Is Wholesale Key for FTTH Deployments?

Are telephone companies uniquely unable to fund fiber-to-home upgrades? Analysts at the Yankee Group say that might be the case.

“Ultimately, the long payback periods of any FTTH deployment of significant scale suggest that perhaps traditional telcos are not the players best suited to deploy and manage the next-generation networks,” says Benoît Felten, Yankee Group principal analyst.

Yankee Group’s analysis is based on fixed cost per home passed at $1,000 and cash margin for FTTH at 45 percent, as well as two variables, penetration and average revenue per user.

The model is a lot more sensitive to penetration than it is to ARPU, Felton says. “In other words, it’s more important to connect a high proportion of the customers targeted by the network than to make them pay a lot for it.”

It’s virtually impossible for FTTH to pay for itself in less than five years unless penetration is at least 30 percent, and even then a time frame of seven to eight years is more realistic, Felton notes. That would be the case for a network with ARPU of $80 per customer per month.

But it is worse. That initial calculation does not account for the cost of borrowed money to finance the builds. Discounting for the cost of capital at 12.5 percent, the payback takes considerably longer.

In fact, says Felton, payback takes 16.5 years. That is going to be a sobering fact for executives whose typical framework is payback within five years.

There are some conceivable variables that could improve payback time. Service providers can take steps to lower the investment cost below $1,000. In most cases that will be politically impossible though, as it requires limiting investments to areas where construction costs are lowest.

That could mean areas with easily-available aerial attachment rights or ducts with spare capacity. It almost certainly means areas of reasonable customer density.

Dense urban areas with mostly multi-tenant accommodation likely can be built for $500 per home while lower-density urban areas with a higher proportion of single-tenant homes can be built for about $750 per home, says Felton.

Another way to lower the deployment costs is to make the customer contribute to part of these costs. There are three essential ways to significantly decrease the costs per home connected, says Felton. Service providers can target only high-density and, or, low-deployment cost areas.

They can add drop cables and gear only for customers who sign up for the service.

The two-step build separates the trunk network construction from the “activating service” operations. The initial deployment brings the fiber within a few meters of the customer’s premises, but the home drop is only built if and when a customer actually signs on to the service.

Verizon, for example, has done this, allowing the company to save about $600 for each home passed that is not a current customer.

This clearly improves the business case in the early stages and converts a significant amount of capital expense to the “success-based” model.

The drawback is economies of scale since each new customer addition requires an additional truck roll, but such install activities would be required in any case. The approach simply requires more labor time, as a new install might require burying a new drop cable.

The two-stage model is penetration sensitive, though. Most of the gain occurs when penetration is expected to remain below 50 percent. Above that level, the one-step build actually costs less.

A variation of the “lower drop cost” method is to require customers to pay a connection fee to access the network. You can decide for yourself whether homeowners and renters are likely to be willing to pay such a connection fee.

It likely is politically impossible to limit the upgrades to areas where construction costs are favorable. It likely is economically or competitively impossible to get homeowners and renters to pay for a portion of the upgrades.

Perhaps new services will emerge that can drive penetration above 30 percent. So far, that has proven difficult, though.

That leaves only the wholesale model. That said, wholesale access could reduce the payback model to nine years or less, Felton argues.

“The history of copper unbundling in Europe and Asia shows that the incumbent can maintain a high market share at premium prices despite competition on its own network, that the market grows all the faster for it and that the incumbent earns revenue from the wholesale in addition,” says Felton.

Most telecom service provider executives will blanch at the thought.

But Felton says that might be the only practical way to boost revenues beyond the 30-percent threshold. Sure, telcos will enable their competitors, but will ensure faster payback on their investments, Felton argues.

In essence, Felton argues, a copper-based business model is profoundly different from a business model based on more-expensive optical access. FTTH is, in a sense, disruptive to the old vertically-integrated model.

Felton says the fundamental choices likely

are opening up the network to wholesale customers, or “accepting that the network is a business for another entity,” says Felton.

That could include either functional separation or structural separation, in some cases going so far as to leave optical access facilities to other entities such as electrical or water utilities or municipalities. Given the expected state of financial markets long term for government entities (unrelated to the current recession), that suggests utilities might be a solution.

If telco executives cannot convince investors that they can achieve payback in three to five years, they generally will not be able to get the investments, or will have to settle for pressure on their share prices.

Aside from the palatability of the wholesale option, much depends on the assumptions one makes about what this does to or for the incumbents retail prospects. Felton says one has to assume the shift to wholesale has no impact on retail market share.

U.S. telco executives might not agree this will be the case. “One might expect that opening the network would diminish the market share of the network owner, but experience in copper unbundling shows that the incumbent always retains a significant—if not majority—market share and that the number of overall customers grows,” says Felton.

In fact there are a number of examples (like BT Openreach) where embracing unbundling resulted in the incumbent’s market share increasing, says Felton.

The argument also hinges on penetration of unbundled competitive offers remaining at about 20 percent market share. Felton also models wholesale ARPU for an unbundled connection at $20 a month. Felton also assumes 60 percent margins on wholesale lines.

In short, the fiber-to-the-home business requires a fundamental rethink of the wireline operator’s mission and organization, Felton argues.

A strategic positioning that might have been successful on the existing copper network might not be as relevant, or even viable, in the context of the deployment of a new and costly network.

Felton argues telcos have to sort through these issues before other contestants decide to move. Felton’s analysis probably is influenced heavily by the experience in European markets.

That said, some possible tactics have too little impact on the payback model. A five-percent increase in profit margin or two-stage build, or decrease in cost of capital to 10 percent do not affect the model fast enough.

The three variables that can lower payback to less than nine years include slicing the cost of the network to $750 or less; higher penetration rates of 40 percent or more; or opening up the network to wholesale customers.

In a U.S. context, in the near term, only the lower capex or higher penetration scenarios are likely to make sense. U.S. incumbents are making open access moves, but carefully and related more to devices or applications than physical access, which continues to be seen as a strategic asset.

Still, Felton’s argument might someday make more sense if major U.S. providers are unable to create new services at a healthy clip or cannot boost penetration beyond 30 percent.

And the wholesale argument isn’t getting any more traction in the cable TV segment of the market, either. Both telcos and cable companies seem to agree that the scarcity of wired broadband access assets makes network ownership quite valuable.

Wireless providers arguably have a slightly-different perspective on wholesale access, but even there some clear steps have been taken to reduce the attractiveness of wholesale offers, typically by restricting the availability of the latest handsets.

Felton’s argument seems to have greater resonance in European and some other markets than in the United States. But it is provocative. And U.S. competitors would like nothing better than robust optical access wholesale. IP