On 5th May 2022, TRAI sent a response back to DOT on its reference dated 9th March 2022 on the subject of Licensing framework for Satellite-based connectivity for low-bit-rate applications. The purpose of this note is to explain the real issue that is being deliberated, and if the TRAI recommendation gets finally accepted by DoT, then how spectrum charges (SUC) for services delivered via satellite compare with those offered through terrestrial means. And will it create any contradictions? To prevent the discussion from getting complex, I will limit the scope of this note to only Spectrum Usage Charges, and leave other issues to be analyzed in a follow-up article.
What is the problem with DOT’s Current Rule?
Current Rule
Currently, the spectrum charge is computed as per the DOT order dated 22nd March 2012. Charges so computed are applicable for each link separately. In other words, each link is treated individually for the purpose of this calculation.
Annual Royalty, R (in Rs) = 35000 x Bs; [Bs is defined in the table below]
The Problem
The above rule was devised at the time when the satellites used low bandwidth and for closed user group applications. But as the technology has matured over time and both bandwidth/data speeds have increased significantly. Now, as per the existing formula if an operator wants to deploy a satellite with BW of 500 MHz of both uplink and downlink, then he has to pay Rs 7 Cr annually for each site that he plans to operate (assuming the site operates both uplink and downlink simultaneously). This makes the current rule infeasible for the new generation of satellites using more bandwidth for offering high data speed internet services.
What did TRAI recommend for resolving this problem?
TRAI has recommended two different approaches for charging SUC (Spectrum Usage Charge) for the satellite spectrum.
A) NLD (Backhaul)— 1% of AGR (Adjusted Gross Revenue) generated by the long-distance traffic flowing ONLY in the assigned satellite spectrum. In other words, the revenue so generated (within the satellite spectrum) needs to be separated from the other NLD revenues of the operator for the purpose of calculating SUC. According to TRAI, for NLD services, this accounting separation is simple and CAN be done easily without any difficulty.
B) Access (Last-Mile) — 1% of AGR (Adjusted Gross Revenue) of the total access revenue from all access services. In other words, the satellite revenue needs to be added to all other revenues for the purpose of calculating SUC. According to TRAI, for access services, accounting separation is complex and CAN NOT be done easily.
Why did DoT send the recommendation back to TRAI for review?
DOT is asking TRAI to examine and provide clarity on the feasibility for an NLD service provider to segregate revenues (and consequently separate accounts) accruing from satellite connectivity. As support to its request, the DOT has pointed to clause 2.87 of TRAI’s earlier recommendation dated 28th July 2020, wherein TRAI has mentioned the following.
“…Migrating spectrum usage charges for using satellite frequencies from the formula-based mechanism for Access Service licensee will be a complex issue, as it is difficult to do the accounting separation for revenues accruing through the use of satellite connectivity…..”
In other words, the DOT asked a very simple question — IF it is not possible to segregate revenues emanating from access services, why should it be possible to segregate them for the NLD services?
What is TRAI’s Response to the DOT’s backreference?
As per TRAI, its recommendation is based on the premise that satellite-based NLD is a separate identifiable vertical, and therefore, the revenue emanating from it can be segregated and maintained easily compared to satellite-based access services (para 2, Page 9 of TRAI’s Response dated 5th May 2022).
Doing so (as per TRAI) will enable the NLD operators to provide satellite services in a cost-effective manner.
What contradictions does the TRAI approach end up creating?
TRAI is repeating the same mistake which DOT committed in 2010 when it recommended 1% SUC for the BWA spectrum (being used for access with revenues segregated and reported separately) in the 2300 MHz band when it was put out for auction. This made the process of calculating SUC very complex. Till Feb 2021, the revenues from the 20 MHz block (auctioned in 2010) were separately treated for the purpose of calculating SUC floor % (minimum SUC payable to the operator). Later on 28th Feb 2021, DOT removed this complexity and made the SUC order simple.
Hence, the question — Will history not repeat itself in NLD with TRAI suggesting segregating revenues emanating from the satellite backhaul services? If this is so simple to implement, then why did TRAI not recommend a similar approach for calculating SUC for microwave spectrum for running backhaul services (% of which increases proportionately with the number of MW carrier frequencies the operators hold, but SUC is calculated on the total operator’s revenue)? If it is so easy to segregate revenues from satellite backhaul services, the same should be possible for MW backhaul too, no?
Then, what about the SUC structure for the E-Band spectrum which DOT is thinking of assigning to the operators. Will TRAI recommend a similar SUC structure for the E-Band spectrum too? If they do will it create an accounting mess for DOT to manage, and the possibility of arbitrage?
Are there any other issues that we must be aware of?
Yes, what? The difference between the SUC calculation formula for access spectrum and satellite spectrum. Now SUC for access spectrum is calculated based on a weighted average formula. On 15th Sept 2021, the cabinet, through a press note, announced a set of reforms for the struggling telecom sector of India. In this press note, the reforms were bucketed under two heads — a) Structural Reforms; b) Procedural Reforms. Point number 7 of the structural reform is reproduced here for easy reference.
“No Spectrum Usage Charge (SUC) for spectrum acquired in future spectrum auctions.”
Now, as the operators acquire more 5G spectrum, the overall SUC will go down to a level that will be negligible. Read — “Will 5G impact SUC, if yes, how?” But if the SUC for the satellite spectrum is pegged at 1%, then it will add on top to the existing SUC (access+MW) and thereby, increasing the overall SUC of the mobile operator by 1%. This will discourage the mobile operators from using the satellite spectrum for delivering access services.
What should be the Solution?
SUC for access spectrum for both satellite and 5G should be the same, 0%. This will ensure that operators with the satellite spectrum are not unnecessarily burdened with additional SUC (on their total revenue) and therefore NOT demotivated from using it for disseminating access services.
SUC for backhaul spectrum (satellite, E-band, etc) if acquired through auction should be pegged at Zero (at par with 5G). So that operators have the full flexibility to use it for both access & backhaul. Otherwise, it will lead to operational inflexibility. The current weighted average formula should be amended to include such acquired through auctions.
SUC for non auctioned spectrum (access or backhaul) should be pegged at 1%, but no revenue segregation should be allowed. This shall mean the SUC of all non auctioned spectrum including MW will keep cascading on top of each other. In order to prevent the overall SUC from getting too high, the DOT can decide on a reasonable cap.
Conclusion
In my view, TRAI’s suggestion to segregate revenues emanating from satellite NLD services for the purpose of calculating SUC is wrong and will lead to huge problems in the future as described above in this note. Also, to ensure parity between technologies (5G, Satellite, MW) — the calculation and % rate of these technologies should be brought to par. In other words, SUC for all spectrum acquired through auction should be pegged to zero as is being done for 5G, and the operator should have the full flexibility to use it for both access and backhaul. SUC for all non auctioned spectrum (satellite or microwave) should be pegged at 1% with a suitable cap (and no revenue segregation should be permitted) so that the overall payment obligation of the operators does not become too overwhelming due to the cascading effect of the individual SUC percentages.
(Views expressed are of my own and do not reflect that of my employer)
PS: Find the list of other relevant articles in the embedded link.
