Andreessen, Venture Capitalism, Tech Bubbles and predicting Startup Success

Rather interesting talk by Marc Andreessen at D9 (as reported by Liz Gannes) - some nuggets:

If you look at the history of VC, the best firms were generally formed by operators. Over the years, there have been many different kinds of VCs, but we thought it would be kind of fun to go back and do what they did at the beginning.

We’re a throwback in two other ways. We’re focused on computer science and technology, and we’re overwhelmingly Silicon Valley focused. As far as we can tell, we will always be a single-office firm based in Silicon Valley. We think SV is special because there’s a repeatable process for building great companies.

Interesting observation - but I'm not sure they are quite a Back-to-Basics VC, as so far - if you follow the money - they have been investing big money in dead certs quite late in the pre-IPO game. More like Private Equity investment in (soon to be) Public Enterprises.

On Bubbles:

If everybody’s euphoric, then I’m concerned. “If we’re back here in three years and nothing’s changed and nobody’s worried, I’ll be horrified. I’ll wet my pants on stage.” There’s no history of an equity bubble that has not affected the public markets in a major way. Read “The Go-Go Years.” Fast-forward to today, in 2011: Apple’s PE is 12, projected to be 10. Microsoft’s is 7.2, next year 6.8. Google 13.7, next year 11.3. Cisco 7, next year 5.5. “PEs in single digits are what steel mills trade at before they’re going out of business.”

Maybe, but those are not the BubbleCo's - Google, Linked In and other Social media operators are where the bubble is, most with stratospheric P/Es (or virtual P/Es from secondary markets). Marc is quoted as saying "there is no bubble because everyone thinks there is one" elsewhere, but that is not how I read the above. I read it as saying there is risk of a bubble. I think Henry Blodgett is calling this more correctly now, and he should know....

....in 1999, there was actually lots of talk and concern that it was a bubble. But having lived through that one, I'd say there was actually less then there is right now. And the real challenge for decision-makers in that era, by the way, wasn't in determining whether or not it was a bubble--many smart folks thought it was. The challenge was in figuring out when the party was going to end. Because if the bubble lasted longer than you thought it would--which, by 1999, it already had--it didn't really matter whether it would eventually prove to be a bubble or not: You'd have been "wrong" and fired long before you got to do your victory dance.

This is what I believe is really going on in these funds' calculations, and Andreessen is in a roundabout way corroborating that.

And on dotcoms being a predictor of great Web 2.0 startups:

I just had this out-of-body experience where I saw someone reading on an iPad in the lobby and I had a flashback to the Newton. I’m quite confident that there are things we’re funding now that will be great 10 years from now.

Here’s my prediction: Almost every dotcom idea from 10 years ago that failed will succeed. Pet food, diapers, deliveries, they’re all working again now.

Probably true, for any pre-2000 company anyway. A lot of the dotcom ideas before the "end of days" lunacy were sound ideas but failed from a combo of too few customers, too thin bandwidth, not enough functionality in the infrastructure and too high a running cost. Much of this has changed hugely for the positive now.