Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Cultivating Technological Innovation MLLIAM J. PERRY Four significant factors will contribute to continued U.S. Iead- ersh~p in captivating technological innovation: (IJ our accelerating ability to foster technological discontinuities, (2j cooperative uni- versity-irulustry ties, (3J a large pool of risk capital for innovative start-up companies, and (4) strong cultural underpinnings for in- novation. In the first 150 years of the Republic, the United States experienced unprecedented growth and prosperity. In that era, our economy could be viewed as a non-zero-sum game: wealth was being injected into our economy from the new land that was being opened up, the new minerals that were being discovered, and the new people who were coming into the country. Today we live in an era of limits: we are not opening new land, we face a declining mineral production, and we are experiencing relatively small pop- ulation increases. Nevertheless, the country is still experiencing growth and prosperity, and our economy can still be viewed as a non-zero-sum game. The new wealth today, rather than coming from mines and farms, is coming from innovative technology. Indeed, we might say that in the nineteenth century the wealth in California came from the gold in our mountains; today it comes from the silicon in our valleys. Silicon Valley is the wonder and the envy of the world. Countless study groups and panels from such tech- nological superpowers as Great Britain, France, West Germany, Sweden, and Japan are trying to learn what our "secret" is what has caused this unprecedented explosion of innovation. In attempting to explain Me explosion of innovation, I will sew by quoting 443
134 we r lAM J. PERRY G. W. Dummer, the British engineer who almost invented the integrated circuit. At a technical conference in 1952, Dummer said: With the advent of the transistor and the work in semiconductors generally, it seems now possible to envisage electronic equipment in a solid block with no layers of insulating, conducting, rectifying, and amplifying materials, the electrical functions being connected directly by cutting out areas of the venous layers. That was 6 years before the integrated circuit was invented. One does not have to be an electronics engineer to understand that he was anticipating the development of the integrated circuit, and indeed he later spent many years working on that development (supported by He Royal Radar Establishment in England). But it was not G. W. Bummer who invented the integrated circuit, nor was it England that had the opportunity to exploit it. The inte- grated circuit was developed by Kilby at Texas Instruments and Noyce at Fairchild, and Be explosion in technology Hat followed occulted in the United States. Years later, looking back wistfully on this lost opportunity, Dummer said: It is worth remembering that the giant American electronics companies were formed since the war by a relatively few ente~pnsing electronics engineers, setting up with either their own capital or risk capital from a bank. Often a government contract would start them off. Hard work was necessary, and the large home market was a great asset, but the climate of innovation was such that any advanced technical product could be sold. . . The American system of encouraging employees to hold shares in the company is one which should be emulated, as a part share in the fimn's prosperity gives a sense of increased responsibility.... Successful businesses are almost always dependent on a few people who are ~nnovanve and en~us~asiic. That is Durnmer's assessment of why the integrated circuit revolution took place in the United States and not in England. UNDERPINNINGS OF TECHNOLOGICAL GROWTH I would like to generalize on Dummer's points. First, a dramatic change in technology a technological discontinuitycan create an opportunity to form a new industry, and certainly the development of the integrated circuit was such a discontinuity. Since that development, there has been a stream of discontinuities that are best characterized by observing that, for He past two decades, the computer industry has experienced a 20 percent annual improvement in price and performance, largely as a result of technological changes in semiconductors. There is no other industry in which anything even remotely like Hat has happened, and that sustained improvement in price and perfo~ance has led to new products, new companies, and the creation of entire industries that did not exist a decade or two ago. Since technological discontinuities are the key to this growth, one might question
CULTIVATING TECHNOLOGICAL INNOVATION 445 whether we are at the tapering-off point in the introduction of technological discontinuities in the semiconductor and computer fields. I believe not. In fact, I expect that in the next decade we will see a much greater improvement in price and perfo~ance than occurred in either of the past two decades. In particular, I anticipate that in the next decade we will see aggregate price and performance improvements of more than a hundredfold in the field of semiconductors and computers. A hundredfold improvement is not an ex- travagant estimate; in fact, it may very well be conservative. The second critical feature in this creation of new wealth through innovative technology is cultural, and that is primarily the point Dummer was making in the second quotation. In Europe or Japan if the chief engineer of a major company resigned his job in order to start up a new company, his friends and his neighbors would think there was something wrong with him. In the United States, particularly in Silicon Valley, if the chief engineer of a major company did not leave his job to form a new company, his friends and neighbors would think there was something wrong with him. And that sum- marizes the cultural differences between the United States and Europe or Japan. A point that might be added to Dummer's statement is the importance of having great research universities intimately tied to industry. There is no better example of that than the industry-university cooperation talking place at Stanford, which is being emulated at a number of other universities in the country, including MIT and Berkeley. The final point, one that was alluded to briefly by Dummer, is the avail- ability of risk capital. If new technological enterprises are financed, risk capital must be available for they are risky enterpnses. In Europe, if a person wants to start a company, he or she probably gets the money, if it comes at all, from commercial banks. In the United States, banks are an unlikely source of risk capital for an enterprise. In my experience in working with start-up companies, commercial banks play a very limited role indeed, namely the role of providing loans that are secured by equipment, receivables, or the founder's personal guarantees. A second alternative source for financing technological innovation is the public. There was a brisk market in the early l980s in what was called '~penny stock." The penny stock market operated primarily out of Denver, and it provided public financing for high-risk enterprises. That market is now out of favor. Although it may be premature to proclaim its demise, the penny stock market is certainly in very weak condition today. The problem win the public as a source of financing for h~gh-risk, high technology companies is that once a company receives public financing, there exists very great pressure for early financial performance, which is not well suited to the kind of ventures that involve the development of complex, new technological products. Also, while a high-risk venture can have a high reward, that high
446 WIGWAM J. PERRY reward comes with a low probability, and too often the investors forget that a low probability of success means a high probability of failure. Enough of these failures in fact happen that eventually this market gets a bad name. In my view, the public markets are not an appropnate way to finance high-nsk start-up companies; the proper role they play is in the financing of emerging technology companies with products already being marketed. The existence of a vigorous public market for such emerging high technology companies is critical to the existence of a vigorous venture capital market. So, the role of the public market in risk capital is indirect in that it provides a means of liquidity for the venture capitalists. A third way of financing high-nsk, high technology companies is through R&D partnerships, the basic idea of which is sharing the risk the company, the public, and indeed the government all share the risk of failure. Moreover, the partnership is structured so that it operates on a suitable timetable; many partnerships are structured so that the payback on them is not scheduled to occur for 5 or more years in the future. Nevertheless, R&D partnerships have fallen out of favor lately because of several spectacular failures, of which Trilogy and Storage Tech Computer are two prominent examples. One of these companies has gone out of business, and the other has fallen on hard times, each after raising more than $60 million of partnership funding. Such spectacular and highly visible failures have given R&D partnerships a bad name. Despite this, they will probably make a comeback because Hey have fundamental appeal; that is, they are an appropriate vehicle for financing high-risk ventures because of their risk sharing and long time horizon. They will likely come back, however, in a particular way, namely, in the form of partnership-pooled funds that allow the investor to get many of He benefits of an R&D partnership, but with the furler benefit of spreading the risk over many enterprises. Another important way of financing venture operations is through private means, that is, through the founders themselves or their friends and relatives. How much financing of this sort is actually done is not known, but it is used in a significant number of start-up companies. Indeed, ~ started my own company, ESL, in 1964 entirely that way; that company, which today has a quarter of a billion dollars in annual revenues, never at any time in its history received a dollar of venture financing. It was financed entirely with funds from He founders and grew on retained earnings. That mode of fi- nancing puts obvious constraints on the grown rate of the company, but Hose constraints may be a blessing in disguise in that they do not allow the entrepreneur to get into the kind of problems that go with unconstrained grown. (Another company a little better known than ESL Hewlett- Packard financed itself the same way.) Today most founders seem to prefer to get venture capital and only go to private funding if they fail to get venture funding. Private financing has much to recommend it, however, and it should be considered as a first resort, rather than a last resort, for founders.
CULTIVATING TECHNOLOGICAL INNOVATION 447 Of course, the most popular means of financing start-ups today is with venture capital. We have seen in the past few years a dramatic increase in the funds available to venture capital. We have all been bombarded with statistics and other information on this point. To summarize just the principal numbers: during most of the 1970s about $100 million to $200 million a year of new money came into the venture capital community, but in the past few years that figure has risen to between $2 billion and $4 billion a year. Why has this dramatic grown occurred? One very obvious reason is the change in the tax laws that occurred in 197g and again in 1981, which reduced the effective capital gains tax rate to 20 percent. There is no need to repeat the arguments about the critical importance of tax policy in technological innovation, except to note Mat while tax policy plays an important role, it is not the only reason for this dramatic increase in venture funds. Probably the primary reason that we have gone from a hundred million to a few billion dollars a year is the record of performance of venture funds during the 1970s. Harnbrecht & Quist venture funds, which were stated in the early 1970s, showed an average rate of return in excess of 40 percent for Mat decade. A number of other venture firms that were started in Me late 1960s and the early 1970s showed comparable rates of return. Not surprisingly, this caught the attention of the managers of pension funds and other institutions. When they compared the 40 percent growth rates of venture funds wig the relatively low rates they were getting from Weir investments in blue-chip stocks, major fund managers began to invest in venture capital funds. How was this venture-fund perfo~ance achieved? One fundamental factor leading to this performance has been the underlying improvements in tech- nology the 20 percent per year price and performance improvement re- sulting from technological advances allowed the successful companies in this field to sustain very impressive secular growth in earnings year after year. The second fundamental factor has been the related grown in the over-the- counter (OTC) market for high technology companies, which allowed the venture funds to achieve liquidity at attractive multiples. Public market prices depend on cyclical as well as secular factors, as illustrated by the Hambrecht & Quist Technology Index of about 150 stocks (Figure 1), which serves as a surrogate for the price increase of emerging technological companies in the past decade. Two Wings are apparent in looking at Figure 1. The first is that high-tech stocks, compared with the Standard & Poor's 400, are highly volatile. They have been characterized as having a betas of two, actually in He past year or so, it has been much greater Han that. The second is that, notwithstanding this volatility, there is a pronounced secular growth. While the market volatility is exceedingly important, as He past two years have demonstrated all too clearly, it is dominated in the long tenn by the secular *A factor describing the relative volatility of a stock's puce win cyclical market vananons.
448 N D X V A U WILLIAM J. PERRY 1200 _ 1000 400 200 - ~:: 1 1 :~ Or Growth Index Technology Index -a----. S& P400 1> '`.Je o- _ l ~ l a lo. ~' Fl--- 82 83 84 85 ~ . ^. ~ . ~ I' ~ ,` 1 `! 75 76 ~ _ L T 70 7' l 77 78 79 80 8' FIGURE 1 Hambrecht & Quist growth and technology indexes, Standard & Poor's 400 index, 1970-1985. core: For the purpose of companson, all indexes have been set at 100 at December at, 1978. Chart reflects month-end values. Latest values are for June 28, 1985. SOURCE: Hambrecht & Quist, Inc. Repnnted with permission. growth, which is being driven by the sustained improvement in price and perfo~ance resulting from technological improvements. A CLOSER LOOK AT VENTURE FUNDS The public markets are key to understanding the performance of venture funds because the grown in venture funds is driven by the increase in the price of the venture stock when it becomes public. Thus, the capital gains tax rate affects the availability of venture funds in two ways: first, by in- creasing the attraction of venture capital for private investors (they are at- tracted by the potential of retaining more of their earnings in capital gains); and second, by increasing the attraction of high-growth public stocks for investors. This latter effect tends to increase He growth rate of venture funds, which increases the attractiveness of venture investments, even to tax-free institutional funds. There has been a lot of publicity in the past year or two suggesting that
CULTIVATING TECHNOLOGICAL INNOVATION 449 venture capital is drying up. That is a misunderstanding of what in fact is happening. The performance of venture funds in 1984 was lackluster largely because of the lackluster performance of Me public market for high-tech stocks. But this is the cyclical aspect of venture-fund valuation. Venture investments will continue to have appeal because of the secular grown in technology. Even in 1984, which was a dismal year for technology stocks in the market, a number of venture funds were formed at levels of $50 million to $100 million each. So, money continues to flow into venture capital not as much as in 1983, but certainly as much as in any other year in the history of venture capital There will be a cyclical interest in venture capital as a result of the cyclicality of the public market in high technology stocks. But what will override that will be We continuing sustained improvements in technology, which will lead to secular growth in stock earnings, which will in time lead to secular growth of stock pnces, which in turn will bring up the average rate of return of the venture funds. Where do we stand today in venture capital? Can a start-up company get venture money if it wants it? There are literally billions of dollars available in venture fiends, but if you tank to someone Lying to start a company today, he has a different perspective. He sees a very tight market. The reason it is tight does not have to do win the lack of money, but way the fact that venture funds, having made major investments in 1981 and 1982, now are holding companies needing second- and th~rd-round financing, which may not be available from public markets as it was in 1982 and 1983. For that reason the venture funds are tending to reserve their money for these later- stage financings. Thus, while the money is there, it is not being made available to start-up companies to the same extent that it was in 1982 and 1983. LARGE COMPANIES AND TECHNOLOGICAL INNOVATION Thus far, I have talked about innovation through start-up companies. But He amount of money Hat IBM alone spent on research and development last year was about equal to the total amount spent by venture funds on start- ups. So, why do we not look to the large companies for technological in- novation? The research laboratories of large companies are very efficient at developing products that evolve from their predecessor products. However, to the extent that a new product depends on technological discontinuities, He large companies are not nearly as effective. Their size and bureaucratic organization tend to work against the timely development of products that are a dramatic departure from earlier products. But there is another factor Hat may be even more important. Large companies suffer from what I call "the liability of leadership." The following is a case in point. At the time the transistor was being commercialized, Sylvania Electric
450 wIl r lAM J. PERRY Products was one of the three largest manufacturers of vacuum tubes in the world. Sylvania fully understood the potential of the transistor for making the vacuum tube business obsolete, so it established a research center to develop transistor technology. The company brought together the same kind of physicists, the same kind of engineers and technicians, and the same kind of equipment that existed at Shockley Semiconductor and later at Fairchild Industries. They Men took this activity and assigned it to the manager of the vacuum tube division the rest is history. There is a psychological as well as a technological problem in this liability of leadership. A major company that is a leader in a given field has a very difficult time embracing a discontinuous technology that can lead to a product that has the potential of killing off the product that is yielding most of its earnings. IBM had exactly the same problem win the minicomputer as Sylvania had with the transistor. It wasn't that it did not know how to build minicomputers, but it hung back and let a then unknown company- Digital Equipment~efine and create that market rather than develop a minicom- puter that would have cannibalized the low end of its mainframe product line. Later, when IBM decided that the personal computer market was real, it entered that market in a vigorous, entrepreneurial manner because the personal computer was not competing so directly with other of its major products. More and more of Me larger companies are saying: What can we do to participate In these venture activities, recognizing the limitations we have because of our size and our leadership in related markets? Some of Hem have gone into the venture business themselves; Mat is, they have formed and managed venture capital funds. Good information on the results of these ventures is not available, but, In general, they have not worked well. Whether or not Be investments were successful, corporate-managed venture funds have not generally achieved Be primary thing Hey wantedHe transfer of technology and the creation of new products. In He past few years, a number of companies have med a different ap- proach. Known as corporate partnering, this approach is an alternative both to corporate venture funds and to corporate acquisition of smaller companies. An example is the strategic relationship that General Motors (GM) has formed with a small company called Teknowledge. In this relationship, Teknowledge remains an independent company but works with General Motors to develop applications of artificial intelligence to He automobile industry. GM provides funding and market-wise guidance for the R&D projects that are involved. Teknowledge gets a ready market for the product, and GM gets a head start on its competitors with the new products. In addition, GM has made a minority investment in Teknowledge. Thus, corporate partnering has become an alternative way for small companies to raise equity financing. General Motors is perhaps the most experienced practitioner of corporate partnering
CULTIVATING TECHNOLOGICAL INNOVATION 451 today. This in itself is remarkable, since a decade ago the company deservedly had the reputation of believing that all wisdom in the automotive field stemmed from General Motors looking outside for ideas was the last thing that would have occurred to the GM management. General Motors has made dramatic changes in its point of view, and a number of other large corporations are following its lead. This new trend has been noted by William Miller, the president of SRI, International, and a professor at Stanford (where he teaches a course on innovation and technology). He observed that in the past few centuries, the periods of greatest innovation and prosperity occurred when industry was practicing eclecticism. Eclecticism is reaching out for ideas it is the op- posite of the NIH (not-invented-here) complex Hat plagued General Motors during the 1970s. Miller observes that this trend for companies like General Motors to reach out to other companies for ideas is a very favorable indicator for the future. SUMMARY I believe four significant factors will contribute to continued U.S. lead- ersh~p in cultivating innovation across a broad spectrum of technologies. First, ~ anticipate that our ability to foster technological discontinuities will continue and accelerate, resulting in such developments as a hundredfold improvement in He price and performance of computers during die next decade. Second, I expect our research universities to further strengthen their ties to high technology industry, thereby advancing our nation's competitive advantage in world markets. Third, a large pool of risk capital will continue to be available for innovative start-up companies. And, finally, our culture will continue to provide strong underpinnings for innovation. Especially in recent years, the cultural values that promote innovation have spread from entrepreneurial start-up firms to major corporations, which are now reaching out to find ideas for new technology and new products wherever those ideas exist.
