AFTEL'S PUZZLING EXPANSION PLANS

UNCLAS SECTION 01 OF 03 KABUL 000926 SIPDIS SENSITIVE SIPDIS DEPT FOR SCA/FO, SCA/RA, AND SCA/A DEPT FOR EB/CIP NSC FOR AHARRIMAN CENTCOM FOR CG CFC-A OSD FOR BRZEZINSKI E.O. 12958 N/A TAGS: ECPS EFIN EINV ECON KPRV PGOV AF SUBJECT: AFTEL'S PUZZLING EXPANSION PLANS -------------- SUMMARY -------------- ¶1. (SBU) State-owned Afghan Telecom plans to quadruple the number of lines the company has by an aggressive build-out of wireless and copper networks, funded by the GoA and donors. The wireless plans will put AfTel into competition with mobile service providers, which could bring charges of anti-competitive behavior because of AfTel's government funding. The copper network plans appear to be over-sized for the market of business customers for digital services. The expansion is driven by a political imperative to extend service to more of the population and a desire to add quick value for privatization. To duplicate Afghanistan's remarkable success in wireless telecom, the GoA would be better served focusing on a quick sale to get the state player out of the way of private investment. The Embassy intends to start a dialogue with the GoA on its approach to privatizing AfTel. End summary. -------------- SUCCESS WHERE GOVERNMENT ISN'T -------------- ¶2. (SBU) Because it has attracted over $700 million in foreign direct investment, the telecommunications sector has been Afghanistan's biggest economic success story: it has grown from 85,000 total subscribers (including 35,000 satellite phones) in 2002 to over 2.4 million now and has expanded coverage from a handful of cities to all the provinces in the country. Nearly all of this progress has happened on the wireless side, though, and it has happened as a result of the GoA getting out of the way. The same has not happened on the fixed-line side, in part because a phlegmatic state-owned enterprise still has a monopoly and responds more to political direction than to market forces. -------------- DRAMATIC PLANS--DO THEY MAKE SENSE? -------------- ¶3. (SBU) In a recent meeting at Afghan Telecom, senior planning and technical staff briefed Econoff and Afghanistan Reconstruction Group (ARG) advisor on the company's plans for expanding its network. These plans are, to say the least, ambitious: 150,000 new copper lines, completion of facilities for 123,000 new CDMA (wireless) lines, and an optical fiber ring around the country's ring road. When completed, these plans would roughly quadruple the company's capacity. The total cost of the copper and CDMA programs will be $65 million, $17 million of which has been spent (mostly on CDMA equipment). The CDMA project is partly funded by India ($13 million); the $40 million copper project is to be partly funded by the GoA ($30 million),with the balance expected to be provided by other donors. The company has recently awarded a contract for the partly GoA-funded $65 million optical fiber cable ring around Afghanistan to Chinese equipment provider ZTE. ¶4. (SBU) The decision to expand the AfTel network so dramatically is somewhat puzzling from a business standpoint, given that AfTel's largest problem at the moment is not the size of its network but its ability to collect from its customers. AfTel described its collection rate as about 30 percent, and it is uncertain how large its subscriber base is. (NOTE: An advisor to AfTel has said 30 percent is a wild guess--probably on the high side--and that the subscriber base might be something on the order of 70,000. End note.) The company recently installed a new Afghanistan Reconstruction Trust Fund (ARTF)-financed billing system for post-paid customers (currently all AfTel customers are post-paid), which should help with collections once it is operational. The company's recently launched CDMA service is running on a post-paid basis--an approach no other mobile provider dares to take in a country with no street addresses, let alone credit ratings. AfTel plans to roll out pre-paid billing later this year and to migrate all of its mobile customers and most of its fixed-line customers to that system. -------------- KABUL 00000926 002 OF 003 CDMA TO COMPETE WITH GSM -------------- ¶5. (SBU) The CDMA system itself is something of a puzzle. AfTel had been describing this as a build-out of digital fixed-line service (wireless local loop),but has recently begun to describe it as including "full mobility." Because AfTel enjoys the country's only unified service license, it can offer any service it pleases, including fully mobile telephony. (The regulatory authority made a specific decision in 2004 to allow full mobility service under AfTel's license.) But AfTel is receiving GoA and donor funding for this and other projects, so it will be asking for trouble from the GSM operators if it does begin to offer a directly competitive service in areas covered by the GSM operators. At the very least, this potentially anti-competitive behavior on the part of the GoA could dampen enthusiasm for further outside investment in telecommunications. ¶6. (SBU) Meanwhile, the Afghanistan Telecommunications Regulatory Authority (ATRA) has granted a number of district-level CDMA licenses for local fixed service providers (LFSPs) to provide wireless local loop service. The first of these LFSPs is scheduled to launch service this month in Mazar-i-Sharif, followed by a roll-out in adjacent districts. While these licenses are intended to accelerate service roll-out to previously under-served rural markets, in practice they may lay the foundation for real competition to AfTel's fixed-line service in provincial population centers. -------------- COPPER FOR WHOM? -------------- ¶7. (SBU) The expansion of AfTel's copper network makes less than perfect sense from a business perspective, too. AfTel describes this part of its expansion as a move to deliver digital services to its current and prospective business customers. But demand for 150,000 digital business lines is so far in the future that industry insiders say this aspect of the expansion makes no business sense now. That said, another consideration in the deployment of more copper is the scarcity of electrical power in Afghanistan, which makes wireless local loop far more problematic, since electrical generators have to be installed at each base station, and some source of power (usually battery) has to be available with each subscriber. Installing copper lines is a way of avoiding the electricity issue, since land-line technology does not require an outside power source. -------------- COMMENT: A CERTAIN POLITICAL LOGIC -------------- ¶8. (SBU) There is a certain logic to both the copper and the CDMA expansions from the Ministry of Communications' perspective. First, there is a political imperative to bring public utilities to more of Afghanistan's population. At this critical juncture in Afghanistan's history, getting service out to the provinces is a legitimate priority. On paper, a quick expansion looks like a good solution, if one ignores the economic sustainability of AfTel. AfTel has been somewhat corporatized, but in reality it is still part of the MoC. The expansion is essentially a command performance by a state player that may believe it is still operating in a command economy. The key question now is whether the expansion will actually address the needs of the underserved portion of the populace. To the extent that the build-out does that, it will avoid direct competition with the private-sector operators. But it is difficult to judge this from the plans alone; the answer will become clear as the build-out is actually implemented. ¶9. (SBU) Second, the MoC has been promising to privatize AfTel this year. (NOTE: It has been promising, but until very recently not planning, to privatize. This year's draft strategic plan initially alluded to a privatization, but set no timelines and did not include privatization within its lists of goals. In response to Embassy's KABUL 00000926 003 OF 003 input, the plan was changed to include a target date of end 2008 for privatization. End note.) With a small and flaccid subscriber base, paltry revenues, and very modest technical progress to report in recent years, AfTel is an unappetizing privatization offer. In keeping with the usual course of state telecom privatizations, the GoA appears to want to pump some quick value into AfTel to ready it for the market. Thus there may also be some logic to expanding the network before fixing a severe collections problem, and to entering mobile competition against three (soon to be four) larger, more established, and generally better funded GSM providers. -------------- COMMENT, CONT'D.: GET AFTEL OUT OF THE WAY -------------- ¶10. (SBU) At least one knowledgeable insider describes AfTel leadership as undecided about when and how to privatize. As is generally true for privatizations, there is disagreement among technical assistance advisors as well. Some believe a quick sale will make room for private investment to happen sooner, driving expansion of services and tax revenues faster than the state enterprise will ever do. Others want to maximize the value of the sale to ensure serious and committed investors are running things. This takes more time and investment from the state. Both courses hold their own hazards, but inertia dictates the easiest and politically most palatable course: continue to run AfTel as a state enterprise, try to maximize social benefit, and make the privatization decision later, when things might be better. Given the sad record of much more capable and better-funded governments at turning around faltering state enterprises, this easier course is not likely to end well. Thus the Embassy intends to begin a substantive dialogue with the MoC on its approach to privatizing AfTel. End comment. Neumann