The ATP Objective: Addressing the Financing Gap for Enabling Technologies
Mr. Trimble introduced the first panel by reminding participants that the ATP was originally set up to address the “funding gap” for emerging technologies (that is, the tendency of the private sector to underinvest in research and development). Three panelists, from very different backgrounds and perspectives, would present their views on the program. The panel includes
Elizabeth Downing, the proprietor of a start-up company that has received government support from several programs, including the ATP;
David Morgenthaler, a successful venture capitalist, who has not been associated with the program; and
Kathleen Kingscott of IBM, who will outline why a large company values partnerships programs such as the ATP.
THE VIEW FROM INDUSTRY: A START-UP'S PERSPECTIVE
3D Technology Laboratories
Elizabeth Downing—President and CEO of 3D Technology Laboratories, a high-technology start-up company—said that her firm had received several SBIR
awards and was actively cooperating with the NASA Ames Laboratory. It also had a single-company ATP grant about one and one-half years earlier, and was about halfway through the grant.
Start-ups like hers, she explained, have a hard time finding funds to develop their technologies, because of the high risks. This is true even of so-called enabling technologies, which have the potential for widespread application upon further development, but which carry high technical and financial risks. As a result, private investors generally are not willing to support them adequately.
Enabling technologies are technologies that have significant advantage because they
show the potential to make radical improvements in some aspect of society;
tend to be disruptive, by displacing entrenched industries;
have the potential to create whole new industries; but
are too immature and risky to interest private investors because of the long time horizon (even successful technologies may require years or decades to become commercially viable and generate profits).
Why should government fund the development of enabling technologies? Because they have the potential for enormous benefits to society as a whole in terms of wealth and other aspects of the quality of life. Yet private investors will not adequately support their development, because profts are too uncertain or too distant or both. The uncertainties may arise from poor scientific understanding, the need for major improvements in materials, or severe manufacturing challenges.
Traditional sources of equity financing, Dr. Downing said, are not always sufficient at this “seed” stage of development. Venture capital, for example, is an excellent source of funding for companies whose technologies are well enough established to be predictable and scalable in the near term. But for companies built on technology that is more speculative, this source of financing has dangers. Venture capitalists, if they are denied a predictable return on their investments after a few years, will often sell the company for the value of the patents and move on to other investments. Many venture capitalists have neither the time nor the background to be ideal partners.
So-called angel capital (from family, friends, or wealthy investors) may have a longer time horizon. But it is extremely scarce.
Corporate investments in technology-based start-ups are often attractive, when there is a genuine mutuality of interest. But large companies may have many motives for investing. They may, for example, take control of the entire company for the sake of only one part of its technology portfolio, and abandon the rest.
Government funding, through programs such as the ATP, helps mitigate these problems for small companies. In particular, it helps the founders of small firms keep control of their technologies and pursue their development.
A Potential Enabling Technology: Crossed-Beam Volumetric Displays
Dr. Downing's company, 3D Technology Laboratories (3DTL), was formed to exploit an invention called crossed-beam volumetric displays (CBDs). The CBD concept produces a true three-dimensional display in a volume of a special active material, by scanning the material's interior with two independently steered laser beams of different wavelenths. The transparent material, when illuminated by two of the beams at once, emits a photon. By controlling the scanning, the material can be made to display a three-dimensional video image.
The technology, if it can be successfully developed, would certainly be an enabling technology. It would offer many advantages over conventional displays, and might be expected to displace two-dimensional displays in applications throughout the economy.
Reducing Technical and Market Risks
Reducing Technical Risks
At the same time, it presents high risks, which would deter any private investor. The technical risks include the following:
uncertainties about whether it can be scaled up from its current small size;
questions about the properties of the material; and
questions about the availability of software to support this novel display.
But all three aspects of technology risks are being reduced through federal funding. 3DTL has used nearly the full range of federal seed funding that is available to a technology company—Small Business Innovation Research grants (phases I and II) from several agencies; an ATP grant; a Defense Advanced Research Projects Agency (DARPA) Broad Area Announcement; and Cooperative Research and Development Agreements (CRADAs), along with federal contracts, to carry out work on scaling, system architecture, and software development. The goal of this work is to develop a working prototype. If successful, this will substantially reduce the technical risks as perceived by private investors.
Reducing Market Risks
Market uncertainties are also great. At this point, it is impossible to estimate the size of the potential market. Early systems will be expensive, so it is important to identify “early adopter” markets, in which cost is less of an object. In addition, significant markets for the software may be available; the company is selling “development time” to customers in the software research community that want to write software.
The company's business goal is to “develop and manufacture versatile engines,” with software that can be configured into specific products. It wants to sell directly to various markets, including medical, defense, scientific visualization, and others.
Advantages of ATP for Small Companies
Downing concluded by saying that federal funding programs have made it possible to reduce the technical risks of the technology sharply. Recent technical breakthroughs have allowed the company to move to “limited manufacturing.” The company is now at the point at which it is able to look for private sources of capital, including venture capitalists and corporate partnerships.
Of all the federal funding programs that the young company has participated in, she said, the ATP was by far the most useful, because of its large funding commitment, long time duration, and administrative simplicity.
Mr. Trimble asked Downing how many employees she had. “Six,” she answered, of whom all but one are technologists. Mr. Trimble observed that this suggests the ATP is a very efficient way to deliver money to technologists, involving virtually no overhead.
John Newman of the Council on Competitiveness, asked for the “decision tree” leading her to apply to the ATP program. Downing answered that she had started the company as a graduate student and had begun with SBIR grants which are significantly smaller than an ATP award. An ATP employee at a meeting had suggested that she apply for an ATP grant. She was selected to give an oral defense of the project and ultimately won the award. She added that “It's absolutely the best way to go for a small company with high-risk technology.”
THE VENTURE CAPITAL PERSPECTIVE
Morgenthaler Venture Capital
David Morgenthaler, founder of Morgenthaler Venture Capital, said that venture capitalists are probably not as bad as Downing had implied. It is important to understand the nature of the business. Institutional venture capitalists must be accountable to their limited partners. Pension funds, endowments, and others give them money to manage, and they demand in return a positive and fairly predictable internal rate of return. They do not like to wait 10 years (the normal lifespan of our funds) for a return on their money. They will not invest in a company that will pay off in 15 years.
Morgenthaler Ventures is 32 years old, and has a national scope, with offices in Cleveland, Silicon Valley, and Atlanta. It has raised more than $1 billion and has more than $550 million to invest today. Mr. Morgenthaler remarked that he is often asked by local authorities how they can build a dynamic entrepreneurial high-technology economy in the next two years. The answer is that it cannot be done. Many of the nation's older industrial cities lost their entrepreneurial edge at some point and do not have the critical mass of Silicon Valley.
The government has, of course, had a role in the creation of Silicon Valley. An Ohio lawyer once boasted to him, after a couple of drinks, that he had created Silicon Valley.
This was a bold claim, so he asked the lawyer to elaborate. “As a procurement officer at Wright AFB,” the man said, “when they were setting up the space program, I was in charge of finding sources for the associated communications technology. I discussed the concept with all of Ohio's industries. They knew nothing about the technology, and were quite complacent about their positions in steel, automobiles, plastics, glass, and so on. In short they were not interested. So I took it to those long-haired hippies in California. And that was the birth of Silicon Valley.”
That story, of course, is very much an oversimplification. But those government contracts had a great deal to do with the growth of the electronics industry. And Ohio did not have the entrepreneurial resources to make that leap. This is one of the reasons that my firm decided to go national.
Venture Capital in the United States
Venture capital investments have grown dramatically in the past few years. In 1993 $3.9 billion were raised in the United States. By 1999 the total was $46.6 billion. This does not include buy-out funds, so-called mezzanine funds, funds of funds, or other private equity. Remember that the largest single source of venture capital—dwarfing the share of venture capital firms—is of course the so-called “angel investor,” who are relatives, successful business persons, or private investors willing to finance start-ups.
This rapid growth in venture capital suggests that there is plenty of investment money. The question, Mr. Morgenthaler observed, is how to evaluate its risk profile, and whether it can be a satisfactory source of funding for start-up companies with early-state, or immature, technologies.
How Venture Capitalists Evaluate Potential Investments
To elaborate this point, Mr. Morgenthaler described how Morgenthaler screens its opportunities for investment using a simple seven-step checklist, which includes
the need you are addressing, including the size of the potential market;
the product or service to fill the need;
the plan or program to fill the need;
the people, and their ability to implement the plan;
financing (the ability of the above to attract funds);
the exit strategy, that is, the potential for liquidating the investment within 4 to 7 years, such as by going public or selling out to a bigger company; and
the internal rate of return.
A key question is at what stage in the maturity of a technology should a venture capitalist invest? Maturity is inversely correlated with risk, so we prefer to invest when technical uncertainties are reduced. We like to invest in improving or broadening an existing product line, or applying it in a new application. We are often willing to invest in applying an enabling technology to develop or improve a new product. We are rarely tempted to invest in developing an enabling technology to implement a knowledge principle. We would never invest in the scientific work of discovering a new phenomenon or principle, and we practically never invest in work to prove such a principle.
Sometimes we do invest in fundamental research, but never intentionally. I am on the board of a public company that has invested $140 million in pursuing a Nobel-prize-level science project, which may have a tremendous market at some point, if all of the technical and market uncertainties are reduced. However, this in not a typical investment for us; nor should it be. This has to do with the final item on my list, which is the internal rate of return. In many ways this is the limiting criterion on venture capital investments, because we always have our limited partners looking over our shoulders. If you make someone 10 times his money in three years, that's wonderful. If you do it in five years that's good. If it takes 10 years, it doesn't meet the partners' expectations for internal rate of return, and if it takes 25 years they'll never invest with you again.
The Venture Capital Horse Race
Understanding the appropriate role, and limitations, of the venture capital business is important. In many ways, the venture capital business can be likened to a horse race. In this analogy, the technology in which one has invested is the horse. The management of the company is the jockey. The market is the race. The venture capitalist is the owner and trainer.
The race can have any number of outcomes. The horse may be great, but a poor jockey may fall off, losing the race by default. Or the horse may be inadequate, so that even a great jockey—able to get every ounce of performance out of the horse—will still not win. Or you may have a great horse and a great jockey, but decide to race at the county fair (that is, in the wrong market), winning easily, but capturing a trivial prize.
The ideal situation for a venture capitalist is to have a great horse and a great jockey, and to enter them in the Kentucky Derby (a huge market, in our homely analogy). The stakes here are tremendous, because all of the greatest competitors will be at the starting gate ready to enter in the race. If neither the technology nor the management is world-class, then we have no hope of winning. For example, we were early investors in Apple Computer, which originated the mass-market personal computer. We were worried that, once the market for such things was proven, cheap Asian producers would undercut the market. But we were looking in the wrong direction. Apple management predicted a more direct threat: that IBM, the great 800-pound gorilla of technology, would jump in once the market was proven, and Apple would have to find a niche market to survive. That is what happened.
Generally, when we lose, we can blame it on the jockey. Management failures, according to an internal Morgenthaler survey, account for 60 percent of the failed investments, and technology failures for only 10 percent. To be a successful venture capitalist, one needs both world-class people and world-class ideas.
Creating Better Races with the ATP
Today there is a great deal of venture capital available—by far the most ever. The great opportunities in the Internet and biotechnology have generated great interest among investors. Some of these opportunities are real, and some are illusory. Too many investments have been made in both areas. There are 300 public biotech companies and about 1,000 private ones, all but 50 of which need money. Some of these companies will not succeed. In the Internet your guess is as good as mine, but there, too, some companies will go out of business. Venture capitalists cannot finance all of the companies that need additional finance. As you may know, Internet public financing is starting to experience serious trouble. It is hard to predict, but a prudent investor should foresee consolidation and pain.
Public financing for biotechnology was scarce from 1992 to 1999. We brought a biotech company public in 1991, so we know how hard it was to raise funds. The industry underwent a case of investor mania last year, but that was just a blip on the chart and seems to have collapsed.
For a region to develop a technology-based economy, the ATP program or something like it is vital. You may have fine research universities and other scientific talent, but if you do not have the reserves of management and technical skills to bring those great ideas to market, then high-technology development will go elsewhere.
We need better horses, with better bloodlines, to back. But private venture capitalists cannot be expected to support those early, risky projects. The government can help by financing both more discoveries, through basic research, and more “platform technologies,” which lead to new and better products and services. These technologies will be the horse that will attract the better jockeys, to create bigger races.
Mr. Trimble asked Mr. Morgenthaler what was the average size of a single venture capital investment. “Three to five million dollars for an early-stage company,” he answered, “and we would expect to invest about as much again in the next round.”
Mr. Trimble said he had started a company in 1978 to exploit GPS technology. It took six years to secure venture capital and 12 years in all before going public. The early years were supported by angel investors. His experience with the SBIR program agrees with Elizabeth Downing's: the gap in funding between Phase I and Phase II is disruptive and not conducive to growing a company.5 Programs that provide multiple-year funding directly to small groups of technologists—such as the ATP—are far superior.
LOWERING HURDLE RATES FOR NEW TECHNOLOGIES
International Business Machines Corporation
In addition to her experience at IBM, Ms. Kingscott noted that she co-chairs the Coalition for Technology Partnerships, a group of many companies, large and small, with an interest in partnership programs. Her talk was intended, she said, to reflect that broader perspective.
IBM and the Changing Process of Product Development
To put the scale of IBM's research in perspective, Ms. Kingscott reported that IBM spends about $6 billion annually on research and development and has facilities throughout the world, including New York, Austin, Almaden, Beijing, Tokyo, Delhi, Haifa, and Zurich. About 15 percent of the research is considered basic or exploratory. IBM researchers have been awarded two Nobel prizes (for work leading to the important technologies of the scanning tunneling microscope and high-temperature superconductivity). Research has a long tradition at IBM. These investments are designed to create and maintain technology leadership. IBM is, after all, a technology company.
Of course, most of the work is technology development, and IBM is successful in this effort. IBM awarded 2,658 patents in 1998, making it the biggest recipient of patents in the world, with more than twice as many patents as the nearest competitor.
5 The Department of Defense has successfully addressed this gap with the SBIR Fast Track Initiative to speed awards for companies that can attract outside investment. See National Research Council, The Small Business Innovation Research Program: A Review of the Department of Defense Fast Track Initiative, Charles W. Wessner, editor, Washington, D.C.: National Academy Press, 2000.
Partnering is Key
Despite this success, the IBM leadership has recognized that it is not enough to carry out this research by itself. Partnerships are essential. This is because the nature of research has changed dramatically for industry in the past 20 years and collaboration is increasingly vital. We once sent our researchers off to do their work, and when they were done they would throw it over the wall, in the form of some kind of publication for those in the company interested in applying it.
It is no longer possible to rely on that haphazard process. In the 1980s, IBM began to focus on its lack of success in the marketplace despite its predominance in research. Ralph Gomory, who was then head of IBM research and is now President of the Sloan Foundation, began to evolve a new approach, of “incremental improvement,” driven by product or manufacturing needs. That approach required closer ties between researchers and the technology developers and engineers, working as teams.
In the 1990s we began to connect that approach more directly with the marketplace, by adding leading customers to our teams. We identified 12 or 13 customer sets and worked with the best in each set—as teams—with each team made up of researchers, engineers, managers, and business developers. For each customer set, we tried to develop the best applications possible.
The researchers themselves, surprisingly, liked the new team approach, because they could see the applications and market significance of their work. They were so enthusiastic that we changed our organization to link all of the different product and technical divisions with the research divisions.
Cooperating with Others
Although IBM is a large and successful company, we never forget that we are engaged in global competition. Product cycles are becoming steadily shorter; at IBM they are now measured in “web years,” of 90 days. The pace of innovation is illustrated by the fact that half of IBM's hardware revenues come from new products, that is, products introduced in the past six months. Consequently, there is enormous pressure to bring technology to the marketplace more quickly all the time. It is no longer enough to have the best idea; partnerships are needed to develop new ideas and technologies and to move the product to the marketplace. The marketplace moves so quickly these days that if you don't have access to the best people and best technologies inside and outside your company, you will lose. With the need to draw on outside expertise and acquire new technologies and processes, partnerships are increasingly seen as essential. That is where ATP comes in.
ATP and the Value of Partnerships
Partnerships have value in reducing cost and risk, and providing access to technology. They also make technology transfer—which has been called a “contact sport”—more efficient, by improving communication from person to person.
These partnerships involve three kinds of partners: universities, industry, and government. Universities bring breakthrough ideas and talent; industry brings marketplace experience. Government brings scale and scope, resources, and interesting combinations of technologies. The ATP offers the opportunity to bring all three together.
The program's structure has been examined ad nauseam, for the benefit of those who consider it “corporate welfare.” It is not corporate welfare. It is a means of developing technologies with real and substantial benefits to society. The tinkering that has taken place over the past decade to assuage these accusations risks the success of the entire program. The program's evaluation criteria and technical milestones are clear. Its processes generally work well. It is competitive, merit-based, and peer-reviewed. IBM has participated in nine ATP projects, and we believe the program works.
The program also has unexpected advantages. For example, one underappreciated advantage of the ATP is its impact on the quality of students' training. Thoughtful companies and organizations should look to ATP for the chance to work with students. It is a valuable screening and training tool for industry. Students like the ATP because it exposes them to real-life industrial problems and provides hands-on experience with industrial work.
Synergy Between Large and Small Companies
ATP's most important function is as the enabler of new kinds of partnerships. The business partnerships it supports would be hard to find anywhere else. Both small and large companies find the program useful. Small companies like ATP because it brings them into contact with large companies, with technologies, skills, and management systems that are sometimes the best in the world. It allows them to shift from being simply a supplier to these companies to being a full partner in an ongoing relationship. For large companies, the benefits are access to the niche expertise and unique talent of small companies, which complement the broad expertise of a company like IBM.
The Role of Government
Government has an important role in partnership programs. First, it serves as an intermediary, or link, between the private and public research enterprises, which sometimes have little contact. Government research has a better chance of paying off in the marketplace under these conditions.
Second, ATP offers a neutral ground, on which companies can come together to develop what David Morgenthaler earlier this morning called “platform
technologies,” and others call “enabling technologies.” Those precompetitive technologies serve as de facto industry standards, which can then be carried into the competitive marketplace by various companies.
ATP also makes possible highly productive relationships. IBM, for example, participated in an ATP joint program working on technology for high-definition television. We would never have been able to work with the broadcast television networks and the other members of the partnership without the ATP. The program partnerships stimulate the diffusion of technology more widely. This kind of precompetitive enabling technology development will not get done by private industry operating alone. The result is greater than the sum of its parts.
Challenges to the ATP
As many of you know, the ATP does face a number of challenges. Perversely, one of the main challenges the program faces is the uncertainty generated by the unstable political environment with regard to funding of ATP. This uncertainty has eroded business's confidence in the ability of the program to sustain its commitments. The volatile ups and downs of the ATP budget make companies think twice before they participate. Industry needs stability and predictability, especially for R&D investments.
The political debate has been not only about the budget total, but also about the program's makeup. For example, some have objected to the participation of large companies. The cost-matching rate for large companies has also been debated in Congress.
To perform its mission, the program needs stability. It needs the stability of multi-year budget commitments, to support the kind of long-term view that is implied by the notion of support for enabling technologies.
Policy Issues for ATP
In addition to the uncertain political environment, ATP faces a number of issues:
Declining ATP budgets. This narrows the window for applicants, constraining applications.
How close to the marketplace the program should draw the line in defining technologies as “precompetitive,” and therefore suitable for government funding.
Whether or not to focus programs on particular areas of technology. There is a need to strike a balance between identifying leading technologies on the one hand, and developing an industry consensus on the other.
Whether large companies should be allowed to participate singly, or only as members of joint ventures.
Through its participation in this program, IBM has learned some important lessons about partnerships:
The formation and maintenance of partnerships presents challenging technical, structural, and cultural problems.
The structure of a partnership contract is critical. It must define the anticipated business model, and take account of the natural trends of technology. If there is no longer-term market for a technology, development efforts will stop. A market must be kept in sight.
Partnerships must have mutual advantages and strategic synergy for all partners, including attention to human factors, skills, and cultures. All parties must see a win-win situation in order to produce results.
Partnerships are agreed by senior executives, but to succeed, they must be reinforced throughout the organization.
Clear goals and measurements of progress and outcomes are important. When these conditions are in place, the results can be very positive.
A focus on results is vital. The expected outcome must be precisely specified in terms of who is to do what, when, where, and why.
“Venture capitalists tend to focus on particular areas of technology. What impact does that specialization have on start-up companies in other technical areas?” one audience member asked David Morgenthaler. In response, Mr. Morgenthaler responded that the specialization is necessary to develop expertise. One of his partners—a highly successful venture capital investor with long experience—recently tried to invest in an Internet company, but lost to a competitor that had more experience in that specific field. Promising start-ups look for deep experience in their area when they are comparing competing offers from venture capitalists.
Maryellen Kelley of the ATP Assessment Office noted that some have criticized ATP for making multiple awards to the same companies. She asked Kathleen Kingscott how many of IBM's nine ATP grants involved new partners. Ms. Kingscott replied that the company's first two awards were as a single company, but all of the rest have been with joint ventures with a shifting but overlapping set of partners.
A member of the audience asked David Morgenthaler for his view, as a venture capitalist, of how effectively ATP helps small companies move their technologies to the point at which private investors can take over. “Very much so,” he answered. “It is an excellent program for developing enabling, or platform, technologies, which can have broad applications but are long-term, risky investments. Venture capitalists are not going to fund these opportunities, because they are at too early a stage of maturity. Government can and should. It should do more than it is doing.”