In 2008, he said:
The worst sort of business is one that grows rapidly, requires significant capital to engender the growth, and then earns little or no money. Think airlines. Here a durable competitive advantage has proven elusive ever since the days of the Wright Brothers. Indeed, if a farsighted capitalist had been present at Kitty Hawk, he would have done his successors a huge favor by shooting Orville down.
The airline industry’s demand for capital ever since that first flight has been insatiable. Investors have poured money into a bottomless pit, attracted by growth when they should have been repelled by it. And I, to my shame, participated in this foolishness when I had Berkshire buy U.S. Air preferred stock in 1989. As the ink was drying on our check, the company went into a tailspin, and before long our preferred dividend was no longer being paid. But we then got very lucky. In one of the recurrent, but always misguided, bursts of optimism for airlines, we were actually able to sell our shares in 1998 for a hefty gain. In the decade following our sale, the company went bankrupt. Twice. (http://www.berkshirehathaway.com/letters/2007ltr.pdf)
Tech is also weird in that software development has significant diseconomies of scale, while software operations has significant economies of scale. Small teams can develop a product faster, but once the product exists and has been proven in the marketplace, it's cheaper to scale from 100 servers to 100,000 servers than it was to go from 1 server to 100 servers. This has historically fueled the cycle of small tech startups being founded, gaining traction, getting bought, and then the founders quitting to found other small startups. It's economically rational for them to do so, because they can develop a product faster with a small team than inside a big organization.
I'm curious how the existence of cloud-hosting like Amazon EC2 will change this market. That's had the effect of splitting software development and software operations into separate markets: before, it used to be cheaper to integrate them into one firm, but now a commodity product exists that makes it feasible for the development firm to remain independent and simply pay a fee for all the operations support.
I suspect it's actually bad for startups in the long run, much like the power loom was bad for textile manufacturers. Since there're economies of scale to operations, that side of the market will tend to a few big players (right now Amazon has a virtually monopoly, and only a couple others even have the capability to offer something like that), while the startup side will tend towards many small firms. In a situation like that, the big firms have all the bargaining power, and so most of the profits will accrue to them.
Of course, software development doesn't have to start out capital intensive. A couple of kids working on their startup in a cheap apartment can work cheaply. They can defer the capital intensive part until they are successful. An airline doesn't have that option.
The costs of servers scale extremely well, while the costs of airplanes and jet fuel don't. So web companies have much higher potential profits than airlines, and they're consequently worth a lot more.
...that was the point all along.
So web companies have much higher potential profits than airlines, and they're consequently worth a lot more.
This isn't a competition between tech companies and airlines. Airlines merely act as an example an industry which, at least as of 1995 when Buffett first brought them up, had a net loss over the entirety of the industry's existence. What's the net lifetime profit of the tech sector? What will be the lifetime profit of social websites in another ten years?