Niagara Networks - a New Mobile Player in the Canadian Market?

CBCNews.ca reports that a mysterious new company, Niagara Networks Inc, has entered the list of prospective bidders for the Advanced Cellular Service (AWS) spectrum license auction scheduled to commence end of May 2008.(CBCNew Canada on AWS and Niagara)

Niagara Networks has indicated that it would seek 6,510 bid points. This amount would be sufficient to launch a new country-wide mobile network in Canada. Niagara's ambition level required the company to make a credit deposit of about 880 million Canadian Dollars to the Industry Canada (i.e., Canadian Ministry of Industry).


The Canadian AWS Framework (Industry Canada AWS guidelines)

The Advanced Wireless Services (AWS) frequency band is primarily intended for UMTS-based (or IMT-2000) services and is similar to the AWS band auctioned two years ago in the USA.

The AWS spectrum being auctioned is in 3 blocks of 2x10 MHz as well as in 3 blocks of 2x5 MHz. The main frequency band is 1,710 - 1,755 MHz (uplink: User Terminal to Base Station) and 2,110 MHz - 2,155 MHz (downlink: Base Station to User Terminal). A total of 2x45 MHz (90 MHz) will be auctioned intended for Frequency Division Duplex operation (i.e., FDD). The Industry Canada has set aside a minium of 2x20 MHz for a Greenfield mobile operator.

The definition of a new entrant is taken to be an entity with less than 10% share of the Canadian mobile market. Thus, in this respect the local Canadian mobile operator MTS would qualify as a new entrant.

The minimum bid price for a country wire AWS license would be C$5 million per MHz or an average (based on minimum bid) C$ per MHz per Pop between 0.11 and 0.15 C$/MHz/Pop depending on the type of service area.

In Europe it is in general accepted that in most major markets (i.e., Germany, UK, France, Italy, etc..) 2x10 MHz would not provide sufficient UMTS radio capacity for the expected huge uptake in data services. As a minimum a mobile operator, in a major market (such as Canada), should target 2x20 MHz. Thereby reducing the risk of becoming growth limited.

In context of acquiring 2x20 MHz, it is worth noting that such bandwidth would be in line with the Next-Generation Mobile Network minimum guidance for bandwidth needed to provide peak-performance (i.e,. 100 Mbps @ 20MHz).

Industry Canada is currently also developing a release plan for the 700 MHz spectrum (i.e., digital dividend) currently being used for analog TV transmission over-the-air. It is not foreseen that significant spectrum will be released prior to mid-2011. The regulator will for the 2.50 GHz - 2.69 GHz frequency band (mainly in the hands of incumbents) demand introduction of mobile services in order for current licence holders to keep their spectrum position. Industry Canada is formulating a policy for un-assigned spectrum in the 2.5 GHz band. In addition to the 700 MHz and 2.5 GHz frequency ranges, there are several other bands below the 1,700 MHz which are being studied for possible release.

Canadian Mobile Market

From a spectrum portfolio perspective the Canadian market would not be an easy entrance for a greenfield company. Most of the existing mobile spectrum is concentrated around the the Big 3; Rogers Wireless, Bell Canadian Enterprises (BCE) and Telus. For the fourth player MTS, the existing spectrum position is relative weak. This situation would not change substantially after the AWS auction as the 3 spectrum-fat players still would have a superiour spectrum position.

The Canadian mobile market is different from the European one. Typical Western European Mobile penetration is in the neighborhood of 110%, while in Canada only around 60%. Furthermore, post-paid customer share (ca. 77%) in Canada is much higher than Europe (ca. 45%). In addition the blended ARPU in Canada (ca. $58) is likewise better than in Europe ($40). The amount of mobile minutes of use (MoU) per user per month generated on average in Canada (385 MoU) also significantly higher than what is observed in Western Europe(142 MoU). The Canadian market is timing-wise behind when it comes to non-voice service (incl SMS) revenue share out of the total revenue. In Canada this was in 2006 only 8% compared to the Western European 16% share both including messaging.

Greenfield Canada Inc - The Challenge

Deploying mobile networks are not cheap and in a mature market as Canada possible not the easiest venture. To provide coverage (to European GSM levels) across Canada and for the whole population would be very expensive above 900 MHz. If the regulatory requirements for license compliance only stipulates population coverage, an operator can get very far with covering the top-10 to top-20 cities. This said if competitors have country-wide coverage it could possible be a major customer dissatisfaction if the same would not be available to them. This issue can be solved by National Roaming wholesale deal(s) with a competitor(s) having country-wide coverage. Although it does lead to additional whole-sale operational expenses, it might nevertheless be more attractive than rolling out a network to the same level as existing incumbent.

Let's look at new entrant Greenfield Inc business model and challenges such a company may face;

Inspecting the framework for the AWS auction it can be estimated that the minimum population coverage condition would be between ca. 47%. This is roughly equivalent to covering the top-20 Cities in Canada (with some extension to metropolitan areas). The needed service areas would not amount to much more than one thousands of the whole Canadian surface area or 12,000 square kilometers.

Our Greenfield Inc would at 1.7 GHz require between 2,000 and 3,500 radio nodes to cover the above service areas. Thus covering ca. 47% of population. The final number of nodes would depend on the targeted quality of coverage, of which 5 years deployment time is provided in order to fulfill the regulatory minimum population coverage criteria.

The capital investment level needed to rollout to the regulatory requirements would be between $400 to $600 million cumulated (5 yr period); after which the Greenfield Inc most likely would capital invest a maximum of 5% annually of their service revenue.

WARNING - The following analysis might offend the Financially inclined due to shockingly simple estimates and intentionally overlooking complexities of which consultants can earn lots of money on making appear even more complicated.

Assuming that Greenfield Inc becomes successful in the Canadian market and after the first 5 years have achieved a stable market position of 20% customer market share, i.e., 2.7 million subscribers. With an ARPU of $50 per month (2007 level was $58) one would expect annual revenues of $1.9 billion; which results in an annual Capital expense (Capex) level of ca. $100 million (i.e., 5% of revenue). This Capex level would be sufficient to support customer demand, traffic increases (depending on acquired spectrum bandwidth) and infra-structure obsolescence management (note this is only valid if city/metropolitan based coverage strategy is followed).

Our Greenfield Inc takes care of its operational expenses (Opex) and after the initial 5 years of network deployment can keep the EBITDA margin stable at 30% (i.e., EBITDA $570 million). Thus, Opex should not be higher than $1.1 billion.

Depreciation & Amortisation (D&A) level is kept at a level of $100 million annually (steady-state). Furthermore, Greenfield Inc has an effective tax rate of 35% (i.e., among the developed economies Canada has one of the highest corporate tax rates).

Now we can actually estimate the free cash flow (FCF) Greenfield Inc would generate from the 6th year forward:

Revenue $1.90
-Opex $1.33
=EBITDA $0.57
- D&A $0.10 (ignoring spectrum amortization)
- Tax $0.18
+ D&A $0.10
=Gross Cash Flow $0.41
-Capex $0.10
=FCF $0.31 billions

assuming zero percent FCF growth rate and operating with a 10% Weighted Average Cost of Capital (i.e., WACC) the perpetuity value from year 6 onwards would be $3.1 billion. In Present Value this is $1.9 billion, net $0.4 billion for the initial 5 years discounted capital investment (for network deployment) and considering the first 5 years cumulated discounted EBITDA ($0.3 to $0.5) billion provides

a relative strong case of $1.8 to $2.0 billion valuation prior to spectrum investment where-of bulk valuation arises from the continuation value (i.e., 6 year onwards).

If our hypothetical Greenfield Inc was careful they would most likely not like to pay much more than $1.0 billion for AWS spectrum. In particular as the valuation is dominated by the perpetuity.

As is often the case (in finance) the value can be made more impressive by choosing lower WACC, assuming a non-zero FCF growth rate, higher margin, improved operational efficiencies, higher ARPU, more customers, etc...

Greenfield Inc would have a good positive business outlook as a new entrant to the Canadian market.

Niagara Networks Inc - Positive Outlook

Niagara has been reported to provide a credit deposit in the order of $880 million with Industry Canada (i.e., Ministry of Industry). This might be seen as an indication of the value Niagara has assessed for a Greenfield operation in Canada (note: one may hope that the value is significantly higher than what they intend to spend on spectrum).

That Niagara appears to aggressively pursue the AWS spectrum is not really surprising. It is in line with the above estimated positive outlook for Greenfield Inc which ended with a valuation estimate of ca. $1.8 billion.

Niagara should target a spectrum acquisition of at least 2x20 MHz as it would ensure them both sufficient growth possibilities (compared to competitors) as well future proofness moving to a Next-Generation Mobile Network (NGMN) infrastructure somewhere in the next 5 to 10 years.

The expected release of the 700 MHz frequency band (expected full scale after 2011) could reduce the value of the Niagara business / spectrum acquisition. The network economics for network deployed at 700 MHz is significantly better than at 1.7 GHz as the operational cost for the 1.7 GHz network will be much higher than 700 MHz (comparable coverage). Furthermore, the 700 MHz network would highly cost-efficient be able to reach a much larger proportion of the population and surface area than would be financially viable with 1.7 GHz.