One mobile commentator in the form of Chetan Sharma Consulting thinks there is ample space to allow AT&T to merge with T-Mobile USA, without it necessarily distorting the competitive market for US cellular, applying what he calls the ‘Rule of three.’
Again, mobile is not Faultline’s area but the merger between AT&T and T-Mobile seems such an important subject, which will strengthen or weaken one of the real threats to US pay TV, that we thought we’d stay with it for a while.
The thrust of Sharma’s argument is the application of a well-used economic index called the Herfindahl-Hirschman Index (HHI). Sharma doesn’t actually say that the merger should go ahead, but in terms of market weight, parallels the AT&T T-Mobile deal to the situation in other countries, including the UK and Canada. The US Department of Justice uses the HHI as one of its ways of evaluating mergers.
In essence his rule of three is to point out that many markets around the world in many industries, there are often only room for three players, in particular 3 cellular operators often end up dominating a national market. Because applying the HHI index of HHI3 for the top 3 players, is carried out by squaring the market share of the three leading players, adding them together and then normalizing it as a fraction of 1. Hence a market split 30% 30% 30% among the top three players, which looks a pretty competitive market, gives a result of 900+900+900 = 2,700 which is 0.27 in this measure. He says that this is a way of comparing with other markets. We’re not sure it’s any better than just adding up all three market shares and seeing what’s left.
In a market with two players, however you cut it the HHI has to be something like 50% and 50% with the HHI being 0.5, if it is 75% and 25% it’s actually worse at 0.62 and a monopoly has to be 1.0.
Sharma argues that financial regulators use this and other formulae to calculate if a market contains sufficient competition, and that the critical level is a comparison with other markets. Before the merger for the US it is a market split of 33%, 31%, 16% = 0.23, similar to the competitive market shape we looked at above. After the merger AT&T jumps to 42%, Verizon is still on 33%, and Sprint 16% so this becomes 0.31, about the same as the UK.
He does put together a graph showing the cumulative market share of the top three operators in each market around the world, with almost all of them being above 80% but right now in the US sitting at just 73%. After the merger this would be 94%.
Sharma also argues that the rate of decline in US cellular voice minutes is about middle of the league table with India’s prices declining the fastest and South Africa, Argentina and Denmark declining the least. This rate of decline of voice minutes is a clear indicator of competition, in that if operators do not lower voice prices, customers will move.
Another thing he points to is the fact that when looked at against other industries in the US, the top three cellular operators have a more competitive HHI3 than mobile operating systems, the soft drinks industry and way more than mobile search ‘ so heavily dominated by Google. However Google did not reach its monopoly position through mergers which could have been denied it, it did it purely by winning the war.
We think viewing the HHI3 values is a neat way to see that in at least one context, the AT&T merger is not extraordinary, China and Mexico are the least healthy markets according to Sharma and he’s probably right given China Mobile’s crushing lead (although the government has effectively intervened to try to set up three equal parties there and given China Mobile a tougher route to 3G).
When we asked him Sharma added, ‘The point I was trying to make in the paper is that in isolation, the merger will indeed change the structure of the US market, but in the context of global markets and the manner in which markets evolve, it is just following the usual evolutionary theory. That T-Mobile in US is struggling as a #4 player is not an accident .. this is how markets work ‘ by design or by market forces’
And on the subject of whether or not the merger should be allowed, he said, ‘It is not a black-and-white decision, there are a number of factors at play and they should be considered in parallel, otherwise it is easy to come to wrong conclusions.’
There certainly are a number of factors that need to be considered in parallel. AT&T and Verizon are both quad play operators, and they have rock solid revenues in other monopolies (do the HHI3 index in fixed lines and it’s a very uncompetitive market) which can be used to subsidize new markets. But also they can use their assets in the ground to offer bundles that can shape the market. In the UK, one of the markets which he compares unfavorably with the US, Vodafone has no fixed line assets. Also the two market leaders which have merged, Orange and T-Mobile UK, to create Everything Everywhere, aren’t even operating under that brand, but retain their separate brands, so this looks more like a RAN sharing agreement (though this may change). They have no fixed line assets either except a few hundred thousand ISP lines at Orange. The previous market leader O2 also has no fixed assets and certainly 3 UK has no Telco lines. So really this is a five player market, and MVNO Virgin, sat on the T-Mobile network and British Telecom’s MVNO on the Vodafone network, are the only players with fixed lines. There just isn’t the level of bundle control in the UK. Also backhaul is almost exclusively in the hands of British Telecom, which has no cellular offering apart from that MVNO.
If you add the long lines AT&T business, and all of its enterprise success, and look across the board at ways in which it can compete with T-Mobile which has no fixed lines, compared to Sprint and Verizon which do, you could make a case that the reasons for T-Mobile doing so badly could just as easily be because it ONLY has cellular and it’s the only company which does. And of course acquiring fixed ISP lines is too tough in the US, because there is no local loop unbundling because the US regulations for it were successfully challenged in its courts (effectively backing a monopoly over enterprise).
Also the fixed line rivals of Verizon and AT&T are not each other, but the cablecos and they have no native cellular opportunity, only an MVNO with Sprint and a remarketing deal with Clearwire (one cable company Cox is building out spectrum). AT&T has been able to push IPTV services in bundles successfully against cablecos, primarily because it can offer all four services at one price, and the cablecos can’t, not quite.
And if we look back at the US landscape before AT&T bought BellSouth, and Verizon bought Alltel, we are moving from a very competitive market, rapidly to one where there is less and less competition. Is it any wonder that there has been no slowing of cellular voice price erosion? There hasn’t been time yet.
We also have to look at whether or not Sprint remains viable after this deal. It has some fixed lines but these are long distance connections and backhaul is mostly in the hands of AT&&T and Verizon, certainly Sprint doesn’t have enough lines in the ground to offer triple play. So this leaves it in a tough position to survive post AT&T T-Mobile deal, and the HHI3 index after that would go out of kilter with the rest of the world, making the US one of the least competitive markets under this measure.
Of course if the Justice Department said no to the merger with T-Mobile USA, then what opportunities does its parent, Deutsche Telekom, have to sell the operation. It is true that the two companies are joined at the hip by the use of a common technology W-CDMA, in the same way that Alltel and Verizon and Sprint are joined at the hip by CDMA 2000. The issue that T-Mobile USA faced was how does it move up to LTE speeds without sufficient spectrum and while it is losing customers?
Well, in a way that’s not the Justice Department’s problem. T-Mobile could build out LTE or buy wholesale LTE capability from someone like LightSquared, and then it would be in a position to be bought by anyone, because LTE is the natural technology destination for most of the US cellular operators. Hey even in some long distance future it might merge with Sprint and create a balanced three in the US.
Finally we make the point that just because other countries have made mistakes in concentrating their cellular markets into too few hands, it doesn’t mean that the US has to follow that mistake , especially not at a time when it faces calls for greater and greater mobile broadband capacity. Because a monopoly sure as hell won’t build out spectrum in a hurry.