30 January 2010

Firstly, some facts. ADSL comprises 63%[1][2] of broadband connections globally. In SA, ADSL comprises less than 40% of the broadband market[3], and until recently, an email being sent via fixed-line was in danger of being overtaken by continental drift! Secondly, 6 different network technologies (iBurst Wireless, IPWireless, UMTS (3G/HSDPA), Wi-fi, Wimax and CDMA2000) cover all the major towns and cities resulting in one of the widest choices of mobile/nomadic connectivity in the world.

Another lesser known fact is that, in spite of high retail prices in relation to our peer countries, wireless broadband is very competitively priced. I know the margins for a pure wireless broadband network, and believe me, they’re as thin as Kate Moss in vertical stripes… Vodacom, MTN, iBurst, Sentech and Neotel are constantly having a ding-dong at each other to gain advantage in the marketplace, and price has become the only real differentiator!

So what’s the problem? The problem is that Telkom has a de facto monopoly on fixed lines under South African soil, one of the key cost components for any wireless broadband network. When iBurst builds a network it must purchase diginet lines from Telkom in order to connect its base stations to the Network Operations Centre (in Woodmead). The cost of linking a base station from CT to Joburg is the same as connecting CT to London! No one has an installed cable infrastructure that matches Telkom’s footprint, so all the wireless operators are forced to use Telkom for their short- and long-haul transmission.

And international connectivity? Up until the recent arrival of SEACOM and the new telecommunications licenses, there was only 1 legal way to send data from SA to the rest of the world. SAT3, the undersea cable part-owned by Telkom. When I joined iBurst in 2006 the average price of 1Meg of international bandwidth was R36,000pm. Today it is less than R10,000pm, a drop of 72% in 3 years but still crazily expensive when you consider that the price for an equivalent distance from Europe is R700pm.

Wireless broadband is expensive because the biggest variable expenses (local and international connectivity) are controlled by Telkom and are exorbitantly priced. The wireless operators have to pass on this cost to consumers. Bring down the cost of local and international links and the cost of broadband will drop across the board.

What is the local loop? Local loop refers to the copper lines that already run into households (originally installed for your landline). A major reason for fixed lines comprising 63% of the global broadband market is that the respective fixed-line operators already have lines into most (previously advantaged) households in the country (installed over the past 60 years using taxpayer money.)

One of the most effective ways to improve broadband in any country is to allow other companies to use these existing lines. To this end the regulator forces the incumbent (ie: Telkom) to share these lines at a fair price (also determined by the regulator using cost as a guideline), thereby giving competitors the opportunity to sell ADSL services without having to dig up all the streets and lawns in Joburg!

The root of our problems is that Telkom’s primary objective is the maximization of profits. Without an effective regulator they have been able run amok. If Telkom were still 100% government-owned you would not need to regulate it. The government could simply instruct Telkom to share its infrastructure (local/international connectivity and local loop) at a fair price.

But the fact that Telkom is a publicly listed company (private investors own roughly 50%) means that management has a duty to maximize shareholder value. That means charging as much as possible for the use of assets (ie: cables in the ground) over which it has a monopoly for as long as possible. So you can’t really complain if Telkom’s management is doing its best to maintain the monopoly and keep prices high. Their bonuses and promotions are based upon Telkom profits, not upon the percentage of South African’s that have broadband.

That’s why we need a strong, independent, well-funded regulator. Of course, it’s easy to attack ICASA from the sidelines. Sadly, and mostly through no fault of its own, ICASA has been hounded over the past decade with interference and fiscal constraints. It is difficult doing your job when your boss is constantly undermining you, and then not giving you the budget you need to fight companies the size of Telkom! A great example occurred in 1996 when the government granted a guaranteed 5 year monopoly to Telkom in order to entice the Thintana Consortium to buy a 30% stake. It’s like being a policeman tasked with arresting thieves, but the judge issuing an order giving the thieves lifelong immunity!

Was it worth it? The government raised R4.3billion from the listing and it is estimated that the cost to the economy of extending the monopoly was R67billion…

Has ICASA been effective at reigning in Telkom? Let’s compare countries: Telkom makes a net profit margin of 24%[4], compared to a net profit margin of 2,4% for Deutsche Telekom[5]. That means Telkom is 10 times more profitable that Deutsche Telekom. Yes, 10 times more profitable.