Bubbles on the brain

It has become commonplace in Silicon Valley and in the blogosphere to take the position that we are in another bubble — a Web 2.0 bubble, or a dot com bubble redux.

I don’t think this is true.

Let’s examine the theory of a new bubble from a few different angles.

First, recall that economist Paul Samuelson once quipped, “Economists have successfully predicted nine of the last five recessions.”

One might paraphrase this for our purposes as “Technology industry experts have successfully predicted nine of the last five bubbles”… or perhaps more like five of the last one bubbles.

The human psyche seems to have a powerful underlying need to predict doom and gloom.

I suspect this need was evolved into us way back when.

If there is a nonzero chance that a giant man-eating saber-tooth tiger is going to come over the nearest hill and chomp you, then it’s in your evolutionary best interest to predict doom and gloom more frequently than it actually happens.

The cost of hiding from a nonexistent giant man-eating saber-tooth tiger is low, but the cost of not hiding from a real giant man-eating saber-tooth tiger is quite high.

So hiding more often than there are tigers makes a lot of sense, if you’re a caveman.

But as with other habits ingrained into us by evolution, the habit of predicting doom and gloom when it isn’t in fact right around the corner might no longer make sense.

On Wall Street, investors who have this habit are known as “perma-bears” and generally are predicting the imminent collapse of the stock market. This habit keeps them from being fully invested. Sure, they’re well protected during the occasional crash of 1929 or 2000, but by and large they massively underperform their peers who take advantage of the fact that most years, the economy grows, and the market goes up. They have disappointing careers and die unhappy and bitter.

In reality it seems very difficult to predict either a bubble or a crash.

Lots of people predicted a stock market crash… in 1995, 1996, 1997, 1998, and 1999. They were correct in 2000. But as soon as the stock market recovered in 2003 and 2004, they were back at it, and there have been similar predictions from noted pundits ever since — incorrectly.

Similarly, in the technology industry, there were people calling a bubble starting in 1995 and continuing through to 2000, with a short break for about two years, and then more bubble-calling ever since.

If you’re going to listen to people who predict bubbles or crashes, you have to be ready to stay completely out of the market — the stock market, and the technology industry — almost every year of your life.

Second, historically, bubbles are very, very rare.

It’s significant that in books and papers that talk about bubbles, there are simply not that many examples over the past 500 years of capitalism.

You’ve got the South Sea bubble, the Dutch tulip bulb bubble, the bubble in Japanese stocks in the 1980’s, the dot com bubble, and a few others.

They just don’t happen that often, at least in relatively developed economies.

And they don’t tend to happen more than once in a generation.

(Perhaps because many of the people who go through one are so traumatized that all they can do is sit around and worry about another one.)

Interestingly, modern economic research is in the process of debunking a number of historical bubbles.

It looks increasingly plausible that had US monetary policy been better run in the early 1930’s, our view of what happened in the 1920’s would be far more benign.

It also turns out that the Dutch tulip bubble is largely a myth.

So generally speaking, if one is going to seriously call a bubble, one has to be aware that one is calling something that is extremely rare.

Third, in the technology industry, lots of startups being funded with some succeeding and many failing does not equal a bubble.

It equals status quo.

The whole structure of how the technology industry gets funded — by venture capitalists, angel investors, and Wall Street — is predicated on the baseball model.

Out of ten swings at the bat, you get maybe seven strikeouts, two base hits, and if you are lucky, one home run.

The base hits and the home runs pay for all the strikeouts.

If you’re going to call a bubble on the basis of lots of bad startups getting funded and failing, then you have to conclude that the industry is in a perpetual bubble, and has been for 40 years.

Which may be fun, but isn’t very useful.

Lots of people running around starting questionable companies, launching marginal products, pitching third-tier VC’s, throwing launch parties, shmoozing at conferences, blogging enthusiastically, and otherwise acting bubbly does not a bubble make.

That’s just life in this business.

Note also what you don’t see in the theoretical Web 2.0 bubble of 2007.


Lots and lots and lots of IPO’s.

For a theoretical bubble, that is just plain odd.

Fourth, getting more specific about Internet businesses — things have changed alot since the late 90’s.

It is far cheaper to start an Internet business today than it was in the late 90’s.

The market for Internet businesses today is much larger than it was in the late 90’s.

And business models for Internet businesses today are much more solid than they were in the late 90’s.

This is a logical consequence of time passing, technology getting more broadly adopted, and the Internet going mainstream as a consumer phenomenon.

People smarter than me have written about these factors at length elsewhere, so I won’t dwell on them, unless there is specific interest.

But my back of the envelope calculation is that it is about 10x cheaper to start an Internet business today than it was in the late 90’s — due to commodity hardware, open source software, modern programming technologies, cheap bandwidth, the rise of third-party ad networks, and other infrastructure factors.

And the market size for a new Internet business today is about 10x bigger than it was in the late 90’s — there are about 10x more people online (really!), and they are far more used to doing things on the Internet today than they were in 1999.

(Want evidence of that last point? Clothing purchases are now bigger than computer hardware and software sales online. I can guarantee you that nobody who was involved in ecommerce in the mid-90’s ever would have predicted that.)

The Internet is a fully mainstream medium now, people love it, people are willing to do all kinds of things on it, and it’s getting really cheap to offer new services to those people.

Fifth, and finally, there’s the simple fact that the Internet businesses that are succeeding in 2007 are for the most part incredibly valuable, compelling services that lots of people like and that are in general either making a lot of money or will be making a lot of money quite quickly.

People laughed when Fox bought MySpace for $580 million, but that’s a business that will generate nearly $300 million in revenue in 2007, and more in 2008.

As an independent asset today, MySpace would probably be valued at between $3 billion and $5 billion today — perhaps higher.

Call that the deal of the decade.

Similarly, Facebook is bringing in a lot more revenue than people think.

And then there’s Google.

These companies aren’t pulling in all that revenue via some kind of Ponzi scheme.

This is money coming from real advertisers and real users for real services with real value.

Which makes total sense, amid the enormous mass migration of consumer time and attention away from traditional media towards online media.

These same factors apply all the way down the foodchain.

A high-growth online startup that gets bought for $100 million or $200 million by a large Internet or media company isn’t getting that kind of acquisition price just for the hell of it, but rather because the acquirer can plug that startup’s service into its broader portfolio of services and make real money with it.

These are big numbers, but remember, there are more than a billion people online now. That is a very large market — a lot of people, spending a lot of time, buying a lot of things, in totally new ways at the same time as they are abandoning older services like newspapers, magazines, television, movie theaters, and print catalogs.

So, my view is that to call a bubble, you have to find evidence of it outside of the mainstream of the kinds of Internet businesses that are being built, sold, and run in 2007.

In closing, I’d be the last person to say that I never roll my eyes at the next startup that’s doing online wiki-based popularity-ranked video-podcast mobile social dating widgets for the dog and cat owner market.

But a bubble?

I doubt it.

Postscript: I didn’t see this until after I wrote my post, but you have to love the opening line: “Grandpa lived through the Depression, and life thereafter was indelibly shaped by haunting memories of soup kitchens and hobos.”