Why going public sucks: it’s not governance issues

By Felix Salmon
July 17, 2012
Marc Andreessen agrees with me that it sucks to be a public company. But he disagrees on the reasons why:

" data-share-img="" data-share="twitter,facebook,linkedin,reddit,google" data-share-count="true">

Marc Andreessen agrees with me that it sucks to be a public company. But he disagrees on the reasons why:

Basically, it was very easy to be a public company in the ’90s. Then the dot-com crash hits, then Enron and WorldCom hit. Then there’s this huge amount of retaliation against public companies in the form of Sarbanes-Oxley and RegFD and ISS and all these sort of bizarre governance things that have all added up to make it just be incredibly difficult to be public today.

Andreessen, of course, as the lead independent director at Hewlett-Packard, knows all about “bizarre governance things”. But Sarbanes-Oxley and RegFD and ISS don’t even make it into the top ten. And as for the idea that they’re “retaliation against public companies”, that’s just bonkers; in fact, it’s really rather worrying that we have directors of big public companies talking that way.

The three things mentioned by Andreessen come from three different places: Sarbanes-Oxley came from Congress; RegFD came from the SEC; and ISS is a wholly private-sector company which allows shareholders to outsource the job of ensuring that governance at the companies they own is up to par. If there wasn’t significant buy-side demand for such services, ISS wouldn’t exist.

It’s directors’ job to care deeply about governance, rather than to dismiss serious concerns from legislators, regulators, and shareholders as “bizarre governance things”. And for all that companies love to bellyache about the costs of Sarbox compliance, the fact is that if you’re looking to governance issues as a reason why it sucks to be public, you’re looking in entirely the wrong place.

The real reasons it sucks to be public are right there in the name. Being public means being open to permanent scrutiny: you need to be very open about exactly how you’re doing at all times, and the market is giving you a second-by-second verdict on what it thinks of your performance. What’s more, while the glare of shareholder attention on the company can be discomfiting, it tends to pale in comparison to the glare of public attention on the company’s share price. I was on a panel last week talking about IPOs in general and Facebook in particular, and I was struck by the degree to which Mark Zuckerberg’s enormous achievements in building Facebook have already been eclipsed by a laser focus on his company’s share price.

Zuckerberg — or any other CEO — has very little control over his company’s share price, but once a company is public, that’s all the public really cares about. If Facebook’s share price falls, all that means is that external investors today feel less bullish about the company than they did yesterday; if it falls from a high level, that probably just says that there was a lot of hype surrounding the Facebook IPO, and that the hype allowed Facebook to go public at a very high price. Facebook could change the world — Facebook has already changed the world, and did so as a private company — but at this point people don’t care about that any more. All they care about is the first derivative of the share price. And maybe the second.

When companies go public, people stop thinking about them as companies, and start thinking about them as stocks: it’s the equivalent of judging people only by looking at their reflection in one specific mirror, while at the same time having no idea how distorting that mirror actually is. Many traders, especially in the high-frequency and algorithmic spaces, don’t even stop to think about what the company might actually do: they just buy and sell ticker symbols, and help to drive correlations up to unhelpful levels. As a result, CEOs get judged in large part by what the stock market in general is doing, rather than what they are doing.

And meanwhile, CEOs of public companies have to spend an enormous amount of time on outward-facing issues, dealing with investors and analysts and journalists and generally being accountable to their shareholders. That’s as it should be — but it isn’t pleasant. When Andreessen says that it was very easy to be a public company in the 90s, he’s right — but he’s wrong if he thinks the 90s were some kind of halcyon era to which we should return. They were good for Andreessen, of course, who took Netscape public in a blaze of publicity and more or less invented the dot-com stock in the process. But they were bad for shareholders, who saw their brokerage statements implode when the bubble burst. And, as Andreessen rightly says, they were bad for the broader institution of the public stock market, which has never really recovered from the dot-com bust.

As a venture capitalist, Andreessen has a fiduciary responsibility to his LPs, and he needs to give them returns on their investment, in cash, after five or ten years. It’s a lot easier to do that when you can exit via an IPO. But the way to make IPOs easier and more common is not to gut the governance that’s in place right now. Instead it’s to embrace it, and make public companies more like private companies: places where management and shareholders work constructively together and help each other out as and when they can. Mark Andreessen is a very active and helpful shareholder of the companies that Andreessen Horowitz invests in — and he has a level of access to management and corporate information that most public shareholders can only dream of. If he likes that system, maybe he should start thinking about how to port it over to public companies, as well.

Update: See also this great post from Alex Paidas, who’s found a quote from Andreessen saying that all of his companies should have a dual-class share structure. Despite the fact that I can’t ever imagine Andreessen himself being happy with non-voting shares.

9 comments

Comments are closed.