[pg]

There is a difference between no plans to exit, no exit occurring. No exit ever occurring is currently going to be a failure for any technology investor, because currently, at least, there is no custom of technology companies (unlike perhaps restaurants) paying dividends. Maybe such a custom will develop, but there isn't one yet, and it's not what YC is designed for.

The best illustration of the distinction between no plans to exit and no exit occurring is 37signals. Jeff Bezos's investment in them is a sign that he believes there will eventually be an exit. As far as I know he didn't insist on them paying him dividends; that would have been very unusual; so without an exit there would be no way for him to get his capital back.

We too would have bet on 37signals, because it's hard for technology companies to stay medium-sized. They either peter out, in which case exits are a moot point, or they grow so large that they either go public or eventually receive an acquisition offer the founders are willing to take.

[staunch]

FYI: 37 signals is an LLC and pays out their profits at the end of each year. So Bezos is getting a check once a year (though obviously that's not why he invested).

[pg]

Wow, I didn't realize that. It will be interesting to see how this turns out.

[dood]

Can you imagine a future where investors somewhat like yourselves (if not necessarily YC) invest considering their expectation to likely be a combination of dividends and some possibility of exit?

Or perhaps better phrased, do you see dividends playing a role in angel/YC-like investment decisions anytime in the nearish-term?

[pg]

It seems very, very unlikely. Dividends have been done before. They were how "startups" used to pay investors back in the railroad days. But the rates had to be set in advance. Without preset rates, investors would have to trust company managements not to skim profits (which there are 101 semi-legit ways to do) and claim there were none to return. And while a railroad could predict profits with reasonable certainty, how could a tech startup?

[patio11]

If you do not trust management, they can act to detriment of investors even without dividends. All they have to do is pay themselves market wages, for the right value of "market", rather than investing in growth. Or they can run company like it was Japanese, where it exists to buy perks for employees off their own taxable income. (The four star chef is for recruiting! Honest!)

This is a two way street, since investors have an incentive to push for swinging for the fences even when a bunt would be life changing for the founders. Consider a two man team who hits a million in sales, but seems to stall out (say, ran out of channel, but has good ongoing relationships with customers). Pivot and risk company to hit ten million, which would justify exit, or continue executing and make two families rich with little risk? Not hard to see dynamic.

[pg]

Larger funding rounds include provisions to protect against such abuses. The ultimate of course is the right to replace the CEO if necessary. In practice they seem to work.

[AmberShah]

Could this be accomplished by paying company management a low but reasonable wage and having them also get paid dividends? That way they'd be motivated to have more profits and get more money through dividends themselves. I suppose in the 101 semi-legit ways there's probably still a way around this.

[dood]

Makes sense, thanks. Seems a shame that there is no reasonable way to invest in the companies which have good potential for growth and profit, but have no clear exit.

[garply]

Why would you do business with, let alone invest in, people whom you did not trust?

[pg]

The problem is the difficulty of the second half of "trust but verify."

[AmberShah]

What I read about VC's definitely leads me to believe that you should have an exit plan and be on the same page with them about it. If I wanted to tell them, come back in 5-10 years and I'll think about it, that would be a problem. Perhaps it is not true or perhaps a seed fund like YC is different.