The most recent Pitchbook has some interesting information on private equity exits. If university administrators have been sold on the idea that there’s fast cash in startup company equity, there are also data to dampen administrative spirits.
The Pitchbook identifies three primary forms of exit: acquisition, buyout, and IPO. For investment success, this is the real meat of the matter.
*According to Pitchbook data, the holding period before exit via acquisition is now about 4 years, and 6 to 7 years for IPO or buyout. These periods are similar to estimates for the time it takes from license to the time royalties from sales might be seen.
*More than 90% of the exits for companies raising less than $25m is by acquisition. If folks are raising small amounts of investment, then if there’s going to be an exit, they will be acquired. What’s interesting, then, for university claims to economic development and innovation, is what happens to the company, and its products, when it is acquired. The company may be shut down or moved, or its products pulled in favor of those of the acquirer. In times of fever growth, I’ve seen big companies buy small ones just to expedite the acquisition of engineers–get them 20 or 30 at a time rather than having to hire them one by one. One is essentially buying the work product of someone else’s HR department as the primary asset of value.
*Nearly half the exits are software-related, involving 80% of exit capital. Leading areas of activity are enterprise software, commercial services, and media–not particularly the high-tech university thing, just as health services show lots of growth but not a lot of high tech. It is strange, then, how fixated university licensing offices are on biotech, as are state economic development offices. At one point a few years ago, 47 states had identified biotech as the industry that was going to save their economy. Holy cow. But here it is that enterprise software, media, and the like are doing well for exists–stuff that four or five or seven years ago universities might have been focusing on, but mostly were not. The windows of opportunity for investment and exit shift around, across hundreds of areas of endeavor. Some windows are only months open, while others, like biotech in the 1980s, might run for a decade. An internet speculative investment window ran six years, from about 1996 to 2004. Most university tech transfer offices ignored it, didn’t hire folks who understood it, and stuck to their biotech and pharma activities, hitting closed panes of glass like clueless birds. Arguably, there was a nanotech investment window, too, but also arguably university technology licensing practices adapted from biotech destroyed the window before it could stay open very long. When all the rotten university patents on carbon nanotubes expire, perhaps there will be another investment window, this time without having to deal with universities. If it happens, it will be a much more lively time.
*The median venture capital raised prior to exit is $76.5m for IPOs, $11m for acquisitions, and $19.7m for buyouts. If companies are not raising on the order of $10m over four years, then they are not positioned for a median exit. There is a concept out there of the “therapeutic dose.” If you don’t take enough of that aspirin, it doesn’t actually help your pain. There’s something similar with regard to investment in business activities. If you don’t spend enough, you don’t get the desired result. You show all the behaviors of doing business, but you are not working at a scale sufficient to make success. Steven Blank argues (see pp. 126-27), for instance, that if you are going to challenge an incumbent with greater than 41% market share in a given industry, you have to outspend them, on average, by a factor of 3 in your marketing budget to gain any market share. It matters a great deal, then whether a company has raised a therapeutic dose of investment for the structure of the market in which it intends to compete. If those fresh happy university startup companies are going up against entrenched market leaders, and are raising their happy $1m in capital, then they are likely just eating investor money, and the “jobs” being created will disappear again in a matter of months. No one reports, however, the number of jobs lost at university startup companies. There’s a reason for this.
*Biotech represents only about 3% of exits by acquisition, 2% by buyout. 21% of exits by IPO are biotech. But companies in biotech are raising on the order of $75m prior to IPO. So if a university is hyping investment in five biotech companies totaling $100m and you are impressed, keep in mind that that level of investment is not nearly enough to get the company to a typical IPO, and it is highly unlikely the company will be acquired (3%!) (unless, of course, it’s one company with $75m of investment and four companies getting hardly anything to speak of, to make it look like more companies are doing well than really are–the beauty of rhetoric by averages is that it eliminates undesirable information about what is going on).
*California had 233 exits last year, New York had 54, and Massachusetts had 53. Everyone else was bugs in the grass. So if you are starting companies like mushrooms out in Utah, and you have got 100s of them now, and you have made a big splash in the press about it, there’s this basic problem with exits. From the University of Utah point of view, at least at one time, the rhetoric about startups was very effective. It led the state to allocate $93m for a five-year program (called USTAR) to build on this “success”–that is, divert resources to what “appears to be working.” The result? At the end of those five years, the university could identify four startups with 13 employees. To their credit, the report authors emphasized that they could only track the economic contribution of the state’s spending, not the economic impact. That is, the contribution is what spending $93m does in the state’s economy, not what the results of spending that money does. But folks ignored all that and hyped the report’s findings as impact anyway. In reality, the spending’s contribution was barely a hiccup in the state’s overall gross domestic product, and there was not any innovation or impact worth mentioning.
This points to a huge gap in state and university economic development programs: they tend to spin their successes as if they were companies not subject to securities laws for forward-looking statements. If they were held to that standard, their reports would be actionable as fraud. And there is no question that most state and university reports on economic impact are intended to deceive. Those involved won’t put it this way, of course–they won’t ever out and admit it–they will say that they are rallying the public to this new initiative of the state and if everyone works together it could in the end live up to expectations but in the mean time the job is to inspire confidence in those expectations so that even if the program does not accomplish its stated objectives, it will look good, make those running it look good, and make those that decided to fund it look good–even if they should be shocked and dismayed. But it’s fraud. And coming from public sources, it is also a betrayal of public trust. Worse, money and opportunity gets wasted on administrators slurping up the fat of the land while innovation languishes.
What’s needed, then, is for a state or university to have a loyal opposition, a kind of independent auditor, like the teenager tasked with trying to break into the database, to test it for vulnerability. This role used to be filled by the press, but investigative reporting and challenging the powers that be has gone with corporate ownership of media and the disappearance of media competition within regional markets. This role also used to be filled by universities, but that was when they valued their independence and kept themselves out of institutional conflicts of interest. State economic development councils could play this role, if they chose, but for that they have to forgo being the cheerleaders and spend more time examining what is happening, questioning the nature of data being reported, and asking tough questions. An answer of “we don’t really know” would be, in all of this, refreshing, if true. At least policy-makers would have a level playing field.
In a number of university settings, folks put a lot of emphasis into how much investment money a company has raised, and then attribute that investment to the fact that at one point the company did a presentation at a university-hosted entrepreneurial event, as if the two are correlated. Rarely, they actually are. But most of the time, it’s just spin and hype, intended to impress–and deceive–a public with the sight of apparently big numbers. The reality is, if investors are putting a lot of money into a company, and there is no exit, then they are losing their investments. That big number bandied about by a university–“companies going through our new [each word is a separate link, btw] venture program raised $250m” means something along the lines of “investors have put this much up for potential loss, in part believing that an association with a university will mean greater honesty and mitigation of risks.” What the writers want people to think is “Gosh, that sure sounds like a successful program–don’t bother to check it for problems, be happy that something is working out.”
Yes, rounds of investment do indicate investment activity. And there is something useful in that. Investment can be a good thing. But so are loans, and especially so are profitable sales of new products. It is important that investment activity be put into context–what other investment activity is happening? Is the university self-investing but making it appear that the investments are independent? Is the university diverting private investment from other things that would be valuable to the community to its own techish startups? Any university determined to be honest in their communications will provide that context. But few do. You don’t read text that says, “Our startups raised $100m, but 99% of this figure was raised by one company that moved out of state three years ago but which we still like to think of it as a startup, and it no longer works on what it licensed from us in 2006. As for the rest of the new companies we have licensed to in the past three years, five didn’t raise any money, one has a 9-month SBIR grant for $100,000, and a couple got seed investments totally $375,000. None of these companies has yet to field a new product based on university-licensed technology, but we remain hopeful.” You don’t read anything like this, ever. There is a reason for that.
Most importantly, however, it is not the amount of money going into a venture, but rather the outcomes of that venture, and there are two critical outcomes: one is a company that becomes profitable through the sales of new products based on university discovery. This is the ground truth claim of innovation based on public funding of university research. It is amazing–and dismaying–that no university, and no professional organization–reports this activity on a regular basis. Yes, there are “success” stories, but there is nothing that accounts for first commercial sale or use of licensed technology, even though it is one of the few expressly required reporting elements under Bayh-Dole’s standard patent rights clause. It’s just that Bayh-Dole makes such reports exempt from FOIA, so we will never get the information from the government directly. And the universities are not offering the information, either. There’s a reason for that.
The other important outcome is that there is an exit for investors, whereby they recover their investment. In the VC world, this might be one in five, or one in ten, companies, but there are other forms of exit, with other kinds of private equity investment. In the coarsest, most material way possible to put it for universities: if they are trying to find instant wealth from startups, that wealth comes, if it comes at all, generally, from an exit, not from running royalties on sales of a licensed invention. An exit is not the same as a new product or use–an exit is not even a proxy for innovation. But it does show that one set of investors has persuaded another set of investors that something is valuable enough to acquire–and there are any number of reasons for that, including that the company is failing and is being offered at a really cheap price that allows one to poach the talent and take over the lease to do something quite different. But an exit does also indicate that there is a return on investment–whatever that may be, for better or worse–and it is in that return that a university licensing operation makes a major part of its case for being involved in company stocks.
In many cases–perhaps half–of the startups that I have been around, the licensed technology from the university is used to raise money, which is then used to hire engineers and scientists, who then build a product that the investors want to see, which often is not within scope of the patent(s) that were licensed from the university. Thus, one takes an interest in company equity as an alternative to putting some draconian diligence clause into the license that requires the company to build product within the license or else, which in turn makes it difficult for the company to raise investment funds, which in turn defeats the whole purpose of getting the deal with the university to legitimize the effort. In short, the patent exists to put a happy face on things, and is important to attract investors, but not because the investors need the safety of the patent in order to invest–the patent they will rely upon will be one filed by the company, on something probably related but not what was licensed from the university. So the university patent gets used, but not necessarily to promote the utilization of the university-hosted invention.
It’s true that Stanford made a huge amount in the Google IPO. But it makes absolutely no sense then to rig a program to focus on building companies that will go to IPO this way, just because it happened once at Stanford. Whatever one wants to say, one is not following Stanford’s “approach” in doing so, nor even “imitating” Stanford. There are ways to build companies to go to IPO, but universities are not doing that, and have no way, really to do that. And there are ways to model one’s program on Stanford’s. MIT, did, for instance. But then MIT is plop in the middle of a comparable tech environment, so there’s a little sense in it, at least.
Stanford’s licensing program has worked based on the “plant a lot of seeds and see what happens” approach. It “works” because it does not fixate too much on any particular way of doing business. Because Cohen-Boyer was a big time non-exclusive licensing program doesn’t mean that Stanford dedicates itself to finding another Cohen-Boyer. It might come up with something, and would know just how to do such a program again, but no, what it finds instead is a Google, and that’s because it licensed a software algorithm for enhancing internet search to a startup that discovered its income was from advertising, and in an IPO it caught the tail end of a speculative bubble of irrational exuberance in all things internet. But that doesn’t mean that Stanford then turns all its attention to starting more Googles. It might come up with something, and it would know what to do if it did, but no, it’s likely that the next big thing will not look like the past big things.
The approach Stanford uses is remarkably robust. Rather than fixating on some past success to repeat it by the imposition of administrative processes, as if by dint of proper planning and allocation of funds, one can do repeatedly what one just got hugely lucky about because one was open to possibilities. Once one has put even a few million into starting companies because once Stanford made a lot of money from one company IPO, there’s no going back to be open to other stuff. One plans, and spends, one’s way into a dark hole of unlikelihood, all the while making it plausible, if not attractive, for senior university leaders to be around it, cheerleading and supportive.
Nassim Taleb has laid this kind of situation out with regard to risk, with regard to rare events, the black swans of investing. Technology innovation has its own black swans–for good or bad. There is a long tail of prospects–many ways in which one might introduce innovation to the world, and profit from doing so, if seeking profit (rather than, say, honor, or duty, or delight, or service) is the thing that most drives you or your institution to action these days. Just because one prospect happens to go true in 2004 in the heart of Silicon Valley (or at least in Mountain View) does not mean that this is where the next prospect will also go true. Could be, but it is likely not to be. And that’s the problem for planning in innovation. There are too many prospects to do a decent job of investing in them all, and it is simply foolish to then pick one and invest in it, simply because someone else got lucky a decade ago. Chance may favor the prepared mind, but bureaucratic, institutional planning may eliminate the possibility of chance. And it is chance more than anything that drives innovation, not institutional plans. Just as it is impossible to address every possible risk that we face in our lives each day, nor even to adequately estimate the likelihood of the rare events–the asteroid explosion, the union strike, the cracks in the floating bridge pontoons, snow in May–it is also impossible to anticipate every rare opportunity for innovation.
Hayek made an argument for the conflict of central planning with personal liberty and innovation. Central planning–institutional planning–if it already knows where innovation is going to be, then is not planning innovation but something else, because innovation by its nature is a change in the established order, not an extension of that order. One cannot plan in that way for it. Yes, one can assert a direction and make a plan, but that direction and plan takes one element from one thousand, or ten thousand, and tries to make it come to pass, while the world wells up as it will anyway. All that the plan accomplishes, besides rolling the administrative dice, is the suppression of the institution’s own response to the rest of the world. So while innovation perks along in media, say, or mining, university administrators rig to get the next Google, as if it were 2004. But all they accomplish is attracting investors to their cause. They don’t have the exits to show their investment in startups amounts to a hill of beans–even if all they wanted from their activity was profits from handing off a speculation to new speculators, to heck with new products.
The reality is, innovation is a fickle thing that defies institutional planning. The practice of institutions generally is to suppress crazy innovation, to create stability and order, to evolve slowly and conservatively. If an institution wants to encourage innovation, then it has to run against its own values of orderliness, process, consistency, and mitigation of risk through proper paperwork. That does not mean not being prepared, but it changes what it means to be prepared. Being prepared does not mean having a plan, or program, or a focus, or hyped rhetoric that makes every institutional program into a success when it may be an abject failure.
For instance, at the University of Washington, they are now hyping the hope of creating of 20 companies in three years, which they claim will double their rate of startups. But UW reported creating 30 companies in three years starting in 2006–and dismantled that program in favor of a focus almost entirely on company creation. That is, they decided to plan for innovation to do more of what they were doing (or at least for huge amounts of income from companies that at least are made to look innovative) and even cited Google as a motivation for this “critical period of ascendancy” where UW’s reputation will be enhanced for starting such a company itself. UW has a “4-year vision” that coincidentally ends the year that the University’s primary source of income, the Hall patents, expire. One might expect the vision to expire then, as well, having exhausted all of its resources, as well as university reserve funds, and state and federal grant money running to the millions–in all well over $100m in five years.
University senior officials hope, publicly at least, to get back to 2/3rds of their previous startup output, but make it appear to the public as if they are going to double their output of companies–with no mention at all of any exits. A recent press release is full of information that doesn’t hold up. In the release, UW claims 8 startups in six months, but public information shows that they have started only 3 companies in the past year. (Two companies in six months, one with modest venture investment; one in seven months but with a four-year-old product and that company appears to be a shell for what would otherwise be a software distribution activity, not a startup; one is a not a startup at all, but a marketing name for a licensed product; another is a year-and-a-half old with no product or company presence on the web but for a splash screen created by the university licensing office, apparently; others were started two or three years ago.) That’s 1/3 the rate of 2008, before the university adopted their cutting edge “commercialization” model and the hyped rhetoric that goes with it. Sadly, this is a signature program of the university president, not some backwoods operation. It appears that the only thing left is for the university to figure out how to manage an exit.
*****
[Postscript, April 2016: In July 2014, at the height of her apparent success (judging by the hype, not the outcomes), the Vice Provost in charge of the UW licensing-cum-startup program suddenly quits. Six months later, in February 2015, UW president Michael Young abruptly leaves for Texas A&M. The Hall patents expired in April 2014, leaving the startup program with the prospect of only one more year of self-funding from royalties. UW promptly dismantled the startup program and in January 2015, rebranded what remained of the licensing program, dedicating it to promoting “entrepreneurial thinking, innovation mindsets, creative problem-solving, and experiential and team-based project learning” and generally aiming to promote regional economic development.
The new UW program aims to “go beyond IP and startups”–so, it largely abandons the reason why it existed in the first place, to manage IP brought to UW for management, confusing that with trying to make money for UW, which is what from 2008 to 2014 the recently deceased startup program claimed to be doing. Actually, what the program did was spend itself silly on itself and its image, burning through $100m+ when it should have been scaling back and building up a productive set of licensing relationships that could support–and justify–a new, much smaller scale (and ego).
Perhaps, too, what mattered is that UW’s startup-focused program took six months to do a deal with a startup backed by a UW regent, while Georgia Tech took only two weeks to negotiate its part of the technology package to the same startup. Perhaps, then, behind it all was lavish incompetence tolerated by UW officials so long as they felt flattered by the deceptive hype–“it might not be so true,” they may have reasoned, “but it sure feels good, gives us something to brag about. Success breeds success, you know.” No doubt some liars (to others, to themselves) fervently hope that the world will change and make their lies turn out to be true. Planning for innovation invites such a confusion between statements of fact and statements of desire. (Calling it “confusion” puts a nicer spin on it, doesn’t it?) Serious investors aim to separate out the hype and bullshit in an entrepreneur’s pitch to get at what is factual, reliable, worthy of investment. It’s easy to see how UW officials made for lousy investors, and bought a sales pitch uncritically. I chronicle the UW mess in various articles, including this one.
What continues to astound me is that no one at UW appears (publicly, at least) to care that UW wasted $100m chasing unicorns, paid off elected officials (one, perhaps two or more), lied to–or bullshitted (no regard for the truth)–the public and the state legislature about their performance, misrepresented their situation to the federal government to be part of a multi-million dollar grant to the state Department of Commerce, approved huge personal conflicts of interest, attempted to change university IP policy in absurdly draconian ways, and generally ruined what had been a modestly workable IP management program. Instead, the leader of this program was sent off with happy praise, having spent not only current funds but also reserve funds.
Perhaps the deal was, play nice (six months of extra pay, say, and keep up the facade that the program was wildly successful). Or perhaps when a prophecy fails (in this case, making so much money from startup equity that UW will be able to change its financial model with all the new income), folks making a public commitment to the prophecy double down and evangelize a changed purpose even more fervently. Or, perhaps it is just the embarrassment of realizing that one has been taken, duped out of a huge investment. Then again, maybe the UW program and those involved in supervising it have been so messed up that trying to come up with reasons for the mess is itself an exercise in metaphysical grieving. As the Pitchbook figures point out, the UW startup program, but for a thin sliver of wild luck, was doomed from the get-go. It was a $100m startup ruse that took in the senior management at a major research university on the premise that it would make them hundreds of millions.]