In an aggressive step, the South African parliament’s Portfolio Committee on Communications has announced its intention to push for a major cut in South African mobile termination rates. South African mobile networks currently pay each other ZAR1.25 ($0.17) per peak time minute for terminating calls on their networks. The Committee has indicated that it would like telcos to reduce this rate to ZAR 0.60 by November this year, and by a further ZAR0.15 annually over the next three years. [Click title to read full analysis]

 The net impact of a sharp MTR cut on operators will be materially negative in our view; from a consumer and regulatory standpoint, however, this move may be too little too late.

 - South African MTRs are among the highest in Africa. They are at the same level as rates in such markets as Gabon and Congo-Brazzaville (hardly examples of progressive economies) and 2x-3x higher than at least a dozen other African markets. Regulators in other countries have been more aggressive in cutting MTRs; to say that South African MTRs are due for a downward correction is a hard form of euphemism.

- From a regulatory standpoint, and welcome though this measure would be, there are reasons to be skeptical of its medium term impact on headline retail prices. For one, South Africa is a mature market, with mobile penetration rates around 110%. Competition has settled into a comfortable triopoly, while most other African markets now have four to six operators. The optimal time to make this move would have been prior to the licensing of a third operator in 2001. South Africa’s high prices are the direct result of its policy approximations over the past 15 years, which have created a Gordian knot that will not be easily undone.

- MTRs are merely the tip of the iceberg underneath South Africa’s mobile cost structure. Peak time on-net prepaid rates are 70% higher than the African median. The differential between peak on-net and off-net airtime prices is less than 20%; in other words, on-net calls (which do not include any form of termination fee) are nearly as high as off-net calls. A cut in termination fees would not in itself impact on-net calls, which account for more than half of all voice traffic for the main operators.

- There are other issues; according to the AfricaNext Research report The Future of African Mobile Profitability, South Africa’s network costs are high, at about $4,500-$5,000 per site per month, higher than at least a third of African markets, though increased self-provisioning should start alleviating the burden of this item. Average per employee costs are 2x-3x higher than in most other African markets.

- Strategically, we expect MTR cuts to impact the South African market in a number of ways:

      -Push operators to offer better on-net deals, so as to increase the amount of traffic they control

      -Increase the differentials in prices between on-net and off-net calls

      -Increase overall average usage

 Slashing termination rates may not make much of a difference on a stand-alone basis, but should provide enough of a threat that operators will move to cut on net calls without additional regulatory prodding; moderate on-net price cuts would provide some arguments against deep MTR cuts, with increased usage making up for the decline in revenue.

 -We expect the overall impact of an MTR cut to be a net positive for mobile operator Cell C and fixed operator Telkom, both net interconnect payers. For MTN and Vodacom, the proposed interconnect cut would have a material impact if implemented within the next three to six months. Net interconnect margins typically account for 15%-20% of EBITDA; a 50% cut in termination rates, for example, would slash 3-4 percentage points in EBITDA margin, assuming no change in prices and traffic patterns. As a result, the optimal approach for the operators would be three-pronged:

(1) slow the process somewhat until alternative sources of revenue and cost-savings processes make MTR cuts inconsequential

(2) move in front of the public relations issue, ideally by slashing on-net prices before MTRs are cut

(3) accelerate network self-provisioning.

In the West African nation of Togo, the government and second mobile operator Atlantique Telecom Togo (ATT, a unit of Etisalat) are locked in a high-stake, “pay or else” poker game reminiscent of Benin’s negotiation with its mobile operators two years ago. As in Benin (and more recently, in Senegal), the Government of Togo (GoT) has suspended the operations of the operator, subject to full payment of the license fee under the government’s terms.  

- The GoT is asking for $45m (XOF 20bn); at a top level, the amount is not outlandish. As the GoT has noted, it is consistent with recent license valuations in Benin and Cote-d’Ivoire. The problem with the Togolese license is that it is stripped down to near-nothingness. The proposed license does not include an international gateway (a must for profitability) or direct interconnect (it would require ATT to connect to other mobile operators through state-owned Togo Telecom); it is not technology neutral (ATT would need to buy more licenses to offer broadband services), and includes no tax holidays. Without those elements (which were included in other markets), Togo’s license is worth only half as much as is requested in a best case scenario, for an operation that by our estimates, will see 1.5m-2m subscriptions at best and low ARPUs over the long term.

- Another sticky point is the fact that the GoT requires the entire license fee to be settled by the end of 2009; the general practice in the region is for the operator to pay for a portion (typically 50%) of the license at renewal, with the balance settled over a 2-3 year period, so as to avoid the impact on the balance sheet of the one-time capital outlay. The government has already allocated the money in its budget, and Ministers are now under pressure to deliver it.

- According to Togolese Press Agency Savoir News, Etisalat has made a counterproposal  including alternate offers ranging from $20m to $45m, with variations based on the features of the license and the possible licensing of  a third mobile operator. The payment schedule is bound to remain a bone of contention, but Etisalat is handling the issue just right in our view. We expect the parties to ultimately reach an agreement, but the entire episode has been rather damaging and wholly unnecessary.

Following initial reports that Middle Eastern Group Zain had put its African operations up for sale, the company has now indicated that it was in preliminary talks with three potential buyers, including two Indian companies. In a previous Investor Note, we posited that Zain had good cause for selling its African operations. With an initial (and in our view, on the high side) valuation of $12bn floated around, the central, ultimate question now centers around the fair value of the Zain Africa business (excluding Sudan and Morocco, which would not be part of the deal). We add our own rather speculative grist to this mill, and find valuing Zain Africa somewhat of a Gordian knot.

Will investors value the African unit on the basis of what it has been, a highly profitable first tier pan-African operation? Or will they put more weight on company performance over the past 18 months, which offers some hints of vulnerability in a market whose catalysts of profitability are rapidly mutating? In the answers to these questions lie the variation between a valuation at $6bn and a valuation closer to $10bn.

 In a new AfricaNext Investor Research Note, we argue that Zain Africa has to be valued conservatively, as an operation with good upside, which faces significant internal and external challenges going forward. We estimate Zain Africa’s enterprise value at about 5x-7x 2008EBITDA, a range of $7bn to about $10bn, including a premium for extensive pan-African coverage. We also say that Zain should either sell the entire Africa unit as a whole, or not at all.

 The full Research Note is available in the Member-Only section of our web site (subscription is easy and free, but required).

At last, it was here. The SEACOM cable landed on East Africa’s Coast over the summer, and has now gone live in Kenya and Tanzania. We summarize some initial observations from our own recent trip to East Africa, press reports and other anecdotal evidence. Our verdict: promising on bandwidth utilization, wait and see on retail prices.

1. While no official estimates have been announced, we estimate that SEACOM has been able to pick up slightly more than a dozen customers since its launch, each purchasing anywhere from an STM-1 (155mbps) to 2.5Gbps.

2. Undeniable though the enthusiasm from ISPs and corporate customers for the new cable may be, it is coupled with some caution. The main ISPs are purchasing enough bandwidth to meet immediate demand and stay competitive, but are moving cautiously as they (1) seek to reduce the cost or get out of long term satellite contracts and (2) wait to see the impact of the arrival of two additional cables on wholesale bandwidth prices.

3. Service providers with stakes in other cable projects (namely the TEAMS and EASSY cables expected in 2010) are nonetheless picking up capacity on SEACOM, lest they be at a competitive disadvantage, a trend that highlights SEACOM’s first-mover advantage.

4. Taking capacity inland and reaching landlocked markets will remain significant challenges; cunningly, SEACOM has been buying up terrestrial cross-border capacity into landlocked markets (e.g. to Rwanda and Uganda) to expand its offering beyond seaport routes. Its main challenger in East Africa will be Kenya Data Networks.

5. The submarine cable impact on wholesale prices is unquestionable. By our estimates, SEACOM has slashed wholesale bandwidth costs by at least 60% in Tanzania (for a dedicated E1); in Kenya, a dedicated international E1 is already going for less than $1,500 (excl. VAT). Kenya Data Networks, one of the country’s largest wholesale players (and a major SEACOM client) announced that it would slash its wholesale rates by 90%. The retail price impact will be slower, though there is some movement here as well. The initial step will be for increased bandwidth for a similar amount of spend; we anticipate a six to nine month lag for a decline in retail prices.  

6. The impact on wholesale bandwidth requirements is already perceptible. In Kenya and Tanzania for example, estimates from the Broadband and Wholesale Forecasts released by the AfricaNext Data Unit project a 150% and near 200% increase in international bandwidth utilization respectively for 2009, and that’s with less than 6 months of submarine cable availability.

Reports emanating from the Middle East suggest that Middle Eastern Group Zain has put its African operations up for sale (excluding Sudan). Kuwaiti newspapers Al Qabas and Al Watan have reported that the process is at a preliminary stage, with Zain receiving substantial interest from a variety of potential investors. Other reports point to discussions with a “French company”, for a deal that could be worth “up to $12bn”.

Consider us mildly surprised by this news.

1. The sale of Zain’s African operations would go against Zain’s oft-stated ambition of becoming a top 10 telecoms company in the world by 2011, with $6bn in EBITDA and a customer base of 110m. Over the past two years, Zain has implemented a number of initiatives to leverage its scale in the African market, from its “One Network” initiative spanning 19 countries to an ongoing operational restructuring dubbed “Drive 2011”. Late last year, Zain also indicated that it remained on the prowl for potential acquisitions or greenfield licenses in the African market.

2. Zain’s African operations are typically well positioned; the company is a market leader or a number 2 in most of its African markets. It has the widest coverage footprint in population terms among all pan-African operators, and is present in markets that, by our estimates, will generate the bulk of incremental cumulative mobile revenues over the next five years. Its subscriber base in sub-Saharan Africa rose 50% in 2008, and revenue rose 30% in 2008. Some operations have performed well (Zambia, Niger, Gabon), while others continue to struggle with the twin combination of economic downturn and intense competition complicating their ability to get better (Kenya, DRC).

3. Ultimately, the fact that Zain would even consider selling its Africa business points to heightened concerns about the deterioration of fundamentals in African mobile markets. Competition has intensified, taxation levels have risen as tax holidays have expired, the markets are as capital-intensive as ever (at a time Zain is seeking to cut CapEx levels by half) and African currencies have plunged against the dollar. By contrast, Zain’s Middle Eastern operations have become more profitable, but require more capital.

4. Excluding Sudan, Zain’s African operations accounted for about 65% of the group’s subscriber base, 56% of its revenue and 50% of its EBITDA in 2008. Perhaps more significantly, they take up more than 75% of capital expenditures, yet only account for 15% of the group’s net income. Zain’s net income rose 6% in 2008; excluding Africa, net income rose 34%. For all the lofty subscriber numbers, African operations are arguably a drag on the entire group, at least for now.

5. Selling may be strategically questionable over the long term; some operations (e.g. Nigeria) are bound to perform better. In the short term, however, and assuming a good price (and $12bn would be good), selling would help pay high short term debt and maximizes returns by re-allocating capital spend to higher-profit Middle Eastern and North African operations. Questionable though this move may be in some respects, we would certainly understand it.