered with IVCA. Of these rms, eight are in the public sector and seven have received funds from multilateral funding agencies (IVCA, various years). The IVCA, though not yet able to present a united front, has continued to mature. If the Indian venture capital industry continues to expand and a regulatory framework that ends the bene t for o shore registration is enacted, then the IVCA will be able to present uni ed positions. This will enable it to become the voice of the Indian venture capital sector. 8. RECENT GOVERNMENT EFFORTS TO ENCOURAGE VENTURE INVESTING In the late 1990s, the Indian government became aware of the potential bene ts of a healthy venture capital sector. Thus in 1999 a number of new regulations were promulgated. Some of the most signi cant of these related to liberalizing the regulations regarding the ability of various nancial institutions to invest in venture capital. Perhaps the most important of these went into e ect in April 1999 and allowed banks to invest up to 5% of their new funds annually in venture capital. As of 2001, however, they have not made any venture capital investments. This is not surprising since bank managers are rewarded for riskaverse behavior. Lending to a risky, fastgrowing rm could be unwise because the loan principal is at risk while the reward is only interest. 14 In such an environment, even if bankers were good at evaluating edgling rms, itself a dubious proposition, extending loans would be unwise. This meant that since banks control the bulk of discretionary nancial savings in the country, there is little internally generated capital available for venture investing. The bureaucratic obstacles to the free operation of venture capital remained signi cant. There has been a confusing array of new statutes. The main statutes governing venture capital in India included the SEBI s 1996 Venture Capital Regulations, the 1995 Guidelines for Overseas Venture Capital Investments issued by the Department of Economic A airs in the Ministry of Finance, and the Central Board of Direct Taxes (CBDT) 1995 Guidelines for Venture Capital Companies (later modi ed in 1999). In early 2000, domestic <a href="/keyword/venture-capitalists/" >venture capitalists</a> were regulated by three government bodies: the Securities and Exchange Board of India (SEBI), the Ministry of Finance, and the CBDT. For foreign venture capital rms the Foreign Investment Promotion Board (FIPB) was required to approve every investment, and the Reserve Bank of India (RBI) had to approve every exit. The stated aim of the Indian regulatory regime is to be neutral with regard to the risk pro le of investment recipients. In sharp contrast to the United States, however, where a venture capital fund can invest in any industry it wishes, in India only six industries have been approved for investment: software, information technology, pharmaceuticals, biotechnology, VENTURE CAPITAL IN INDIA 247 agriculture and allied industries. Statutory guidelines also limited investments in individual rms based on the rm s and the fund s capital. The result of these various regulations has been a channeling of venture capital investment toward late-stage nancing. Regulations regarding venture capital continued to be cumbersome and sometimes contradictory. Income tax rules provided that venture capital funds may invest only up to 40% of the paid-up capital of a recipient rm, and also not beyond 25% of their own funds. The Government of India guidelines also prescribed similar restrictions. Finally, the SEBI regulations did not have any sectoral investment restrictions except to prohibit investment in nancial services rms. The result of these various restrictions was government micromanagement of investment, complicating the activities of the venture capital rms without either increasing e ectiveness or reducing risk to any appreciable extent. Impediments to the development of venture capital also exist in India s corporate, tax, and currency laws. India s corporate law did not provide for limited partnerships, limited liability partnerships, or limited liability corporations (LP, LLP, and LLC, respectively). Moreover, corporate law allowed equity investors to receive payment only in the form of dividends (i.e., no in-kind or capital distributions are allowed). Disclosure requirements were, however, consistent with best international practice. In the absence of seasoned institutional investors, advanced-country standards of investor protection that would normally be imposed by such investors have not developed. There was not even a self-regulatory group. The general regulatory environment continues to hinder investment. For example, all investors in the venture capital fund had limited liability, and there was no exibility in risksharing arrangements. There were no standard control arrangements, so they had to be determined by negotiation between management and investors in the fund. Moreover, Indian regulations did not recognize limited life funds, so in India it was relatively easy to terminate a trust, but this meant that the entire rm was closed rather than a speci c fund within the rm. Therefore, each fund had to be created as a separate trust or company. This process was administratively and legally time-consuming. Terminating a fund was even more cumbersome, as it requires court approval on a caseby-case basis. India s regulatory framework inhibits practices used in the United States to reward employees of startup rms. Currently, in India founders and employees can participate in employee stock option programs, if the rm is private, de ned as those having less than 50 nonemployee shareholders. India s corporate laws allow for exible risk sharing, control and exit arrangements between nanciers and rm management, provided the rms in which they invest are private. But, for rms with more than 50 nonemployee shareholders, India s corporate law does not provide exibility in using equity to reward employees. This is a signi cant handicap in recruiting and motivating highquality management and engineering talent. The restrictions on venture capital extend beyond the framework of corporate law. For instance, tax restrictions on corporations require that corporations paying dividends must pay a 10% dividend-distribution tax on the aggregate dividend. 15 On the other hand, trusts granting dividends are exempt from dividend tax. For venture capital the optimal environment would be a tax regime that is scally neutral. It is also important to be tax-competitive with other domestic uses of institutional and private equity nance, particularly the domestic mutual funds sector. In 2001, the tax code was still disadvantageous for the international venture capital investor. For example, earnings from an international venture capital investor are taxed even if it is a tax-exempt institution in its country of origin. 16 Another signi cant impediment to developing a vibrant venture capital industry was India s foreign currency regulations. Even in 2001, the Indian rupee was nonconvertible. The lack of convertibility hampered venture capital in ows from o shore because speci c, timeconsuming governmental approvals from multiple agencies were required for each investment and disinvestment. Just as the currency regime inhibited international venture capital rms from investing in India, domestic venture capital rms were not allowed to invest o shore. Synergistic investments in overseas rms that collaborate with domestic rms were next to impossible. The currency regime also frustrated exit strategies. For example, in early 1999 Armedia, an Indian manufacturer of high-technology telecommunications equipment, was in discussion with Broadcom, a US rm, about being acquired, and it also was seeking an investment from an Indian venture capitalist. But, since Indian 248 WORLD DEVELOPMENT venture capital rms cannot own o shore shares, the deal with Broadcom would have had to be changed from a pooling-of-interests transaction to a cash acquisition. Broadcom, therefore, o ered a signi cantly smaller sum to Armedia, because of the Indian venture capital rm s involvement. Fortunately for Armedia, it was able to obtain bridge funding o shore and did not have to use an Indian venture capital rm (Dave, 1999). The Indian legal and regulatory environment continued to inhibit venture capital investors from maximizing their returns. (a) Current political debates An Indian venture capital industry is struggling to emerge and given the general global downturn, the handicaps existing in the Indian environment are threatening. As we have seen, many of the preconditions do exist, but the obstacles are many. Some of these can be addressed directly without a ecting other aspects of the Indian political economy. Others are more deeply rooted in the legal, political, and economic structure and will be much more di cult to overcome without having a signi cant impact on other parts of the economy. A number of these issues were addressed in a report submitted to SEBI in January 2000 from its Committee on Venture Capital. SEBI then recommended that the Ministry of Finance adopt many of its suggestions. In June 2000, the Ministry of Finance adopted a number of the Committee s proposals. For example, it accepted that only SEBI should regulate and register venture capital rms. The only criterion was to be the technical quali cations of their promoters, whether domestic or o shore. Such registration would not impose any capital requirements or legal structure this is very important, because it would allow India to develop a legal structure suitable to its environment, while o ering tax pass-through for all rms registered as venture capital rms with SEBI. This was an important achievement of the Committee s report. The proposed guidelines continued, however, to prohibit nance and real estate investments. Whether this type of micromanagement is good policy seems dubious. In addition, registered venture capital funds must invest 70% of their paid-in capital in unlisted equity or equity-related, fully convertible instruments. 17 Similarly, related-company transactions would be prohibited, and not more than 25% of a fund s capital could be invested in a single rm. In the United States most of these provisions are not law, but are codi ed in the limited partnership contracts and accepted as common sense. Rather than letting the market decide which venture capital rms are operating responsibly, the Indian government continues to specify a variety of conditions. A number of suggestions were not accepted even though they would assist in the growth of venture capital. Many were related to the much larger general issues of corporate governance. For example, there was no change in the regulations regarding restrictions on currency nonconvertibility, providing employees more exible stock-option plans, allowing domestic venture capital rms to hold equity in overseas startups, and regulations allowing greater exibility in voting and dividend rights. Reluctance to adopt these measures is understandable, because they would strike at some of the fundamental issues of corporate governance in India. Thus they were seen as policy decisions that might set in motion a larger chain of events. At the end of 2001, the Indian venture capital environment contained several unresolved issues. One important obstacle was the inability to pass through unrealized gains or losses through to the venture capital fund s investors through a direct distribution of stock or other securities unless the fund is organized as a trust. In the US, these in-kind distributions are the most common method of compensating investors. This method increases the return for socially bene cial tax-exempt organizations such as foundations and pension funds. Private individuals, of course, pay taxes. Allowing legal structures, such as limited partnerships, will enable such pass-through and encourage investment in venture capital funds. There are also obstacles for overseas <a href="/keyword/venture-capitalists/" >venture capitalists</a> a seemingly unwise set of barriers to a country wishing to attract much-needed foreign capital and to create incentives for the continuing growth of the venture capital industry. Currently, foreign <a href="/keyword/venture-capitalists/" >venture capitalists</a> require permission from the RBI/FIPB for each investment and liquidation. The RBI s formula for disinvestment is based on physical asset value and benchmark price-earnings ratios, both of which are often irrelevant. For example, a rm that has primarily intellectual property but makes a loss may still be valuable, but Indian rules do not allow it to be valued properly. Finally, foreign and domestic rms cannot repatriate capital or earnings without VENTURE CAPITAL IN INDIA 249 further regulatory involvement. There were restrictions on the ability of Indian rms to trade their stock for that of an overseas rm, and it was extremely di cult to sell an Indian rm to a foreign rm something that was simple and common for venture capital-funded rms in Israel, Sweden, Canada, or the United Kingdom. Conversely, Indian venture capital funds today nd it di cult to invest abroad, a signi cant handicap when it is considered that venture capital funds in nearly all other countries were not only allowed to invest abroad, but also did so to secure access to more deals and spread risk. The ability of Israeli venture capital funds to invest abroad, especially in rms established by expatriate Israelis, has been crucial to the dramatic growth of the Israeli venture capital industry. In e ect, the lack of convertibility of the Indian rupee is a structural impediment. Although published data on venture capital investment after the moderate liberalizations e ected in 2000 were not yet available, there were indications that investment has increased, suggesting that further liberalization could elicit even greater venture capital activity. Uno cial gures gathered by SEBI and IVCA 18 showed that over $1 billion was invested in 2000, although the ow has substantially declined in the rst two months of 2001 due to the global tightening of venture capital investment. Interestingly, because the overseas investors were not regulated by SEBI, it was di cult to be certain about the amount of investment. Foreign funds, in the absence of sensible regulations for overseas funds (as discussed above), have preferred the more bureaucratic but ultimately workable process o ered by the FIPB. But, since the FIPB did not classify approvals by investment stage, all statistics are necessarily approximations. 9. CONCLUSION In the introduction we argued that the transplantation of an institution could have four Interactions. In the case of Israel, Interaction (a), which was the successful transfer of the institution with little need to change the environment ruled. In the case of Taiwan, we argued that the institution of venture capital had to be changed, in some signi cant ways, i.e., Interaction (c). For India, we have argued that the environment itself has had to be and needs to continue being changed, i.e., Interac- tion (d). Of course, this is a much more profound process than either Israel or Taiwan has had to undergo. Interestingly, in India we believe that venture capital as an institution may actually need to be changed less than in the case of Taiwan, however this conclusion is not yet entirely veri ed. Earlier patterns of growth or failure in venture capital industries in other countries and regions indicate that the evolution of venture capital seems to be either entry into a self-reinforcing spiral, such as occurred in <a href="/keyword/silicon-valley/" >silicon valley</a> , Israel and Taiwan, or growth and stagnation, as occurred in Minnesota in the 1980s or Germany until recently (on Germany, see Woywode, 2001). In other words, should India wish to develop a high-technology industry funded by venture capital, then it will be necessary to continue improving the environment by simplifying and eliminating regulations that do not perform necessary functions such as consumer protection. The venture capital industry is emerging in India as a result of internal and external factors. Should the venture capital industry be su ciently successful to become self-sustaining, the Indian NIS will be altered. During the 1990s the dominance of government institutions and family-owned conglomerates in the Indian innovation system began to shift somewhat so that entrepreneurship began to play a larger role. Moreover, scholars have generally treated India as a largely self-contained NIS, but during the 1990s the NRIs became an increasingly important component of the NIS through their wealth, contacts, and political and economic in uence. This would be interesting because most scholars have treated the NIS as a relatively xed system. The World Bank, with its agenda of decreasing government regulation, funded the creation of the rst venture capital funds. Though these funds experienced limited success, they were the beginnings of a process of legitimitizing venture investing and they were a training ground for <a href="/keyword/venture-capitalists/" >venture capitalists</a> who later established private venture capital funds. It is unlikely that the venture capital industry could have been successful without the development of the software industry and a general liberalization of the economy. Of course, this is not entirely surprising, because an institution as complicated as venture capital could not emerge without a minimally supportive environment. This environment both permitted the 250 WORLD DEVELOPMENT evolution of the venture capital industry and simultaneously allowed it to begin changing that environment and initiating a co-evolutionary dynamic with other institutions. India still remains in a di cult environment for venture capital. Even in 2001 the Indian government remains bureaucratic and highly regulated. To encourage the growth of venture capital will require further action, and it is likely that the government will continue and even accelerate its e orts to encourage venture capital investing. The role of the government cannot be avoided: it must address tax, regulatory/legal and currency exchange policies, since many of these a ect both venture capital rms and the companies that they nance. More mechanisms need to be developed to reduce risk if funds for venture capital must come from publicly held nancial institutions managed by highly risk-averse managers. India is an example of how purposive action in an environment replete with resources can have long-term impacts on the NIS. The ultimate fate of the Indian venture capital industry has not, as of late 2001, been decided. Even as it grew rapidly during the global Internet-fed venture capital boom of 1998 through early 2000, its continued growth is in no way guaranteed. If the earlier period provided the resources for constituting new rms, then the downturn beginning in February 2000 will be a stringent lter that will remove less viable competitors, in terms of venture capital-backed investments and many of the venture capital funds. Both the venture capital rms and the industry as a whole will be tested severely. NOTES 1. On path dependency, see Arthur (1994) and David (1986). For NIS, see Lundvall (1992) and Nelson (1993). 2. For a more general conceptualization, see Kogut (2000). 3. There is now a proliferating literature on the economics of the venture capital industry. See, for example, Gompers and Lerner (1999) for a guide to this body of work. Economists have attempted to apply principal-and-agent theory, information assymetry perspectives, game theory, and a variety of other models. In each of these there seem to be aws. For example, in the venture capital entrepreneur investment decision, there is a question of which party is less informed. Traditionally, the venture capitalist is considered to not know what the entrepreneur knows. But, the entrepreneur also does not know whether the venture capitalist will perform all the assistance functions promised during the negotiation. Thus, the information is asymmetrical on both sides. Similarly, it is usually the rm s manager who is the agent of shareholders, who are assumed to start a rm and then search for a manager to run it. The agency problem arises because managers may take decisions that will reduce rm value because of their personal goals, e.g., taking overly risky investment decisions because management bene ts from the upside of such decisions but not from the downside of failure. In the venture capital entrepreneur relationship, the agency relationship is unclear. It is the entrepreneur who starts a rm and searches for a nancier while retaining a substantial portion of the rm. The nancier (venture capitalist) also brings operating experience to the relationship. Both parties, therefore, have options to undertake operating actions that can a ect rm value. Thus, the entrepreneur cannot simply be assumed to be the venture capitalist s agent with the possibility of undertaking actions that will reduce shareholder value. Rather than enter these theoretical discussions, we focus upon the actual path-dependent development of the venture capital industry. 4. There are, of course, many important venture capital rms headquartered in other regions. Moreover, in the United Kingdom, the very important rm, 3i, which as part of its remit invested in new rms, was formed in the immediate postwar period. More recently, there has been a proliferation of specialist venture capital rms. For example, there are funds that specialize in retail ventures. Some of the largest venture capital funds, such as Oak Investment Partners and New Enterprise Associates, have partners devoted to retail ventures, though their main focus is IT. So, there is signi cant diversity in the venture capital industry (Gupta &amp; Sapienza, 1992). 5. There is signi cant evidence to support Black and Gilson s observation regarding the relatively greater vitality of the venture capital industry in stock marketcentered economies. But, their reasoning that a public market o ers entrepreneurs an opportunity to reacquire control from the <a href="/keyword/venture-capitalists/" >venture capitalists</a> by using an initial public o ering is odd. There are no cases known where the entrepreneurs have reacquired control at the IPO. In fact, the true reason is nearly the converse of Black and Gilson s assertion. At the IPO, the entrepreneur can also VENTURE CAPITAL IN INDIA begin to cash out, thereby surrendering control. After a successful IPO, however, the entrepreneurs, in most cases, have achieved enormous capital gains on their low cost founder s stock. 6. For example, a recent study found that Brazil has one of the highest rates of entrepreneurship in the world, but it also has almost no venture capital to speak of (Reynolds, Hay, Bygrave, Camp, &amp; Autio, 2000). 7. For a more complete discussion, see Kenney (n.d.). 8. According to a 1991 RBI report, the gross pro t (before provisions on bank assets and taxes) had come down to 1% of assets (a healthy norm would be about 1% for pro ts after provisions and taxes). Moreover, approximately 25% of the total loans were bad. 9. Here we use the term relatively skilled, because Indian postgraduate training was not as good as the training in the best US graduate schools. Put di erently, the Indians who were trained in the elite US schools may have been closer to the cutting edge of engineering. 10. Though we could nd no direct evidence of interaction, it is likely that the World Bank personnel were very much aware of the structure of the SBA and its initial problems. 11. Here they ignored the World Bank s suggestion that they avoid protected industries. 12. The last available information from IVCA is 1998. The 1999 gures have not yet been released. 251 13. One of the authors of this paper, Ra q Dossani, helped develop this mechanism when he was an investment banker. 14. It is true that in the United States, banks have never been an important source of venture capital, even through their SBIC subsidiaries. For the most part, a bank s core competencies are in evaluating and making loans. The problem with loans to small startups is that the capital is at high risk, so any interest rate would have to be usurious. Moreover, since the new rm is often losing money in its early days, paying interest and principal would drain money from the rm during the period when it most requires the money for investment. 15. The tax was to be raised to 20% in the tax proposals for the nancial year beginning April 2000. 16. There is a loophole available to foreign investors, which is to register its fund in a country having a special tax treaty with India providing e ective tax exemption, such as Mauritius or the Dutch Antilles. 17. The 70% rule would prevent nancial rms that are investing primarily in listed rms, in which the risk-pro le of investments could be substantially different from startups, from qualifying for tax passthrough. 18. From private conversations in February 2001 with SEBI s Executive Director L.K. Singhvi and IVCA President Saurabh Srivastava. REFERENCES Adler, F. (2000). Personal interview by Martin Kenney, January 7. Arora, A., &amp; Arunachalam, V. S. (2000). The globalization of software: The case of the Indian software industry. Report submitted to the Sloan Foundation. Available: http://www.heinz.cmu.edu/project/india/. Arthur, W. B. (1994). Increasing returns and path dependence in the economy. Ann Arbor: University of Michigan Press. Autler, G. (2000). Global networks in high technology: The <a href="/keyword/silicon-valley/" >silicon valley</a> Israel connection. Master s Thesis, Department of City and Regional Planning, University of California, Berkeley, CA. Baskaran, A. (2000). Duality in national innovation systems: the case of India. Science and Public Policy, 27(5), 367 374. Bean, J. J. (2001). Big government and a rmative action: The scandalous history of the Small Business Administration. Lexington, KY: University of Kentucky. Besher, A. (1988). Taiwan, US rms team up on venture capital fund. San Francisco Chronicle, June 13, C9. Black, B., &amp; Gilson, R. (1998). Venture capital and the structure of capital markets: banks versus stock markets. Journal of Financial Economics, 47, 243 277. Brunner, H. (1995). Closing the technology gap. New Delhi: Sage. Chitale, V. P. (1983). Risk capital for industry. New Delhi: Allied Publishers. Dalal, Y. (2000). Email communication to Martin Kenney, August 14. Dave, T. (1999). Personal communication to Ra q Dossani, June 1. David, P. (1986). Understanding the economics of QWERTY: the necessity of history. In W. Parker (Ed.), Economic history and the modern economist. New York: Basil Blackwell. 252 WORLD DEVELOPMENT ence on Accessing Venture Capital in India, Stanford University. Nadkarni, K. (2000). Personal Interview by Ra q Dossani, September 27. Naqvi, R. (1999). Business support center of STPI: Proposal for an India infotech center in <a href="/keyword/silicon-valley/" >silicon valley</a> , US, Unpublished manuscript. Nasscom, (1998). Enabling a quantum leap in successful Indian venture creation. New Delhi: Nasscom. Nelson, R. (Ed.). (1993). National innovation systems. New York: Oxford University Press. OECD (2000). A new economy? The changing role of innovation and information technology in growth. Paris: OECD. Pandey, I. M. (1998). The process of developing venture capital in India. Technovation, 18(4), 253 261. Pincus, L. (2000). Personal interview by Martin Kenney, February 7. Posner, E. (2000). Is there a revolution in European venture capital? Berkeley Roundtable on the International Economy Conference Paper Number 4, University of California, Berkeley, CA. PriceWaterhouse (Kesselman &amp; Kesselman) (2001). Moneytree survey Q2 2001 results. Radhakrishnan, N. (1998). State funded: India s local governments try their hand at venture capital. Global techventures, December. Ramesh, S., &amp; Gupta, A. (1995). Venture capital and the Indian nancial sector. Delhi: Oxford University Press. Reserve Bank of India (RBI) (1999). Report on currency and nance. New Delhi: RBI. Reynolds, P., Hay, M., Bygrave, W., Camp, S. M., &amp; Autio, E. (2000). Global entrepreneurship monitor. Kansas City, MO: Ewing Marion Kau man Foundation. Saxenian, A. L. (1998). Regional advantage. Cambridge: Harvard University Press. Saxenian, A. L. (1999). <a href="/keyword/silicon-valley/" >silicon valley</a> s new immigrant entrepreneurs. Public Policy Institute of California. Schrader, H. (1997). Changing nancial landscapes in India and Indonesia. New York: St. Martin s Press. Schware, R. (1992). Software industry entry strategies for developing countries: a walking on two legs proposition. World development, 20(2), 143 164. Shih, S. (1996). Me-to Is Not My Style. Taipei: Acer Foundation. Singhvi, L. (1999). Venture capital industry in India: An agenda for growth. Paper presented at Asia/Paci c Research Center Conference on Accessing Venture Capital in India, Stanford University. Statistical Outline of India 1988 99 (1999). Mumbai, India: Tat Services Ltd., Department of Economics and Statistics. Stinchcombe, A. (1965). Social structure and organizations. In J. G. March (Ed.), Handbook of organizations (pp. 142 193). Chicago: Rand McNally &amp; Company. Stultz, W. (2000). Personal interview by Martin Kenney, February 17. Suchman, M. (2000). Dealmakers and counselors: law rms as intermediaries in the development of Silicon Draper, B. (2000). Personal interview by Martin Kenney, January 11. Evans, P. (1992). Indian informatics in the 1980s: the changing character of state involvement. World Development, 20(1), 1 18. Florida, R., &amp; Kenney, M. (1988a). Venture capital nanced innovation and technological change in the US. Research Policy, 17(3), 119 137. Florida, R., &amp; Kenney, M. (1988b). Venture capital, high technology and regional development. Regional Studies, 22(1), 33 48. Florida, R., &amp; Smith, D. (1993). Keeping the government out of venture capital. Issues in Science and Technology, 61 68. Gompers, P. (1994). The rise and fall of venture capital. Business and Economic History, 23(2), 1 26. Gompers, P., &amp; Lerner, J. (1999). The venture capital cycle. Cambridge: MIT Press. Gujarat Venture Finance Limited (GVFL) (2001). Available: http://www.gv .com/gvcf90f.htm. Gupta, A., &amp; Sapienza, H. (1992). Determinants of venture capital rms preferences regarding the industry diversity and geographic scope of their investments. Journal of Business Venturing, 7, 347 362. ICF Ventures (2001). Investment team. http://www.icfventures.com/ (October 20). Indian Venture Capital Association (various years). Venture activity. Bangalore: IVCA. International Finance Corporation. (1999). Private equity fund portfolio in India. Washington, DC: IFC. Kaufman, S. (1986). H&amp;Q s open door policy into far east venture capital. San Jose Mercury News, November 17. Kenney, M. (n.d.). Agents of innovation, Unpublished manuscript. Kenney, M., &amp; von Burg, U. (1999). Technology and path dependence: the divergence between <a href="/keyword/silicon-valley/" >silicon valley</a> and Route 128. Industrial and corporate change, 8(1), 67 103. Kenney, M., Han, K., Tanaka, S. (2001). Scattered geese: venture capital in East Asian nations. Report to the World Bank, forthcoming. Khosla, V. (2000). Personal Interview by Ra q Dossani, November 10. Kogut, B. (2000). The transatlantic exchange of ideas and practices: National institutions and di usion. Paris: Institut Francais des Relations Internationales. Kogut, B., &amp; Parkinson, D. (1998). Adoption of the multidivisional structure: analyzing history for the start. Industrial and Corporate Change, 7, 249 273. Lateef, A. (1997). Linking up with the global economy: A case study of the Bangalore software industry. New Industrial Organization Programme, International Labour Organization DP/96/1997. Lundvall, B. (Ed.). (1992). National systems of innovation. London: Pinter Publishers. Ministry of Finance (1988). Venture Capital Guidelines. Press Release No. S.11(86)-CCI(11)/87 Department of Economic A airs, O ce of the Controller of Capital Issues, November 25. Mitta, S. (1999). India s technological potential. Paper presented at Asia/Paci c Research Center Confer- VENTURE CAPITAL IN INDIA Valley. In M. Kenney (Ed.), Understanding <a href="/keyword/silicon-valley/" >silicon valley</a> : the anatomy of an innovative region (pp. 71 97). Stanford: Stanford University Press. Sussner, H. (2001). Personal interview by Martin Kenney, March 30. Unit Trust of India (1985). Annual Report. Unit Trust of India (1991). Annual Report. Walden International Investment Group (WIIG) (2000, July 5). Available: http://www.wiig.com/locations/ india/index.html. World Bank (1989). India industrial technology development project sta appraisal report. Washington, DC: World Bank. Woywode, M. W. (2001). The emergence and dynamics of venture capital in Germany: An organizational eld based approach. Unpublished manuscript, University of Karlsruhe. 253 FURTHER READING Bahrami, H., &amp; Evans, S. (2000). Flexible re-cycling and high-technology entrepreneurship. In M. Kenney (Ed.), Understanding <a href="/keyword/silicon-valley/" >silicon valley</a> : The anatomy of an innovative region (pp. 165 189). Stanford: Stanford University Press. Chandrasekhar, K., Dossani, R., Singhvi, L. (1999). The securities and exchange board of India s report on venture capital. Coopey, R., &amp; Clarke, D. (1995). 3i: Fifty years investing in industry. Oxford: Oxford University Press. Dossani, R. (1999). Accessing venture capital in India (working paper). Asia/Paci c Research Center, Stanford University. Jeng, L. A., &amp; Wells, P. C. (2000). The determinants of venture capital funding: evidence cross countries. Journal of Corporate Finance, 6, 241 289. Kenney, M., &amp; von Burg, U. (2000). Institutions and economies: creating <a href="/keyword/silicon-valley/" >silicon valley</a> . In M. Kenney (Ed.), Understanding <a href="/keyword/silicon-valley/" >silicon valley</a> : anatomy of an entrepreneurial region (pp. 219 240). Stanford: Stanford University Press. Lee, C. M., Miller, W., Hancock, M., &amp; Rowen, (2000). The <a href="/keyword/silicon-valley/" >silicon valley</a> habitat. In Chong-Moon et al. (Eds.), The <a href="/keyword/silicon-valley/" >silicon valley</a> edge (pp. 1 15). National Venture Capital Association (NVCA) (2000). 2000 national venture capital association. Arlington, VA: NVCA. National Venture Capital Association (NVCA) (1999). 1999 national venture capital association. Arlington, VA: NVCA. Wang, L. (1995). Taiwan s venture capital: policies and impacts. Journal of Industry Studies, 2(1), 83 94.