Some Implications of the Regents' Proposed UC Ventures

My thinking about the formation of "UC Ventures" is influenced by the fact that today I am flying from London to Berlin to film some thin-film solar photovoltaic researchers and executives who have been living for years in the "valley of death" between important research results and commercial revenues. The photo is of the May, 2011 inauguration of the flagship building for Soltecture, one of the world's best thin-film PV companies that promised to bring zero-energy capabilities to old and new buildings a few years from then. When I stood in front of the building one year after this photo, it had closed, and the company was gone. 

Thus my questions about UC Ventures start with whether it will actually help avoid the collapse--or non-start-- of socially valuable technologies for lack of patient, long-term, adequate financial support.   Will UC Ventures be a "patient investor" that sides unequivocally with the technology--and with the future public that will use it?  Will it offer something special to late-stage technology by entering when others have left?  Will it help original, early-stage research with long-term commitments? Is it fish or is it fowl, or some other, political species?

Here's a bit more background, since it is missing from the Regent's materials.  Venture capital funds  socially-valuable technologies when they are likely to make a lot of money.  They avoid them, no matter how green or clean, when they aren't.  Around 2007, Silicon Valley VC declared clean tech to be the next mega market that VC would chase as it had semiconductors, internet software, and the like: Tony Seba's Solar Trillions (2009) caught the tone.  But energy technology is slow to develop and very expensive to make, unlike router software and apps, and by the time Solyndra went bankrupt in 2011, the Valley had moved on from clean tech to greener financial pastures.  No hard feelings: their job is not to save the planet but to make trillions for their investors, and if solar won't do that, then goodbye solar.  Literally--most advanced solar PV R&D has moved to Asia, often with technology they bought for pennies on the euro or dollar in Western bankruptcies. You can read Vinod Khosla's clear warning here; another denizen of Sand Hill Road assured me face-to-face that the Valley would not invest in clean tech just to keep the manufacturing here.

Soltecture and other thin-film PV companies demonstrate an important fact about the relationship between business and technology, which is that they regularly diverge.  Today's "shareholder" capitalism is all about short-term maximization of returns on investment.   If commercializing a technology will do that then it shall be commercialized; if not, it won't, and the technology will be abandoned.  Businesses exist to make money. Tech development loses money--until the product is finished and sold.  We often ignore this divergence, in spite of frequent critiques from environmentalists, public health advocates, and many many others, not to mention the many business books that discuss the way that the normal pursuit of returns on investment (ROI) conflicts with the R&D that leads to the returns (e.g. Clayton Christensen's The Innovator's Solution, Geoffrey Moore's Crossing the Chasm, William Lazonick's Sustainable Prosperity in the New Economy).  In spite of my recent criticism of Prof. Christensen for his bias against sustainability, I have always appreciated his brutal honesty about what shareholders demand, which is continuous sales growth that increases share prices by exceeding growth expectations.  That does not describe the life of high-impact, high-risk, high-difficulty technologies of the very advanced kind that create the breakthroughs and environmental solutions that we all want.

For a long time, all Very Serious People as Krugman calls them thought venture capital (VC) was the universal solution to tech commercialization in all fields: where universities left off, and mature companies couldn't move quickly, VC would provide capital to get a company from development to delivery to a large commercial market, offering management, development, financing, and marketing expertise along the way.  But VC enters "downstream," for products that are close to commercialization.  They have short time horizons - 3-5 years at most. They expect large multiples of returns on their initial investments, in part to cover their loses on their many bets that fail, in part because their core business is to get very large returns.  

Even this quick sketch reveals the problem: the road from "bench to bedside," from basic academic research to reliable product, is long and winding.  Though you can cash out on a good app in 3 years max, the Internet underlying it took 30 years of growth in the dark, with no market prospects and thus only government funding as the patient investor.  Even if the development frame is short, the social value of a product may greatly exceeds its market value to any given firm.  So firms will underinvest, producing "market failure," which then requires the fix of government support or rich corporations with decades-long horizons, like Japan seems still to have, and like the US used to have when monopolies like AT&T supported some blue-sky research in places like Bell Labs. I am not saying anything original here: mainstream economists like Kenneth Arrow and Richard Nelson began to analyze market underinvestment in research in the 1950s.

But when we talk about technology transfer (TT) from universities to business, we seem to forget this core lesson that business and its private investors are not there to insure the development of technology.  Writers like Joseph "Creative Destruction" Schumpeter and later apostles of "disruptive innovation" have confused the issue by making it sound like they are:  the right entrepreneurial spirit yoked to a big capital stock would automatically benefit society even as it created new wealth. But the bankruptcy of good--and socially valuable--companies like Soltecture, Q-Cells, or even Solyndra are the counter evidence.  Their death flows directly from the withdrawal of capital by investors who decide they aren't going to get major returns in the near future.  They may still love the technology and the company and believe in its enormous future market and great benefit to humanity, but they need to reinvest their capital right now for a higher return--sorry, no hard feelings.  Analysts usually blame the victim--bad management, overrated technology, unlucky price movements.  (If you think this conventional wisdom explains the Solyndra bankruptcy, we have a forty-page refutation that I'll happily send you.) The reality is that VC, start-up companies, and big shareholder-oriented companies (see Lazonick) aren't set up to support the full research and development process, which is better measured in decades than years.  The reality is that financial metrics do not measure science progress or social promise, period.

Like other research universities, the University of California's special contribution to the knowledge ecosystem is basic research--"upstream," early-stage, blue sky, high-risk, wild and crazy, stupid or brilliant, waste or genius, nobody knows in advance.  Yes, I realize universities do lots of applied research, and I certainly agree that no simple line should be drawn between basic and applied, that "Pasteur's Quadrant" is where applied questions produce basic results, etc.  But American capitalism has a huge bias towards both applied and especially the nearly-commercial because that is where the money can be made.  The system has already flushed basic out of the corporate world, which still gets more than 2/3rd of federal R&D funding to do D, making the university's role in basic R more important than ever.  Bear in mind too that STEM research funding is less than it needs to be, and that universities are struggling mightily to support it by digging deeper into their internal funds. 

Enter UC Ventures. How does it define the problem that its existence could solve?   The problem is said to be not enough VC for faculty start-up companies, which are a common--but risky and fragile--step in the commercialization process.   A Regental Working Group on Technology Transfer concluded, "UC should establish a mechanism to invest in UC start-up companies, either through the establishment of or participation in a venture capital fund or funds."  The regents went with the first option, starting their own fund, staked with $250 million from the endowment they directly control, and managed by a "Team" that will be appointed by the Chief Investment Officer while staying independent of that office and of UC in general. No UC employee will serve on the start-up investment Team.  UC Ventures is thus to "support the University ecosystem by providing capital to UC Startups."

Among various details, two others are of special note. The fund will invest in start-ups meant to commercialize UC inventions, but since UC is defined as an "ecosystem," it  includes former UC employees, UC alumni, UC donors, and I imagine anyone who can be designated a "Friend of UC." In other words, current UC researchers will not necessarily have first crack at the fund, so this is not an investment firm devoted exclusively to advancing UC science and engineering.  

The second detail is ambiguity about fund's main role. If it is more a VC fund under the CIO, then the fund's main job is to increase returns for the UC endowment, which is the CIO's actual job.  Accordingly, UC Ventures ultimately reports to the CIO and not, say, to the University Provost, the Office of Technology Transfer, or the Office of Research.  A venture fund, in that context, must have an overwhelming bias for research with near-term commercial promise in the largest market possible, preferably one with low costs. This means that it will necessarily pass over more or less everything that we could call basic research, but also over most late-stage, applied research as well.  If UC Professor Pharma has a start up for a molecule whose patent may yield a high monopoly price in high-income countries, and UC Professor Publichealth has a novel treatment regime for inexpensive drug-delivery in low-income countries, the VC fund, following normal investor practice, must pick Professor Pharma.  Whenever science and markets are at odds, or society and markets are at odds, the VC firm must favor markets.  

Much high-tech business talk dodges the problem by saying that markets, technology, and society all line up sooner rather than later.   But saying it doesn't make it true. The Regents' document takes the same kind of shot (page 3):

Specifically, UC Ventures will aim to maximize financial returns while leveraging the University's unique research and knowledge base, as well as the wider University community, to gain access to attractive opportunities emerging from the University's ecosystem.  The OCIO will design UC Ventures to maximize alignment of interests, minimize costs, and provide a long-term investment horizon.  The OCIO believes there is a compelling opportunity to generate attractive rates of return by selectively investing in early-, mid-, and late-stage companies arising from the UC ecosystem.

There's a tone here of UC Ventures as an especially patient investor, but UC research is one big input in what must be the CIO's strategy to "maximize financial returns."   For late stage R&D, I don't see what advantage UC Ventures will have in the crowded VC world other than having first look at UC product.  For early-stage research of the kind that is harder and harder to fund,  UC Ventures is irrelevant.  (It is, however, already a major PR coup, having gotten quite a bit of media attention of the kind that accepts the popular myth that investing for maximum returns is the same as helping science's public mission.) 

US science needs countercyclical investment, which puts money where other people won't because the work has value Mr. Market can't see. VC firms are cyclical: they do try to get in first, but not too soon, as they don't make money on inventions whose coolness is visible only to them. They need other investors piling in to pump equity values, consumer interest for future sales, and in general a certain obviousness to the value that is never the case in the early or sometimes even in a late phase.  To take Solyndra as an example, when their private investors pulled out in late summer 2011, the Department of Energy, acting cyclically, pulled its loan guarantees.  

What socially valuable research needs--e.g. research like Solyndra's cylindrical solar cells for the flat-roof vastness of the industrialized worlds--is an investor that says, "OK. you've missed every one of your cost milestones, your manufacturing process doesn't work like you said, and five years in, your tech is still crap.  But it's important crap.  And your job now is to make it uncrap, and finally make this whole thing work.  So you said five years? You lied, to us and to yourselves.  So what--R&D is always like that. If it only takes 10 years total you'll be lucky.  So here's five more years of loans, so you can fix your strategy, get more and different engineers, start talking to your customers, create a social buzz around turning Wal-Mart green, do something about your oversensitive equipment, and get back to work."

DOE should have said that. They didn't.  Would UC Ventures ever say that? Not if it's a VC fund.  So what exactly is the research point?

It would be cleaner, and probably more effective, to use some small part of the endowment to have a regular VC fund and not make a big deal out of it, and then use another part of the endowment, say $250 million, to fund fundamental research at UC across the disciplines--in STEM and non-STEM--particularly where the research is so interesting and strange that no one else will fund it.  The seed funding we most need is not for start-ups, where there's no shortage of VC capital sniffing over late-stage ideas, but for embryonic ideas that no outside sponsor will fund.  That wouldn't do as much for UC's corporate image, but it would do more than UC Ventures for knowledge and innovation.