Telecom: Analyzing market share: Spectrum, legacy, execution, brand, timing ::Kotak Sec

Please Share:: India Equity Research Reports, IPO and Stock News
Visit http://indiaer.blogspot.com/ for complete details �� ��


Telecom
India
Analyzing market share: Spectrum, legacy, execution, brand, timing. We analyze
the revenue market share of Indian wireless names to get a broad picture of market
share determinants. Even as spectrum holdings (quality/quantity) and legacy (first-mover
advantage) emerge as key determinants, disadvantages on both these counts can be
overcome through good execution, as seen in a few instances. We reiterate our
constructive stance on incumbents Bharti and Idea with an ADD rating on both
Analyzing determinants of revenue market share - methodology
In this note, we try and explore the importance of spectrum holdings and legacy to wireless
players’ relative revenue market share (RMS). Even as these two turn out to be key determinants of
market share, our analysis throws up (1) a few instances where players have overcome
disadvantages on these counts to gain RMS disproportionate to their relative positioning on
spectrum and legacy - Bharti/Vodafone/Idea in Bihar, Vodafone in West Bengal, TTSL in Haryana,
Aircel in North East are good examples, and (2) several instances where the RMS of players has
been poor despite scoring high on spectrum/legacy parameters – RCOM and BSNL on a pan-India
basis, and even other players (excluding Bharti and Vodafone) in select circles have under-delivered.
For the purpose of our analysis, we clubbed spectrum and legacy into a single metric capturing
both – this metric is expressed in MHz-years and essentially depicts time-weighted spectrum
holding. We also made suitable adjustments for spectrum quality (1800 MHz GSM spectrum
assigned a weight of 1, CDMA and 900 MHz GSM spectrum assigned a weightage of 1.5) to
arrive at circle-level spectrum-legacy metric for various operators. We then compared spectrumlegacy
share versus revenue market share to draw a few inferences. A key element that is missing
from our analysis is relative capex market share – this, and more importantly, the timing of this
capex (whether done during the high-growth phase of the market) would probably explain some
of the observed discrepancies. We have limited our analysis to the top-7 operators. We also note
that we have used adjusted TRAI-reported AGR data for our analysis.
Bharti and Vodafone stand out on maximizing their spectrum-legacy combination
􀁠 On a pan-India basis, only Bharti and Vodafone have revenue market share (RMS) higher than
their spectrum-legacy market share (SLMS). In fact, both these players underperform on this
(RMS/SLMS) ratio only in one circle each – Bharti in UP (West) and Vodafone in M.P.
􀁠 Bharti leads all players with a RMS/SLMS ratio of 1.7X on a pan-India basis. Vodafone is 2nd with
a ratio of 1.3X. Idea is close to 1 while all the other players have ratios less than 1. We also
highlight that Idea has delivered the maximum improvement on this metric over the past two
years; Vodafone has also improved while Bharti has been steady.
􀁠 BSNL/MTNL, not surprisingly, have fared the worst – their RMS/SLMS ratio of 0.35X is just about
20% that of the leader Bharti – valuable spectrum is clearly being under-utilized in the hands of
the PSUs. RCOM is the worst among the private operators at 0.8X.
􀁠 Spectrum-legacy disadvantage can be overcome – Bharti in Bihar and Vodafone in WB are good
examples. Bharti entered the Bihar market as late as 2004 and now has a 45% RMS with an
SLMS <15%. Similarly, Vodafone, having entered WB in 2004, now has a 37% RMS with
10.6% SLMS. We do note that RCOM and BSNL were the leading operators in both these
circles when Bharti/Vodafone entered.


Key inference – challengers have a stiff task at hand unless regulators intervene
to ‘level’ the playing field
Coming from behind to gain reasonable RMS is not impossible, as discussed earlier in the
note. Nonetheless, in the examples we have cited, there are a few common factors that
helped.
􀁠 Reasonable timing – players with RMS disproportionately higher than their SLMS have
been present in the market during its growth phase; even as Bharti was a late entrant into
the Bihar market, it rolled out aggressively (risked massive capex) to overcome network
disadvantage and in fact expanded the market with its coverage leadership. Bulk of
revenue growth in Bihar has happened since Bharti entered this market.
􀁠 Weak competition – in most cases, where we have seen disproportionate market share
gains by challengers, incumbents have been weak, scoring poor on strategy and/or
execution. We repeat that in the both the examples of challengers’ successfully dislodging
leaders off the perch (Bharti in Bihar and Vodafone in West Bengal), the top players when
Bharti or Vodafone entered were RCOM and BSNL.
􀁠 Brand strength – there is a reason why we see most ‘come from behind’ success
examples featuring Bharti, Vodafone, and Idea and very few involving others.
􀁠 Execution and strong financials – coming from behind requires ability to sustain
investments in ‘new’ circles for a long time. Positive EBITDA from established leadership
circles to fund new circle expansion is an advantage as sustained external capital (debt
and/or equity) inflow can become a challenge beyond a point.
In the light of these factors, we believe the challengers in the Indian wireless market (TTSL,
Aircel, and even RCOM/BSNL) have a tough task at hand. Successful incumbents’ spectrumlegacy
advantage can not be wished away. Price discounting has not succeeded in yielding
enough RMS gains to improve business case viability of the challengers. Long-term business
case for the challengers, hence, hinges on one or more of the following.
􀁠 Sharp and sustained increase in RPM without a decline in market volumes or
volume market share – slightly wishful thinking on both counts, in our view. Beyond a
point, increase in tariffs will lead to negative volume elasticity in the market. Also, volume
market share of a challenger is less sticky than those of incumbents’, in our view – a good
portion of this has been gained through the ‘low pricing’ promise; indiscriminate price
increases (even if in line with what the incumbents are doing) could see such market
share go away in a hurry. We believe challengers have to be careful on this aspect –
their long-term business case has to be combination of volume market share gains
and pricing improvement; neither of these alone can deliver a viable business model for
the challengers, in our view.
􀁠 Regulations bridging some of the spectrum-legacy gap versus the incumbents –
regulatory interventions that can do this would include spectrum refarming, meaningful
difference in base pricing for spectrum upto 6.2 MHz and beyond 6.2 MHz, and
substantial increase in recurring spectrum charges for higher spectrum holdings. We note
that some of these regulations have been proposed and may become reality in the
upcoming NTP or Spectrum Act. Nonetheless, these should be seen as necessary but
not sufficient conditions – challengers will have to improve their execution
capabilities several notches or they would fail to convert the positive regulatory
development into improved positioning in the marketplace.


􀁠 More spectrum made available to the challengers at no incremental cost – as per
the license agreement, the challengers are eligible for upto 6.2 MHz of GSM spectrum;
they have been alloted only 4.4 MHz (in most cases) thus far. This factor ties in to the
factor discussed above.
􀁠 More risk capital – most challengers are nowhere close to cash breakeven. More
importantly, even at a volume market share little higher than their current levels, they
would need substantial pricing improvement to achieve cash breakeven. More risk capital
(equity infusion) is a must in our view; also, this has to be deployed efficiently to gain
scale (higher volume and revenue share) through organic coverage expansion and/or an
inorganic move.
􀁠 Improvement in execution and strategic clarity – good execution and focused capex
can help overcome spectrum-legacy disadvantage, as we have discussed earlier in the
note. In addition to improving execution (cliched as it may sound), challengers need to
assess their strategic focus more carefully, in our view – merits of a pan-India presence
and aggressive next-gen technology (3G/3.5G/4G/…) investments need to be
weighed against balance sheet constraints. Even as we have not considered Uninor as
part of our exercise, we suspect it would look better on RMS/SLMS ratio than many of the
other challengers. They have respected their balance sheet constraints and adjusted their
business plan to market realities better than some of the other challengers, in our view.
To conclude, we see low probability of disruption to the current benign competitive
environment in the industry on account of any of the above factors. However, we do
note that (1) over-capacity in the industry is a reality, and (2) RPM improvement
required to make all of this excess capacity viable is improbable without affecting
volume assumptions. This dynamic can change the current benign competitive
scenario in the industry and trigger another price war in a few quarters from now
unless meaningful consolidation happens.
Computation of spectrum-legacy market share (SLMS)
We illustrate the computation of SLMS used in our exercise with a hypothetical example.
Assume there are three players A, B, and C in a particular circle. We compute the spectrumlegacy
holding of each of these as below
􀁠 A – assume A holds 10 MHz of GSM spectrum, 6.2 of which is in the 900 MHz band and
the balance 3.8 in 1800. Also, 4.4 MHz of 900 band spectrum was alloted in the year
1998, additional 1.8 of 900 band in the year 2002, 1.8 MHz of 1800 in the year 2004
and the balance 2 MHz of 1800 in the year 2006. Essentially A has held 4.4 MHz (900)
for 13 years, 1.8 (900) for 9, 1.8 (1800) for 7, and 2 (1800) for 5. A’s spectrum-legacy
holding would be = (4.4 * 1.5 (900 factor) * 13) + (1.8 * 1.5 * 9) + (1.8 * 1 * 7) + (2 * 1
* 5) = 132.7 MHz–years.
􀁠 B – assuming B has held 2.5 MHz of CDMA spectrum for 10 years and 4.4 MHz of GSM
(1800) spectrum for 5, B’s spectrum-legacy holding will be = (2.5 * 1.5 (CDMA factor) *
10) + (4.4 * 1 * 5) = 59.5 MHz-years.
􀁠 C – assuming C has held only 4.4 MHz of GSM (1800) spectrum for 4 years, its spectrumlegacy
holding will be = 4.4 * 1 * 4 = 17.6 MHz-years.
With the above computations, A’s SLMS works out to 63.2% (132.7 divided by (132.7 +
59.5 + 17.6)), B’s works out to 28.4%, and C’s works out to 8.4%.


We present various cuts of our analysis (operator-wise and circle-wise) in Exhibits that follow.
We highlight that the RMS ranks of most private operators would appear better than their
SLMS ranks purely on account of the massive PSU under-performance. It does not reflect the
relative differences in execution between the private operators. This is better reflected in the
RMS/SLMS rank.