Friday, March 5, 2010

What Does 100 Mbps for 100 Million Homes Imply About Monthly Prices?

There are only two major problems with the Federal Communication Commission's upcoming National Broadband Plan, says Dan Hays, PRTM director. The aspirational goal of 100 Mbps service provided to 100 milliion U.S. homes by 2020 is a fine aspirational goal, but it isn't clear how it can be implemented, he says.

The other big initiative is in the wireless area, specificially the effort to get TV broadcasters to give up 500 megahertz worth of spectrum so that wireless service providers can use it.

And there are just two problems there: buyers and sellers. Are there willing buyers? Are there willing sellers? Hays says the broadcasters already have said they are unwilling to sell. Even if they do agree to sell, the cost to acquire that much spectrum would be quite expensive, coming at a time when service providers are straining to justify further investments in their fixed plant.

The 100-Mbps access plan likewise has just two problems: who will pay for the investments, and whether end users are willing to pay substantially more than they now do for the upgraded speeds.

Estimates of how much that might cost range from a wildly-low $25 billion up to $350 billion, says Hays.

Commercial organizations aren't terribly interested in investing now to provide speeds that high, as there is little consumer willingness to pay much more than what people pay today, Hays says.
The percentage of household income spent on communications in the United States is in line with the rest of the world today, he says. So it does not make sense to assume a step level change in spending even if much-higher speeds are made available.

Beyond that, there is a generational and cultural issue at work. A good percentage of broadband non-adopters are older than 65. But as younger users who "cannot live without broadband" move up the age cohort, that particular non-adoption issue fixes itself, says Hays.

Once upon a time one could hear many doubters about why people would buy cable TV when they could get off-air video for no incremental cost. Over time, people decided they really did need it. The same thing has been at work with mobility services and will be true of broadband as well.

None of this is to argue that, over time, access speeds will increase. But investment capital cannot be raised unless there is a plausible business case. So the catch is that investors will want to see some plausible evidence that $300 billion worth of investment will lead to a return on investment.

Assume there are 65 million U.S. households spending $40 a month, on average, for broadband. That works out to about $2.6 billion a month, or $31 billion a year.

Assume a base of about 115 million U.S. households. Assume 90 percent broadband penetration, or 104 million connected homes. That would represent about $4.2 billion in monthly spending, or about $50 billion in consumer revenue.

How much incremental revenue would an investor want to see to justify investing $300 billion? Assume a 15 percent return as a hurdle rate and 104 million customer households.

The debt service implied is $45 billion a year. Assume borrowers also want to repay the principal over a 10-year period. That would very roughly imply a need to earn an incremental $75 billion a year.

Assume 40-percent margins. That implies gross incremental revenue has to be about $187.5 billion a year, or $1630 per customer, or an additional $136 per customer, per month. That implies a monthly price of at least $175 a month.

Do you really think every broadband customer in America, at 90 percent penetration of homes, is willing to pay $175 a month for broadband access?

Of course, maybe I have blown the math here. If not, I think 100 Mbps access, using networks built with private capital, are unlikely to happen. It would require consumers to do something history and logic suggests they will not do.

2 comments:

Cyberdoyle said...

If there was fibre to the home then people could buy as much speed as they wanted or needed. here in the UK many small firms who can't afford their own line willingly pay £60 for one Hour of 100meg to transport their files... Also if there was fibre instead of copper there would be no bottlenecks and data would flow freely without it being a contended throttled service, and once that happens everyone goes faster. Bandwidth needn't be rationed any more, and will be cheap.

Anonymous said...

Cyberdoyle

There is no difference between copper and fiber when it comes to throttled services. The limiting factor is still that every ISP need access to the network and just like retrans for Cable TV the price keeps getting bigger and bigger, but ISP and Cable providers can not charge customers more with out negative impact.

Many Winners and Losers from Generative AI

Perhaps there is no contradiction between low historical total factor annual productivity gains and high expected generative artificial inte...