Is Safaricom Dominant? A Concise Perspective On Analysys Mason's Controversial Kenya Telecommunication Competition Report

WARNING: Long blog post alert!

Background

Approximately one month ago Analysys Mason released a controversial (draft) telecommunication competition market study report as commissioned by the Communications Authority. The report was leaked in late February 2017 and the media, both local and international, were quick to point out that the report clearly states that Safaricom, Kenya’s largest mobile network, is indeed dominant.

The draft report posits that in the best interests of Kenya’s telecommunications sector Safaricom should be broken up and services like M-Pesa be hived off as a separate independent entity. In addition, the report suggests that Safaricom be compelled to share its network and towers with competitors to level the playing field in areas such as 4G service provision. In a nutshell, the Analysys Mason report stirred a hornets nest with various prominent voices for and/or against many of its findings and recommendations.

As it turned out this past week, I received an unsolicited copy of the leaked Analysys Mason report which runs 119 pages long – it’s a very long read! From this perspective, I hoped to take a very pragmatic and objective look at the report without choosing sides per se so as to glean the key insights on the same. One thing to note is that the recommendations in the report are NOT cast in stone and the Communications Authority is not obligated to follow through on them.

In this lengthy blog post, the idea is to fully understand the main report’s findings and its recommendations from an impartial perspective. Saying it as it is so to speak? Here is what I was able to glean from the report, packaged as concisely and simply as possible, even though its a very long read.

Key Highlights

Before taking a deep dive into this blog post, the following are some of the key highlights summarized from the report:

• In the area of towers the report finds that, Safaricom is clearly dominant and that it is not economically viable for smaller operators to extend coverage in areas of low population density given current market share and cost of tower sharing.
• The report recommends it should be required to provide other operators access to its sites in 14 designated counties where there is the largest disparity between the number of Safaricom sites and the number of sites deployed by the other mobile operators in Kenya.
• The market analysis report concludes that Safaricom is dominant in both the mobile communications and mobile money markets due to its very high market share of over 80% in both instances.
• Safaricom controls more than half of the services that are rendered in the retail mobile communications market in Kenya regardless of how this is measured - it has held this position for many years.
• Around 95% of Safaricom’s voice traffic is on-net, while the other operators have 60–80% of voice traffic on-net.
• The report, using data-driven insights, suggests that Telkom Kenya and Airtel Kenya are amongst the least profitable mobile operators in emerging markets anywhere in the world.
• It costs between two and four times as much to transfer a sum of money from M-Pesa to a user of another mobile money platform as it does to transfer the same sum of money to a registered M-Pesa user.
• Safaricom has nearly six times as many agents as any other mobile money platform.
• Safaricom’s M-Pesa agents account for 68% of all mobile money agents in Kenya.
• Safaricom controls more than half of the services that are rendered in the retail mobile money market in Kenya regardless of how this is measured.
• Safaricom’s share of mobile money transactions by volume (82% in 2Q 2016, down from 91% in 1Q 2015) is significantly higher than its share of mobile money subscribers.
• Safaricom’s share of mobile money transactions by value has exceeded 80%.

1. The Market Review Methodology

In order to compile the report, a high-level overview of the current conditions and prevailing trends in Kenya’s telecoms market was required. This enabled Analysys Mason to understand the positioning of the key market players, main trends in the markets as well as how Kenya’s telecoms market compares with its neighboring countries, before starting a rigorous market review.

The market review in Kenya was based on the European Commission (EC) framework (or European framework) for market review, which is often recognized as an example of global best practice. The underpinning logic by Analysis Mason for this approach is that the framework is consistent with Kenyan law and regulation and that it forms a very good reference model for performing market reviews.

1.2 The Three-Criteria Test

To determine whether a particular market is susceptible to forward-looking regulation (referred to as ex-ante regulation) the so-called ‘three-criteria test’ is used in Europe (and elsewhere). The report used the following three criteria to determine if the market in Kenya is relevant for future regulation:

1.2.1 The presence of high and non-transitory barriers to entry

These may be of a structural, legal or regulatory nature. An obvious example in mobile communications might be where the number of infrastructure-based participants is limited by the availability of suitable harmonised spectrum and by the need to establish an extensive network of base stations to offer services.

1.2.2 A market structure which does not tend towards effective competition within the relevant time horizon

This criterion takes account of the dynamic character and functioning of electronic communications markets and thus the possibility of overcoming barriers to entry in the short term, e.g. through technical innovation. A period of three years is often considered to be a relevant time horizon.

1.2.3 The application of competition law alone would not adequately address the market failure(s) concerned.

Competition law interventions are likely to be insufficient where, for instance, the compliance requirements of an intervention to redress persistent market failure(s) are extensive or where frequent and/or timely intervention is critical.

1.3 What Defines Market Dominance?

In the context of the Analysys Mason Report, market dominance is defined as the ability of a company or a group of companies to maintain the prices of its or their products and/or services above a competitive level. Market power arises in a number of industries as a result of factors such as economies of scale that act as barriers to entry and thereby reduce the number of efficient market players, or the use of switching costs that limit the ability of customers to move to new suppliers. Market power is deemed to exist where a company, “either individually or jointly with others enjoys a position of economic strength affording it power to behave to an appreciable extent independently of competitors, customers and ultimately consumers.”

1.4. How Are Remedies Arrived At In The Report?

According to the report, if market analysis establishes that a market is not effectively competitive, then the last phase of the work involves developing remedies to promote competition in that market by addressing the identified market issues/failures. In selecting appropriate remedies, it is necessary to:

• Consider the existing regulations, in order to understand the transition and the burden that the proposed remedies will have on the operator
• Select remedies that are justified in terms of the specific market conditions, and aimed at addressing the market failures identified by the market analysis
• Choose remedies that are in accordance with the principle of proportionality

2. Review of Wholesale Markets & Proposed Remedies

The report identified eight linked wholesale markets and found that call and SMS termination on mobile networks and call termination on fixed networks, are currently already subject to forward looking regulation in the form of termination rates set by the Communications Authority. In the area of towers the report finds that, Safaricom is clearly dominant and that it is not economically viable for smaller operators to extend coverage in areas of low population density given current market share and cost of tower sharing.

2.1 Wholesale Markets Remedies

Given Safaricom’s dominance in this respect, the report recommends it should be required to provide other operators access to its sites in 14 designated counties where there is the largest disparity between the number of Safaricom sites and the number of sites deployed by the other mobile operators in Kenya.

3. Review of Retail Markets & Proposed Remedies

The report identified five retail markets and found that two of these five markets, mobile communications and mobile money, are susceptible to forward-looking regulation. The market analysis report concludes that Safaricom is dominant in both the mobile communications and mobile money markets due to its very high market share of over 80% in both instances. The report goes further to state that under the current scenario in both of these markets Safaricom is able to behave independently of its competitors. In terms of remedies based on these findings, the report recommends the following actions in the future to address Safaricom’s dominance:

3.1 Mobile Communications Remedies

3.1.1 National Roaming

For a period of five years, Safaricom should provide 2G, 3G and 4G roaming on its network to other mobile operators in 14 counties identified in the report which are the same as those for regulated tower sharing.

3.1.2 Replicability of retail tariffs

Safaricom’s standard tariffs, permanent loyalty schemes and promotions should be capable of being profitably replicated by a reasonably efficient competitor. At least five days before launching a new tariff, loyalty scheme or promotion, Safaricom should provide a justification that the proposals can be replicated by a reasonably efficient operator, for which the key parameters (market share, cost structure etc) will be defined by the Communications Authority

3.1.3 Prohibition of on-net discounts

Safaricom should not be permitted to charge different rates for on-net and off-net calls or messaging to any customers under any circumstances (i.e. through its standard tariffs, promotions or permanent loyalty schemes). This includes a requirement that any bonus airtime granted to Safaricom customers should be usable for on-net and off-net calls and messaging at the same rates. To ensure that customers are aware of this remedy, all Safaricom advertising marketing materials referring to tariffs, promotions and customer loyalty schemes should make it clear that on-net and off-net tariffs are the same and that bonus airtime may be used for on-net and off-net calls and messaging

3.1.4. Prohibition on individually tailored loyalty schemes and promotions.

Safaricom may not offer loyalty bonuses or promotions for which the qualification criteria require different levels of expenditure or usage by different subscribers in the same category. This would be discriminatory.

3.2 Mobile Money Remedies

3.2.1 Prohibition on surcharges for cross-platform money transfers

Safaricom should apply the same fee structure and fee level on transfers to registered and unregistered users, including users of other platforms. The fee for cross-platform transfers shall include the fee for cash withdrawal at the Safaricom agent (as it does now) until such time as full wallet-to-wallet interoperability is available.

3.2.2 Mandatory wallet-to-wallet interoperability and potential functional separation

The report suggests Safaricom should work in good faith with the other mobile money licensees, the Communications Authority, the Competition Authority of Kenya and the Central Bank of Kenya to implement a system to enable automatic and near-instantaneous transfers to and from users of other mobile money platforms by 31 December 2017, ending the current practice of effecting cross-platform transfers via cash withdrawal from an agent of the sender’s network and cash deposit at an agent of the recipient’s network. In the event that wallet-to-wallet interoperability is not achieved by 31 December 2017 due to factors that the CA considers Safaricom could reasonably be expected to control, the report proposes that the Communications Authority should move to impose functional separation between Safaricom and M-Pesa.

4. Safaricom’s Mobile Communications Market Share Assessment

Safaricom’s market share in retail mobile communications is as follows according to the report:

• Mobile subscriptions: 65.2% in 2Q 2016
• National minutes on mobile networks: 77.8% in 2Q 2016
• Voice, data and SMS revenues: 86.7% in 2015 (average for year)

The report goes on to highlight that Safaricom controls more than half of the services that are rendered in the retail mobile communications market in Kenya regardless of how this is measured - it has held this position for many years.

There is therefore a strong presumption of dominance by Safaricom in the retail mobile communications market, in line with the Kenyan regulatory framework and European best practice.

4.1 Other Factors That Make Safaricom Dominant In Mobile Communications

4.1.1 Economies of scope and scale and lower resulting unit costs

According to the report, Safaricom has deployed an extensive network on which it enjoys very strong economies of scale (due to this very high market share) and economics of scope in the provision of mobile communication services.

Around 95% of Safaricom’s voice traffic is on-net, while the other operators have 60–80% of voice traffic on-net. Safaricom’s average interconnect out-payment per minute is therefore much lower than that of Airtel and Telkom Kenya. Most capital expenditure (cost of base stations, backhaul and backbone networks) and many items of operating expenditure including spectrum fees, management, strategy, marketing and network maintenance costs do not scale with subscriber numbers or traffic volumes. Consequently, Safaricom is much more profitable than its competitors and can afford to spend more on above-the-line advertising and below-the-line promotions.

Safaricom reported an underlying EBITDA margin (earnings before interest, taxation, depreciation and amortisation) of 44.6% for the year ending 31 March 2016. This has increased to around 50% (or 46.5% when normalised for a one-off adjustment) for the six months to 30 September 2016. This contrasts starkly with the financial performance of the other two operators.

The report, using data-driven insights, suggests that Telkom Kenya and Airtel Kenya are amongst the least profitable mobile operators in emerging markets anywhere in the world: in fact, the database indicates that the only established operator with a lower EBITDA margin is Entel Peru. The benchmark suggests that EBITDA margin is loosely correlated with market share, but the margins for Airtel and Telkom are well below the trendline.

The report further posits that an argument could be made that the poor financial performance of Airtel and Telkom Kenya is the result of inefficiency. However, this would be negated by the fact that according to Airtel’s annual report for the year to 31 March 2016, the average EBITDA margin across all of the company’s African operations was around 22%. Likewise, Telkom, until recently, was majority-owned by Orange Group which in 2015 reported an average EBITDA margin across its operations in Africa and the Middle East of 34.0%.

It is therefore difficult to imagine that management of both Airtel’s and Orange’s operations in Kenya could be so much worse than across the rest of each group that the EBITDA margin should be percentage points below average in the case of Airtel and percentage points below average in the case of Orange. It seems far more likely that the continuing losses of Airtel and Telkom in the Kenyan market must be attributable in large part to market conditions.

4.1.2 Better Network Coverage

According to the report, Safaricom’s high market share means that it can operate base stations profitably in areas of low population density and/or low ARPU where Airtel and Telkom would make a loss. This means that Safaricom is the only network available in some areas, or the only network with 3G or 4G coverage. Superior coverage means that subscribers with multiple SIMs will also tend to give people their Safaricom number as their primary contact number, thereby perpetuating the low share of on-net traffic for Airtel and Telkom.

4.1.3 Better sales network

The report suggests that Safaricom’s high market share makes it easier for the company to recruit and retain agents since they can expect to handle more transactions than an Airtel or Telkom agent. Prepaid users, who make up 97% of the mobile subscriber base in Kenya, therefore find it more convenient to use Safaricom.

4.1.4 Leveraging M-Pesa

The report states that Safaricom’s dominant position in the retail mobile money market via M-Pesa offers a strong incentive for customers to retain a Safaricom SIM. Better marketing, network coverage and distribution encourage such end users to make Safaricom their primary SIM.

4.1.5 Evidence of anti-competitive behaviour

The report expounds that Safaricom is in a position where it would benefit greatly from the club effect if it charges less for on-net calls than for off-net calls. At present Safaricom is not formally required to charge the same for on-net and off-net calls but the Communication Authority’s policy is only to approve Safaricom tariffs with the same on- and off-net pricing.

This arrangement appears to be working for the most part but there do appear to be exceptions. For example, a story in The Standard newspaper dated 26 May 2016 stated that “Safaricom Ltd has lowered its calling rates within the company network in the rural parts of Kilifi where locals will calls for Kes. 2.00 down from the normal Kes. 4.00 calling rate per minute for 60 days from now to give the people an opportunity to make calls”. Moreover, a proportion of Safaricom customers are believed to be on legacy tariffs which do offer lower rates for on-net calling.

Safaricom’s Stori Ibambe na Storo Bonus scheme is an opt-in loyalty scheme for prepaid subscribers. It offers 100% bonus airtime to those subscribers who achieve a daily usage target. The usage target is set by Safaricom and is to be used by midnight on the day that it is granted. The scheme appears to incentivise users who have more than one SIM to continue using Safaricom in the absence of differential pricing for on-net and off-net calls. The scheme lacks transparency because the target-setting process is vaguely defined (the terms and conditions simply say participants’ targets are “based on their average daily usage”) and the scheme is not subject to any external scrutiny.

5. Safaricom’s Mobile Money Market Share Assessment

Safaricom’s M-Pesa’s latest market share of retail mobile money for 2Q 2016 is as follows:

• Mobile money subscriptions: 65.1%
• Transactions by volume: 81.9%
• Transactions by value: 84.3%

These numbers from the report imply that Safaricom controls more than half of the services that are rendered in the retail mobile money market in Kenya regardless of how this is measured. It has held this position for many years. There is therefore a strong presumption of dominance by Safaricom in the retail mobile money market, in line with the Kenyan regulatory framework and European best practice.

5.1 Other Factors That Make Safaricom Dominant In Mobile Money

5.1.1 Economies of scope and scale and lower resulting unit costs

Safaricom has deployed an extensive mobile network that it uses to provide a wide range of services (including mobile money services) over its entire footprint. Safaricom enjoys very strong economies of scale due to this very high market share. It also enjoys economies of scope in the provision of mobile money services. As stated in our analysis of the retail mobile communications market, this results in a very high and growing EBITDA margin which enables the operator to behave independently of its competitors.

5.1.2 High cost and inconvenience of cross-platform transfers

It costs between two and four times as much to transfer a sum of money from M-Pesa to a user of another mobile money platform as it does to transfer the same sum of money to a registered M-Pesa user. In addition, the lack of wallet-to-wallet interoperability means that a person-to-person transfer across platforms requires the recipient to withdraw the money in cash from an M-Pesa agent and then pay it in at an agent for the recipient’s own platform.

5.1.3 Better agent network

Safaricom’s high market share makes it easier for the company to recruit and retain M-Pesa agents since they can expect to handle more transactions than an Airtel or Telkom agent. Consequently, Safaricom has nearly six times as many agents as any other mobile money platform.

5.1.4 Better merchant network

Safaricom’s high market share also makes it easier for the company to recruit and retain merchants who accept payments via M-Pesa. As a result, over 36,000 merchants accept M-Pesa, far more than the number that accept payments using Airtel Money or Orange Money.

5.1.5 Leveraging of mobile communications

Safaricom’s dominant position in the retail mobile communications market was a key factor in enabling M-Pesa to become such an established platform in the first place. The fact that Safaricom continues to have such a high proportion of subscribers undoubtedly helps the company to recruit and retain M-Pesa agents and merchants.

5.1.6 Evidence of anti-competitive behaviour

In the past, Safaricom has threatened M-Pesa agents with loss of accreditation if they acted as agents for other mobile money providers and has arranged to remove or cover point-of-sale advertising for other providers. Airtel documented a number of examples of these practices in 2015. The report states that based on exchanges with Airtel and Telkom, that the situation has improved since then but the two operators believe that the issue has not been entirely eradicated. When asked for more recent evidence, both companies stated that it is difficult to persuade agents to provide it for fear of action by Safaricom.

The MKA USSD study concluded that Safaricom seems to be engaging in conduct that “appears to constrain competition at several levels”. This includes:

• Raising the costs of bank and non-MNO mobile money services providers through unfairly high USSD charges and price discrimination
• Refusing account-to-account interoperability with other mobile wallets to intensify the network effects from which M-Pesa benefits
• Offering preferential treatment to lending products from CBA and KCB in which it has a financial interest.

6. Is Safaricom Truly Dominant In Mobile Communications? Lets Look At The Data

The report goes into great detail to analyze the competitive landscape in Kenya as far as mobile networks are concerned. The findings show that Kenya’s mobile market has the highest Herfindahl–Hirschman Index (HHI) of any country in the benchmark group, at 5082 out of a maximum of 10 000 at the end of 2015 (The HHI measures of the size of firms in relation to the industry and an indicator of the amount of competition among them). This is around 10% higher than the country with the second-highest HHI (Burundi) and 57–94% higher than the remaining countries in our benchmark. This indicates that the market is heavily concentrated.

The report further states that unsurprisingly, Safaricom also has a significantly higher share of subscribers than the leading operator in the other benchmark countries (67% versus 56% in Burundi and 47% in Rwanda and Ghana). Ghana is the only other country in the benchmark group where the leading operator has maintained or increased its market share since 2012. In addition, the report states that based on data provided by the Communications Authority, there has been little variation in Safaricom’s share of total subscribers over the last five years as the figure has ranged from 63.2% to 68.6%.

According to the report, Safaricom’s share of national minutes as reported by the Communications Authority is higher than its share of subscribers. This indicates that the number of minutes per subscriber is higher for Safaricom than for the other operators. Safaricom’s share of total minutes fell from 88.3% in 3Q 2011 to 68.8% in 2Q 2015, but has since climbed to 77.8%.

Lastly, Safaricom’s share of voice, SMS and data revenue is higher than its share of national minutes (86.7% in 2015, down from 90.1% in 2012). There were some inconsistencies in the data on the average revenue per minute (ARPM) supplied by operators, but it appears as if Safaricom’s ARPM is considerably higher than for the other two operators. It also seems plausible that Safaricom’s data revenue per subscriber is higher than Airtel and Telkom.

Further evidence in the report shows that data supplied by the operators indicates that Safaricom’s coverage by population is much higher than that of Airtel and Telkom, particularly in the case of 3G. Using data concerning the distribution of population in Kenya derived from the 2009 census results at district level suggests that Safaricom’s 2G [and 3G] coverage by area is also much higher than that of Airtel and Telkom Kenya. Coverage maps provided by the Communications Authority also indicate that Safaricom’s coverage by area is much higher than Airtel and Telkom Kenya’s.

The report goes on to finally conclude that Safaricom’s market share is unusually high for a large three-player market such as Kenya, and has remained high despite previous regulatory interventions such as reductions in mobile termination rates, the launch of mobile number portability and the licensing of MVNOs. While Safaricom has better coverage than its competitors, the gap in market share is too large to be explained by coverage or efficiency factors alone. This last statement really nails it for me - Safaricom is clearly exhibiting signs of true dominance!

7. How About Mobile Money? How Dominant Is Safaricom’s M-Pesa?

According to the report, Safaricom’s share of mobile money subscriptions has fallen since 3Q 2015, but the absolute decline in the number of M-Pesa subscribers between 3Q 2015 and 1Q 2016 can be attributed to Safaricom’s decision to change the definition of active mobile money subscribers from those who have used the service in the last 120 days to those who have used it in the last 30 days. In any case, Safaricom still has 65.1% of all mobile money subscriptions in Kenya.

The report further states that Safaricom’s M-Pesa agents account for 68% of all mobile money agents in Kenya, although this percentage has slightly fallen from 73% in 2Q 2014, perhaps because of an intervention in relation to Safaricom’s imposition of exclusivity obligations on its resellers.

However, the report indicates that Safaricom’s share of mobile money transactions by volume (82% in 2Q 2016, down from 91% in 1Q 2015) is significantly higher than its share of mobile money subscribers. This decline is largely a result of the growth in Equitel’s transaction volumes. On average, a Safaricom M-Pesa subscriber makes 6 transactions per month, whereas an Airtel Money subscriber makes 0.6 and an Orange Money subscriber makes 0.1. However, the average Equitel subscriber makes 10 transactions per month. Analysys Mason believe this is because a large majority of Equitel subscribers are Equity Bank customers who acquired their subscription primarily because of the banking features that Equitel offers.

The report goes on to note that since 1Q 2015 Safaricom’s share of mobile money transactions by value has exceeded 80%, although it has decreased from 98.4% in 1Q 2015 to 84.3% in 2Q 2016. According to Analysys Mason, they are not aware of any other mature multi-player mobile money markets where one player has as high a market share as Safaricom across the various measures of market share. They go on to state that given the importance of mobile money in Kenya, it appears very likely that Safaricom’s high share of mobile money transactions is supporting its high share of mobile communications.

8. Conclusion? It Would Appear Safaricom Is Indeed Dominant

Looking at all the data and other insights regarding Safaricom’s position in the marketplace for mobile communications and mobile money in Kenya, its plain to see that its operating as a dominant player. None of its competitors even come close to being competitive and they continue to bleed loads of money.

Going forward, what remains to be seen is whether or not the Communications Authority will proceed with any of the remedies proposed in the Analsys Mason Kenya market study report? It seems likely to me that there is a very real risk of Safaricom’s competitors being pushed out of the market - potentially to the detriment of consumers and businesses by being the only commercially viable mobile network in Kenya.