We all know that 5G auctions are just around the corner. TRAI’s consultation process is complete, and as per press reports, the DOT has taken its final decision which will soon get ratified by the cabinet. Hence, it is now time to understand the operator’s bidding strategy -which will depend on the existing spectrum they hold and the price at which the fresh spectrum is being offered. Operators might also indulge in conservative bidding or no bidding in some bands, hoping that the leftover spectrum will get offered at a lower price in the next auction. Now the question is what is the probability of the price of the spectrum falling in a subsequent auction? What conditions the does regulator apply to drive prices down? Is there any precedence? Has TRAI changed its position now? If yes, in what way? What are these specific rules? What were they earlier? The purpose of the current note is to analyze and find answers to these questions.
Operator’s Bidding Strategy
While analyzing the bidding strategy of the operators, we will go band by band. To enable a bird’s eye picture of the overall situation, I have created a matrix, that not only contains the existing spectrum the operators hold in that band, but also the price at which they are offered in a similar band with similar propagation characteristics. This will help sharpen our decisions.
A)800 MHz band
The following figure captures the operator’s bidding matrix. On the right-hand side of the table (under the remarks column), the reasons for making such decisions are clearly laid-out.
From the above table, it is clear that only RJIO will bid for this band. It will do so with the objective of increasing the block size of its current holdings. Both Bharti and VodaIdea will skip this band. RJIO will end up spending Rs 3775 Cr at the reserve price, and much more if 16 carriers get notified in the NIA.
Note, that only 37% of the total spectrum offered will get sold in this band (in the current 15 carrier case). The key reason for this is due to the fact that some spectrum is offered in quantum which isn’t enough for the operator to complete its existing holdings into multiples of 5 MHz.
Note:- The situation for RJIO might drastically change if the DOT ends up putting one more carrier in the fray. RJIO will be able to complete a 20 MHz block in AP, Bihar & UPE, a 15 MHz block in Delhi, Karnataka, Kerela, TN, etc. Since DOT hasn't notified this decision I have not included it in this note. But once that is done (probably in the NIA) I will revise this note to give the picture at that point in time. This will prevent spectrum in this band from remaining unsold. Hence it goes without saying that 16 carriers is extremely important for RJIO and the DOT (more revenues).
B) 900 MHz band
The following figure captures the operator’s bidding matrix for the 900 MHz band.
Note, that the 900 MHz band is of particular interest to Bharti. VodaIdea will bid too. Both with the intent to complete their existing spectrum blocks into uniform sizes of 5 & 10 MHz each. Bharti will end up spending Rs 3534 Cr — much higher than VodaIdea’s Rs 913 Cr. All spectrum will get taken at the reserve price.
Also note, that only 32% of the total spectrum offered will get sold in this band, but if more spectrum gets released in the NIA the number will increase. The main reason is that most of the spectrum offered here is fragmented and the risk of reorganizing them into larger blocks is higher in this band compared to other bands. This is on account of the presence of PSU holders in between the offered blocks.
Note — In the 900 MHz band, some more spectrum might get released in various circles, by squeezing it from the PSUs. Again it is both in the interest of Bharti, VI, and DOT to do that. Once this gets notified formally I will revise this note. This will also help in preventing the spectrum from being unsold.
C) 1800 MHz Band
The following figure captures the operator’s bidding matrix for the 1800 MHz band.
This is a very important band for all three bidders (Bharti, RJIO, & VodaIdea). All of them will bid to complete their existing holdings into multiples of 5 MHz blocks. RJIO will end up spectrum the most — Rs 11494 Cr at RP. This will be followed by Bharti’s Rs 9093 Cr, and then Vodafone’s Rs 1592 Cr.
82% of the offered spectrum will get sold, with only 18% remaining unsold only due to them being of irregular sizes.
Note — Here as well some spectrum is getting released, some as a vacation of guard bands and some in new circles. But the picture in this band will not materially change as in the case of the 800 & the 900 MHz band.
D) 2100 MHz Band
The following figure captures the operator’s bidding matrix for the 2100 MHz band.
This band is of interest to only Bharti and VodafoneIdea. Bharti will end up taking a larger share at Rs 8350 Cr, followed by VI’s Rs 6235 Cr.
Note that 100% of the spectrum will get sold in this band.
Note:- If the operators get enough spectrum in the 900 & 1800 MHz band, they might as well skip this band to buy more spectrum in 3500 MHz. This probability has been discussed in this note.
E) 2300 MHz Band
The following figure captures the operator’s bidding matrix for the 2300 MHz band.
The only bidder in this band will be Bharti. It will do so to complete 40 MHz TDD (20 MHz FDD equivalent). This will cost Bharti Rs 4430 Cr at RP.
Note- a) TDD spectrum is halved in quantum to align comparison with its FDD counterpart, and therefore its per MHz price is doubled; b) 100% spectrum will get sold in this band.
F) 2500 MHz Band
The following figure captures the operator’s bidding matrix for the 2500 MHz band.
The only bidder in this band will be VodafoneIdea. Please note that this band is segmented into two blocks that can’t be harmonized due to some constraints. Hence, I don’t think it makes sense for VI to pick quantum beyond what they can aggregate into one single block. VI will end up spending Rs 2950 Cr at RP.
Note only 39% of the spectrum will get sold and 61% will remain unsold in this band.
Also note if VI is falling short of funds then it might skip this band to buy more in 3500 MHz as discussed further in this note
G) 3500 MHz Band
This is a new 5G band. The circle-wise reserve price is embedded in the following link. The Pan-India price is Rs 634 Cr/MHz for an FDD equivalent. And for the TDD equivalent, this number will become Rs 317 Cr/MHz.
Now since the 5G spectrum is needed at larger chunks, therefore the Per MHz numbers when applied to it, the outflow number inflates. For example, a 40 MHz TDD spectrum in this band will cost the operator Rs 12,680 Cr. In order to create a robust and effective 5G network, the operator will need at least a 100 MHz block. At RP it will cost Rs 31,700 Cr.
Hence, I am not sure whether it makes sense for any operator to pick 100 MHz now, or settle for a lower number with the hope that the remaining will get offered in the future at a lower price. We will discuss subsequently in this note whether sufficient regulatory clarity exists for this possibility.
For the purpose of this note, I have assumed that the operator will take 40 MHz (TDD) and will wait for the remaining portion to be offered at a lower price in a subsequent auction.
H) 26000 MHz Band
This is also a new 5G band commonly known as the Millimeter Wave band. The circle-wise reserve price is embedded in the following link. This band is the most affordable band compared to all other bands. The price is set as Rs 13.98 Cr / MHz when measured as FDD equivalent. The TDD price is half of this, i.e Rs 6.99 Cr / MHz.
Now if the operator decides to take 400 MHz (TDD), then it will end up shelling ONLY Rs 2796 Cr.
But there is only more reason for the operators to become interested in this band, that is due to the SUC (Spectrum Usage Charge) calculation formula (Read — 5G Millimeter Wave — Good for the health of the Indian Operators). By taking spectrum in this band the operator will be able to lower the license fees outflow on account of SUC to the government by as much as 2.5%. The license fee on account of SUC is a percentage number by which the operator’s topline gets reduced.
Operator’s Status (After Auctions)
In this section, we will discuss the operator’s status after the auction, and how it compares with what exists today.
A) Bharti (Current View — Might Change after NIA)
The following chart provides a bird’s eye view of Bharti’s position.
Note after the auctions, Bharti will be able to increase its tally in the 900 MHz band by 18%, 1800 MHz band by 28%, 2100 MHz band by 51%, and 2300 MHz band by 8%. The rightmost section compares Bharti’s position in various band classes — Low Band (800, 900), Mid Low (1800, 2100), Mid High (2300,2500,3500), MMWave (26000).
Note Bharti’s position in the low band is not that good. Hence, it needs to pick up spectrum in the 700 MHz band, but the price of which is pretty high (Rs 3927 Cr / MHz).
B) RJIO (Current View— Might Change After NIA)
The following chart provides a bird’s eye view of RJIO’s position.
Note after the auctions, RJIO will be able to increase its tally in the 800 MHz band by 15%, and the 1800 MHz band by 47%.
Note that RJIO’s position in the low band spectrum is better than that of Bharti. It will have 235 MHz vs Bharti’s 188 MHz in 800 & 900 MHz combined. The low bands are very important to create a high-quality 5G network that can penetrate inside buildings.
C) VodaIdea (Current View — Might Change After NIA)
The following chart provides a bird’s eye view of VodaIdea’s position.
Note after the auctions, VI will be able to increase its tally in the 900 MHz band by 5%, the 1800 MHz band by 7%, the 2100 MHz by 33%, and the 2500 MHz by 24%.
The VI’s position in the low band (800, 900 MHz) is weak compared to others. It will end up holding only 148 MHz, compared to Bharti’s 188 MHz and RJIO’s 235 MHz.
D) Comparative View
The following chart provides a comparative view of all the operators side by side to enable one to evaluate, how one operator stands against the other.
The reason I do not have an aggregated total of spectrum in all bands for each operator as the high frequencies band (especially MMWave) distorts the overall picture.
Operator's Outflow & Options
The figure below provides a snapshot of the operator’s outflow in the forthcoming auction based on the bidding strategy discussed above.
One can clearly see that the outflow in the 3500 MHz band stands out clearly, even when I have assumed only 40 MHz (TDD) that an operator will end up taking in this auction.
Bharti will have the largest outflow of Rs 41K Cr, followed by RJIO of Rs 31 K Cr, and VI of Rs 27 K Cr (Total of Rs 99 K Cr).
This analysis clearly tells us that there is no way the operators will take a jab on the 600 & 700 MHz bands in this auction.
Both Bharti and VI can optimize a little more by not spending in the 2100 MHz band. But instead, consume those funds to buy an additional 3500 MHz spectrum.
Bharti can skip 2100 MHz, and through that saving (Rs 9093 Cr) can buy another 30 MHz in 3500 MHz.
VodafoneIdea can also skip the 2500 MHz band, together with savings in 2100 MHz it will have Rs 9185 Cr to buy a similar quantum of 5G spectrum in 3500 MHz.
RJIO spending the least has a similar cushion to pick up an additional 30 MHz spectrum in the 3500 MHz.
Now, the question is will the operator pick up more spectrum in 3500 MHz by sacrificing other bands, or wait for the price to drop in the subsequent auctions? The next section is an analysis of this crucial question.
TRAI’s Pricing Strategy
Past
For all past auctions, the pricing strategy for the regulator was simple. If there is a bid, howsoever small it might be (i.e only a fraction of the spectrum is taken), the price of that circle for that band will be called out as “Market Discovered Price”. This means, that in order to prevent a band from becoming set as “Market Discovered Price” the only way the operators had was to prevent making any bids. Only then the price of that band in that specific circle will get reduced in the subsequent auction.
Now
TRAI has changed its pricing strategy in its current recommendation (11th April 2022). Instead of branding the spectrum price as “Market Discovered” even when a fraction of it gets sold, it does so now when only the full block of spectrum offered in the auction got sold. Hence, even if a small amount of spectrum is left unsold then the price of that band will now be considered “Market Discovered”. This means that TRAI’s pricing model will drive the spectrum price down in the subsequent auction even when the band saw bids in the forthcoming auction. Those interested to dive down further can go through my earlier note ( Indian 5G Spectrum Pricing) that explains this change of approach in detail.
An attentive reader can understand the implications of this change in pricing strategy. This means, that even if the operators bid for some spectrum in the 3500 MHz band, whatever is left out will get offered at a lower price in the subsequent auction.
Risk of Bidding
Now we also need to understand whether the change in the approach of TRAI described above is cast in stone? In my view, the answer is No. Hence, the risk of bidding in a band whose price is high might also lead to a price escalation in the future. The reason for my making is conclusion is described under.
In the past, TRAI has been changing its pricing strategy for calculating reserve prices. In every recommendation, they made some tweaks, but for the last two recommendations in 2016 & 2018, they more or less held their logic tight — i.e even when a band is partially sold the price of that band will be branded as “Market Discovered”. Now in 2022, they changed this strategy as described earlier. But there is more to this puzzle.
On 9th May 2022, the TRAI sent its response to DOT on its backreference. TRAI in its original recommendation has stated the following.
For LSAs where the spectrum put to auction in a previous auction is sold, the auction-determined price (duly indexed using applicable MCLR if more than one year has elapsed since the previous auction) should be used for arriving at the reserve price. (Clause II (1) Page 13 of TRAI’s response)
For LSAs where spectrum remains unsold in previous auctions, past recommended reserve price (without indexation) should be used (Clause II (2))
Note, in its recommendation TRAI has not defined what it means by “spectrum sold” and “auction-determined”. As if you follow TRAI’s logic in the same recommendation, it reads differently compared to what we all understood and practiced earlier (i.e spectrum price become market discovered even when a fraction of it gets sold).
Looks like the DOT wants the spectrum to be revalued for all auctions even at short intervals. DOT’s views are reproduced here are reference (Page 13 of TRAI’s response dated 9th May 2022)
DOT is of the view that given the fast-changing techno-commercial ecosystem, spectrum valuation at shorter intervals may be desirable. For instant, in LSAs/bands where spectrum remains unsold, there could be a case for a reduction in reserve prices.
Note — In this back and forth, neither the DOT nor the TRAI has converged on the need for revaluation and the strategy that they will end up following for the spectrum that is left unsold.
Hence, the operators are at their own risk when they decide to bid for a spectrum band (like 3500 or 700 MHz) whose price is way high compared to most international markets.
Conclusion
It is clear from the above analysis that all the operators (including VI) will end up bidding in this auction. All of them will solidify their existing spectrum holdings in the 4G bands by buying more. In the existing 4G bands, Bharti will focus on 900 & 1800 MHz. Similarly, RJIO will focus on 800 & 1800 MHz. VI will pick up a little in 900 & 1800 MHz (VI already has a huge holding in 1800 MHz). All three have to decide whether to bid 3500 MHz and in what quantum. The decision to not bid 600/700 and bid 26000 will be easy (we are all convinced about the reasons), but not for the 3500 MHz, as the price of this spectrum is high and the operators can’t be sure that the remaining unsold spectrum will get offered at a lower price subsequently.
Hence, Bharti might have to vacate its positions from 2100, and VI in 2100 & 2500 to release funds to buy additional spectrum in 3500 in the current auctions only. RJIO is the most comfortably placed. It will have enough funds to buy an additional spectrum in the 3500 MHz band. The govt will end up collecting close to Rs 1 Lakh Crores. But since the low spectrum bands (600/700 MHz) are beyond the reach of the operators, the consumers will have to live with 4G for voice, and patchy 5G coverage till these bands become affordable.
(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.
