AT&T is scheduled to reach 30 million U-verse homes passed by the end of the year with their U-Verse service, or roughly 55-60% of their homes. They will virtually stop there according to President John Stankey speaking at Citibank, who announced 55-60% as their ultimate goal. He suggested that 25-30% of AT&T homes will only be offered ADSL. 20% are "not a heavy emphasis for investment," i.e. 5-10 million of AT&T's 50 million homes are screwed unless they have a decent cable alternative. (Yes, rounding means not necessarily equal to 100 %.)

U-Verse is holding and winning customers much better than I and others expected. People like the Microsoft IPTV interface, with reporters in the Houston Chronicle preferring U-Verse over cable. The result is nearly a million U-Verse new customers a year, while AT&T is bleeding customers where they don’t have U-Verse. They are going from 1-3 meg DSL to 10 meg cable, but the 50 and 100 megabit cable isn’t selling because of the $100 typical price. If cable reduced the 50 meg price to the French or British $30-45, that would change. For now, however, U-Verse is doing well.

CFO Ritcher told investors "Consumer wireline is growing again, thanks to our U-verse product. We're really just seeing the benefits of scaling the service and it gives us great momentum. U-verse is transforming our consumer results. Where we offer U-verse, our ARPUs are higher, our churn is lower, our customer satisfaction is better across all of our products."

I didn't expect so severe a cutback, and wouldn't be surprised if they reversed it eventually. I checked with AT&T whether Stankey mis-spoke, but not so according to his pr people. Virtually stopping means that in many cities U-Verse will look like Swiss cheese, with 10-50% of homes in holes that can't be served.

I'm guessing that what's going on is that AT&T decided they had no choice but to raise capex on wireless, accelerating the LTE build to prevent falling too far behind Verizon. Expanding U-verse to 75-85% is almost surely profitable after cost of capital, which to a purely theoretical economist would suggest they would just borrow modestly and do both.

Wall Street perceptions make that virtually impossible, however. It's irrational to underlying value, but stocks get clobbered for raising capex if that has a good underlying return. Remember the madness of Wall Street crowds. Even if it's good business to invest, the current mood on the street penalizes you. One day that may turn around, and the street will re-embrace Graham and Dodd and look at the underlying value of the stock. For the last few years, they've emphasized short term cash flow instead, leading to irrationally low investment in networks.

For example, Jason Armstrong of Goldman Sachs wrote "The Bear Case from 4Q10 Results." He feared the stock price would go down with "higher capital spend impacting carrier FCF. Several companies highlighted plans to increase capital spending in pursuit of wireless data growth, fiber-to-the-tower, or higher customer acquisition costs. Notably, companies with active capital allocation policies are among this group, including AT&T, the RLECs, and DTV."

Armstrong believes increased capital spending brings down the stock price even if earnings may go up. I defer to Jason on analyzing stock prices, especially because his peers like John Hodulik see similar. You make money in the market by guessing what investors will actually do, not telling them a theoretical economist would disagree with them. Market distortions like this one have a dire effect on the real world, reducing investment the country needs. They are as irrational as the over-exuberance of tech in 1999, which led to investments that were unsustainable.

Any competent CFO today would tell his boss "Don't raise capex if you can help it."