Additional frequencies

The government issued a policy directive last month for auctioning of additional spectrum for Next Generation Mobile Services (NGMS). The Pakistan Telecommunication Authority (PTA) efficiently issued an information memorandum (IMF) spelling out the details of the NGMS to be auctioned, auction process and license template.

There is dire need for more spectrum as the existing ones are not enough to cater for the growing demand of data without significant investments. It is a good step from the government and PTA to allow the opportunity to cellular operators to acquire more spectrum. It is yet to be seen if the cellular operators would prefer acquiring the spectrum at a hefty base price or reuse existing spectrum with the significant investment.

There are quite a few good things in the policy directive and IM, which need to be appreciated. Introduction of spectrum cap is a good step to avoid spectrum hoarding by cellular operators having deep pockets. Offering spectrum in multiples of 5 MHz is another good step. Correcting the fractional spectrum allocation in the past by offering smaller chunks of 1800 MHz to make them multiples of 5 MHz, is also praiseworthy. This should have also been done in 900 MHz as part of the renewal process, instead of deferring it for another 15 years.

In principle, the upper cap on spectrum holdings is a right step, but capping it at 40 percent will result into Pakistan being a three player market. In the proposed scheme, two cellular operators can acquire 80 percent of the spectrum and strangulate the remaining 2 cellular operators, or 3 can get 100 percent spectrum leaving nothing for the fourth one or for a new entrant, creating an entry barrier, thus hurting competition.

If a decision has been made that three cellular operators would be the ideal number for Pakistan, then the 40 percent cap makes sense. If not, then PTA may consider tagging the cap with the market share, which is more practical and fair for all four cellular operators. Current market shares of cellular operators as per PTA's July 2021 data shows Jazz at 38 percent, Telenor at 26.74 percent, Zong at 22.01 percent and Ufone at 12.39 percent. There is no need to put a cap on any cellular operator, who has below 30 percent of either the market share or the spectrum holding to keep it a four player market. In any case, spectrum cap is a good step and should be introduced in each band and on overall spectrum holdings.

There are few things, which do not seem right in the IM. The first and foremost issue is very high base price for spectrum. Spectrum is lifeline of the cellular industry and currently spectrum released in Pakistan, is one of the lowest in the region. The base price determined is prohibitively expense, which will curb the desire of the cellular operators to buy more spectrum especially when the Average Revenue Per User (ARPU) in Pakistan, is one of the lowest in the world.

PTA expressed its ambition in the IM for provision of mobile communications services throughout Pakistan by all the cellular operators, vision of maximum population of Pakistan to be served with high quality voice and mobile broadband services and target serving maximum population of Pakistan in 2-6 years. However, these very noble and commendable ambitions/visions/targets are not matching with the very high spectrum base prices.

On top of the base price, operators will have to pay 10 percent advance income tax and deal with aggressive payment terms with interest. These aggressive payment terms with interest were introduced by the last government in 2014 auction for the first time. The auction results may be different, if the operators are allowed to earn and pay as was done in the 2004 auction by allowing cellular operators to pay 50 percent of the auction money in 10 equal annual instalments with no interest.

By putting the purchase price in dollars, government has shown its lack of confidence in its own currency. It is strange that the cellular operators are earning in local currency but will have to pay in dollars to buy a local input/spectrum. Probably the next stage would be to ask for taxes in dollars. Interestingly as per IM, PTA will make refunds only in local currency and not in dollars.

In my previous articles, I suggested to delink the spectrum license from service licence but PTA seems to be in love with the idea of a separate service license with every chunk of new spectrum. All of the potential bidders already have cellular service and NGMS licenses, so why to burden them with another unnecessary service license and increase the cost of compliance to different obligations in different licenses.

It would be interesting to see a cellular operator only buying 1.2 MHz of spectrum in 1800 MHz to top-up its fragmented allocation, given a service license with roll-out and quality of service (QoS) obligations, separate from its existing service and NGMS licenses. Moreover, the cellular operator will have separate expiry dates for its 8.8 MHz and 1.2 MHz spectrum. Ideally these 1800 MHz round-ups or top-ups should have been done as part of license renewal process and not separately through auction.

PTA has planned to sell spectrum through two separate auction mechanisms to the same potential bidders. Both Clock Auction and SMRA (Simultaneous Multiple Round Auction) have their respective pros and cons but without getting into that debate, choosing both does not appear to be a good idea.

It will only add complexity in the auction process and confusion among bidders with varying eligibility points. Auction of both 1800 and 2100 could have been efficiently and easily done under SMRA.

PTA has good experience and tested tools for SMRA.

IM claims that the spectrum allocation is technology neutral but permitted only FDD technologies and impliedly prohibited TDD technologies.

Consumers are the most important stakeholders, but there is no opportunity given to them in the consultation process. Their voice is totally missing in this very important process of finalising terms and conditions of allocation of spectrum, roll-out and QoS. PTA might have gained by keeping sector experts in the loop from their contributions.

With 4G already available in the country and 5G soon to be introduced, there should be sun-set on 3G. It is not understandable why PTA is encouraging operators to keep 3G with average speed of 512 kbps. An average speed of 512 kbps for 3G and 2 mbps for 4G has been proposed in the draft license template.

These speeds might have been OK in 2014, when these technologies were new in Pakistan, but in 2021 these speeds are ridiculously low. With these speed requirements, operators will have no motivation to improve.

Neither in the IM nor in the draft license template, one could find an mandatory obligation to connect X percentage of towers with fibre for better broadband experience. In the absence of such intervention, QoS is not likely to improve in near future. Interestingly Universal Service Fund Company (USFCo) has made efforts to fill this gap by contractually binding the operators laying fibre, to connect 25 percent of the cellular towers in the contracted area on a special tariff, to increase quality of broadband in rural and under-served areas. If USFCo, with no regulatory power, can do it, why PTA cannot intervene and introduce such obligation through regulations or license obligations.

Government has not given clarity on spectrum roadmap, when would be the next spectrum auction and which spectrum will be made available when. Government should stop the practice of starving the operators before issuing new limited spectrum at hefty base price, in the hope of extracting maximum dollars.

Such practice only hurts the economy and the consumers. This same approach led to unsold spectrum in 2014 auction and later in 2016 and 2017 there were only single interested cellular operators, desperate for spectrum. I can clearly see in the crystal ball that the history is going to repeat itself in the upcoming spectrum auction.

On spectrum rationalisation, IM reveals that consent of cellular operators has been obtained for spectrum, however there is possibility that any cellular operator may withdraw its consent or otherwise delay the process or start litigation, when the actual process starts. PTA should have developed the plan before the auction to give confidence to the cellular operators of what their revised bands would look like, as the winner would like to use the allocated spectrum immediately. PTA needs to have a risk mitigation plan ready to avoid delays in the rationalisation process. There also needs to be clarity as to who will bear the cost, if incurred, to make such adjustments, eg, change of filters, buying of new equipment or software license from vendors etc.

Rationalisation process should not be limited to 1800 MHz only. PTA should also do this exercise for 2.1 GHz, as the contiguous spectrum to existing allocations would make the use of spectrum more efficient for the winning operators.

Active sharing and spectrum trading was mentioned in the Telecommunications Policy of 2015 and still there is no regulatory framework. One wonders, how many more years it would take for PTA to have such framework in place.

In the end, let me make some predictions about the upcoming spectrum auction with a sincere wish that I am wrong:

• No new entrant will participate in the auction;

• There will not be an overwhelming interest from operators;

• Only operators desperate, or with loads of money, will participate in the auction;

• There will not be aggressive bidding;

• Auction will be over on day one of bidding;

• Most of the spectrum will be sold on base price or close to it;

• Some spectrum will remain unsold despite the fact that cellular operators need it; and

• No significant improvement will be made in QoS in the next one year.

Even if half of the above predictions come true, government and PTA should rethink their rent seeking strategy and work towards long-term benefit to the people and the economy of Pakistan.

The writer is an ICT Regulatory expert