Venture capital has a self-dealing problem

This week, I did something very dumb: I got into a Twitter fight with a bunch of venture capitalists.

The proximate cause of my fight was this Wall Street Journal article about 21, a secretive Bitcoin start-up that has raised $116 million in venture capital, with much of it coming from Silicon Valley mega-firm Andreessen Horowitz. What made this particular story interesting was that 21’s co-founder, Balaji Srinivasan, has a day job as a partner at Andreessen Horowitz, the same firm that funded his company. I’ve since learned that Srinivasan was working on 21 before he took a job at Andreessen Horowitz (and is moving to a part-time role at the firm in order to focus on 21 full-time), but at the time, the whole thing had the feel of a venture-backed ouroboros. So I tweeted this:

https://twitter.com/kevinroose/status/575327668159188992

The answer, I was quickly informed, is that not only are self-dealing arrangements considered kosher in Silicon Valley, they’re everywhere. Khosla Ventures partner Keith Rabois is also the co-founder of house-hunting start-up OpenDoor, which raised $9.95 million in a Khosla-led round last year. Workday co-founder Aneel Bhusri was a VC at Greylock Partners when the firm led Workday’s $15 million round in 2005. Peter Thiel, the consummate player-coach, has played both the investor and co-CEO roles at Palantir.

These aren’t like Entrepreneur In Residence arrangements, in which a VC firm hires a start-up founder in order to incubate his or her fledgling company in-house, using the firm’s connections and resources in exchange for a hunk of equity. And they’re not strategies like the ones employed by “start-up studios” like Betaworks, where the entire point of the firm is to find promising sectors in which to plop down new businesses. These are a different beast altogether—big, established VC firms using outside money to fund new start-ups run by their own investing partners.

In trying to explain why deals like these seemed ethically suspect, I invoked a (dated) pop-culture analogy:

https://twitter.com/kevinroose/status/575486009888002048

At which point the invective really started. VCs alleged that I was ignorant about the mechanics of VC, that I didn’t get why a VC firm would invest in a partner it trusts over one it doesn’t know, that I was simply stirring shit. (For a bunch of dudes with more money than Croesus, venture capitalists sure are tetchy!) So, in the interest of taking on all my Twitter foes as a consolidated whole, I put together a list of some of the criticisms I got and responded to them.

What do you mean unfair? VC investors are uniquely positioned to create great start-ups! And all of these start-ups have gone on to make billions of dollars!

This was the most common response I got, and I have to say, it puzzles me. Because it doesn’t really address the central conflict of interest. As I put it on Twitter: Sure, maybe Simon Cowell is an incredible musician, and would win American Idol on his own merit. But you still wouldn’t trust a system that awarded him the top prize. Because Simon Cowell has information about American Idol that most of us don’t. He knows which judges like which Aretha Franklin songs, which judges are suckers for Kenny Loggins ballads, and exactly how to structure an audition to make it to the next round. And, more to the point, in the early rounds, he’s one-third of the people doing the choosing.

In venture capital, unlike in TV competitions, unfair advantages are seen as a good thing. (In his reply to me, Rabois touted “asymmetric ability & info” as a crucial part of what makes VC investors qualified to start companies.) Part of the reason you would invest in a VC’s start-up over a non-VC’s start-up is that the VC knows things others don’t. He has an informational edge.

But how do you think he got that edge? In many cases, it was by sitting through hundreds of pitches by entrepreneurs who don’t work at venture capital firms, and who may have divulged proprietary information about their businesses in the course of asking for VC money.

Looked at through the eyes of a spurned start-up founder, the ethical conflict is clear as day. How does an entrepreneur know whether the VC he’s pitching his burrito-delivery app to is earnestly hoping to invest in him, or whether he’s fishing for trade secrets to start a burrito-delivery app of his own? (One start-up founder, whom I spoke with after 21’s $116 million funding total was announced, told me, “If I’d been in [Andreessen Horowitz] pitching a consumer Bitcoin wallet or chip-based service in the past year, I’m throwing up.”)

Not all of these fears are founded, of course, and the best VC firms have conflict of interest policies that mitigate some of these risks. (A person close to Andreessen Horowitz tells me, for example, that the firm hasn’t taken any Bitcoin wallet pitches since it invested in 21 and Coinbase, another Bitcoin start-up.) But these policies don’t exist everywhere, and even the ones that exist aren’t totally reassuring to start-up founders. And I don’t blame them. Now, in the most competitive start-up environment in history, these founders no longer just have to worry about being beaten by other start-ups in their space. They have to worry about potential investors picking off their ideas and using them for personal gain.

Venture capital isn’t supposed to be fair! It’s supposed to generate returns for investors. And the investors don’t care about self-dealing, so what’s the issue?

Many of the responses I got boiled down to, “So what?” As long as both the LPs (the investors in a venture capital fund—who can include public pensions and endowments, along with sovereign wealth funds and other institutional investors) and the GPs (the investing partners who work at the venture capital firm) are okay with the self-funding arrangement, what’s the harm? Many of the companies started by VCs have gone on to make billions of dollars for their early backers, after all. And isn’t that a VC firm’s mandate—to get into top-tier deals and multiply investors’ money by any means necessary?

Well, first, we don’t know that all LPs know what’s being done with their money. I would love to ask the investment officers at a firefighter’s pension fund in Wyoming what they think of venture capital GPs using their money to fund their colleagues’ personal projects.

But even if all LPs know about self-dealing and have signed off on it, that doesn’t cleanse the conflict of interest. (To use a different analogy: Goldman Sachs’s shareholders would love it if the bank could tear down the Chinese wall separating its investment banking division from its research division—they would make a lot more money!—but the bank isn’t allowed to do it, because we’ve decided that having those walls in place is in the public interest.)

There’s also no guarantee that venture capital LPs will be happy with these arrangements forever. It’s easy for an LP to support self-dealing when everything is going up and to the right—the start-up’s making money, the investors are making money, incentives are aligned, everyone’s happy. But when the company goes downhill, the incentives get tangled. Now, the founder’s job is still to keep the failing company afloat, but the VC’s job, in part, is to get the VC firm’s capital back. If OpenDoor is failing, do Khosla’s LPs really trust that Keith Rabois will run the business in a way that allows them to recover their capital, rather than running it to preserve his founder’s stake?

VC isn’t a zero-sum game. One person’s win isn’t another’s loss.

Chris Dixon, an Andreessen Horowitz partner whose opinions I respect, said this in a reply to my original tweet:

I can see his point. In theory, venture capital is a benevolent capital allocation scheme that enables the growth of companies whose products improve society. And if those companies happen to be started by venture capitalists, so what? The good start-ups will rise to the top naturally, benefiting everyone, and the bad ones will be destroyed by the market.

But in practice, venture capital is a battle like any other, with winners, losers, referees, and a finite supply of money to go around. Before the market gets to choose which start-ups succeed, there are a series of gatekeeper institutions that give their imprimatur. And these institutions matter. A dollar from Andreessen Horowitz is worth much more to a start-up than a dollar from a no-name venture capital firm, simply because Andreessen Horowitz has so many connections and resources that it can create successes where others can’t. The same is true of Y Combinator, which accepts fewer than 3 percent of the companies that apply, and greatly increases their chance of survival.

If the tech industry is as important to America’s future as Silicon Valley thinks it is, they should want these gatekeeper institutions to be as meritocratic as possible. And that means making sure that they operate with as little in-group prejudice as they can. Sometimes, that might mean missing out on a big deal over a conflict of interest. But it would give aspiring tech founders some confidence in the integrity of the system they’re entering.

Also: given the amazingly homogenous nature of the VC industry’s senior ranks (96 percent men!), we should oppose the self-funding VC trend out of diversity concerns alone. VCs are already a clubby, cliquish bunch. Why should we institutionalize one more way for rich men to throw money at each other?

But everyone in VC does it!

I got this one a surprising number of times. And I guess I don’t know what to say in response, except that universal corruption is still corruption.

Right now, the VCs who are self-dealing are mostly well-known investors at top-tier firms. But what happens if this trend spreads down-market? Would all the people supporting Keith Rabois and Balaji Srinivasan’s new ventures be okay with a mid-level VC taking investor money to make yet another Snapchat clone? What if the VC wanted to start a Mexican restaurant? What if the VC made a really, really convincing pitch to his fellow investment committee members, and promised them that his restaurant would be a $50 billion unicorn that would blow away Chipotle and Taco Bell? Would they be willing to shovel a few million dollars from the Oklahoma Teachers Retirement System into his pockets?

I don’t think they would. And I think that, as more venture capitalists strike these kinds of deals, there will be a backlash from start-up entrepreneurs, who will see it as both a barrier to their success and an ethical morass in its own right. As one astute observer put it to me recently, “It’s a tight group of millionaires giving each other outsiders’ cash to buy lottery tickets that might make them billionaires.”

So, VC is corrupt and clubby. Duh. What are you proposing we do about it?

This was the portion of the Twitter fight that flummoxed me the most. Because I don’t think venture capital investments should be regulated by the SEC in the same way that, say, hedge fund investments are. And I don’t necessarily think that the solution to VC self-dealing is more regulation.

But I do think that top-tier VC firms should take the ethical high ground, and institute some conflict-of-interest rules around these deals. Maybe it’s as simple as “we won’t invest outside money in start-ups in which our current GPs hold executive-level positions.” Or maybe it’s a little more nuanced—something like “if we’ve taken a pitch in a given space, we won’t invest in a GP-led start-up in the same space for a period of at least two years.” Maybe it’s a financial arrangement that preserves the ability to invest in GP-led start-ups while reducing the temptation to double-dip: “If we invest fund money in a GP’s start-up, that GP forfeits a corresponding amount of his carry from that fund, and is no longer allowed to attend pitch meetings in the same space.”

Rules like those might produce some weird externalities, but they’d be better than nothing. Most big venture capital firms already have conflict-of-interest policies that prohibit certain GP activities, so adding a rule preventing self-dealing would be easy enough. And really, what’s the harm of having fewer self-deals? That amazing, GP-led companies won’t get funded? If Peter Thiel’s next start-up idea is so hot, there will be plenty of people lining up to invest—there’s no reason the money has to come from Founders Fund.

Changes like these wouldn’t fix venture capital’s clubby, back-scratching tendencies overnight. But they would go a long way toward making the system fairer for entrepreneurs. And they would reassure venture capital LPs that their money is truly being spent finding and funding the best start-ups in the world, rather than simply being funneled into the pockets of a preordained few.