GX: Details on Lease vs IRU and Capacity Situation
Global Crossing Ltd(GX)
Rating: 1S
As of 06/19/2001
Last Changed 11/09/2000
Global Crossing Ltd. (GX)#
GX: Details on Lease vs IRU and Capacity Situation 1S (Buy, Speculative)
Mkt Cap: $7,653.7 mil.
June 18, 2001 SUMMARY
* Contrary to comments by a competitor, GX mgmt did not
TELECOMMUNICATIONS alter prev. published financial guidance. In this note
SERVICES we address broader issues raised by competitor
Jack B. Grubman * The pricing environment is no different than what
we've been saying. The rate of change of decline in
bandwidth pricing is clearly slowing rel. to 12 mos ago
Christine R. * On add'l capacity in subsea mkt, we incl. table that
Gochuico shows there is ~15% less cap. being planned in
upgradable form than thought at YE00 as cos like TSIX
are on verge of going out of bus. & other cos as FTHL &
Charles Simonds TCM are jt building.
* Re: econ. cond. & demand grwth, obviously, macro-econ.
climate challenging but reality is GX's overall data
svc. demand growing at a 3-digit rate YOY.
* GX did not say it was changing bus. model to a leased
vs. IRU bus. GX threw out notion that, over time, there
may be financial & strategic adv. to do so. We est. that
even if GX went to more lease, company would remain
fully funded.
* Pressure on stock overdone. Reiterate Buy and $30
target.
FUNDAMENTALS
P/E (12/01E) NA
P/E (12/02E) NA
TEV/EBITDA (12/01E) 7.7x
TEV/EBITDA (12/02E) 5.8x
Book Value/Share (12/01E) $11.68
Price/Book Value 0.7x
Dividend/Yield (12/01E) $0.00/0.0%
Revenue (12/01E) $7,114.3 mil.
Proj. Long-Term EPS Growth 0%
ROE (12/01E) (27.2%)
Long-Term Debt to Capital(a) 41.7%
GX is in the S&P 500(R) Index.
(a) Data as of most recent quarter
SHARE DATA RECOMMENDATION
Price (6/15/01) $8.66 Current Rating 1S
52-Week Range $25.75-$8.66 Prior Rating 1S
Shares Outstanding(a) 883.8 mil. Current Target Price $30.00
Convertible Yes Previous Target Price $30.00
EARNINGS PER SHARE
FY ends 1Q 2Q 3Q 4Q Full Year
12/00A Actual ($0.44)A ($0.61)A ($0.65)A ($0.70)A ($2.11)A
12/01E Current ($0.69)A ($0.93)E ($0.89)E ($0.86)E ($3.38)E
Previous ($0.69)A ($0.93)E ($0.89)E ($0.86)E ($3.38)E
12/02E Current NA NA NA NA ($3.99)E
Previous NA NA NA NA ($3.99)E
12/03E Current NA NA NA NA NA
Previous NA NA NA NA NA
First Call Consensus EPS: 12/01E ($3.14); 12/02E ($3.21); 12/03E NA
OPINION
We recently met with GX management, as did many others. The fact is that GX
management did not comment whatsoever on its financials. Furthermore, we
think that a lot of hysteria has been caused regarding industry pricing,
capacity, demand growth, and most importantly, the notion of IRU to lease
transition that is unwarranted and which we will discuss in this note. The
bottom line is that the rate of change in industry pricing is actually
getting better. There is actually less capacity now in the market than what
people thought six months ago. Demand for most data services is still
growing at a healthy rate.
The transition from IRU to capacity leases is one that has been thought
about, but is not yet happening. Buyers of network capacity, especially the
40 or 50 very large buyers, need predictability of capacity and certainty in
quality of network based services in order for them to represent to their
end-user customers the same. You do not get that type of service level
agreement (SLA) and predictability on short term leases but only in IRU
format. While dark fiber is probably not going to be a very robust market,
we believe IRU wavelength sales for the large buyers of capacity will
continue to be the norm versus short term leases. If the shift from IRU
sales to short term leases does happen, we demonstrate for illustrative
purposes the financial impact in our note, and, by the way, short term leases
are actually positive for the NPV of Global Crossing. Short-term leases are
actually positive to the NPV of GX and will not result in an unfunded
situation.
We believe that there are historic opportunities to buy world-class assets
such as GX that are evolving into world-class operating businesses at an
unbelievably compelling value---GX is currently trading below its book value
of $11.68 and at a TEV/2001E Adjusted EBITDA multiple of 7.7x, or less than
30% our estimated 2001 - 2006 EBITDA CAGR of 26.5%. We would be very
aggressive buyers of GX stock.
Putting some Wall Street comments in perspective on GX, GX management who met
with us and many other analysts and investors over the course of the last
week did not, in one way or another, alter previously-published guidance on
financials.
Putting that aside, we would like to address broader issues. Number 1: the
pricing environment is no different than what we have been saying. The rate
of change of decline in bandwidth pricing is clearly slowing relative to 12
months ago. We estimate terrestrial bandwidth pricing is declining at a 30%
annual clip vs. a 40% annual clip 12 months ago. Subsea pricing in the
Atlantic plummeted over the past 12 months, but forward pricing is declining
at a more modest rate, probably closer to 25% than 35%. At the service layer
of the network, frame relay, ATM and other corporate data pricing is actually
flat to maybe even slightly up on a per port basis. Even voice pricing has
seen no new price points in the market for several quarters.
Turning to capacity coming into the market, particularly in subsea, as the
table below suggests, there is actually less capacity being planned than was
thought at the end of last year. According to Telegeography, there was to be
about 52 Tbps of total subsea capacity in a fully-upgradable form planned as
of YE2000. Knowing what we know today, namely that companies like TSIX on
the verge of going out of business and other companies such as FTHL and TCM
are doing joint builds, we now believe that the fully-upgradable subsea
capacity is going to be roughly 45 Tbps. We expect that this trend of less
available capacity will accelerate. Already, in the first six months of this
year, we have seen a 15% decline in expected subsea capacity. As far as
stranded assets of bankrupt players, or near-bankrupt players, they are
likely going to go to existing entities such as GX, or TCM, on the subsea
side. Similarly, as far as terrestrial networks in North America and Europe
are concerned, there is much less network capacity in the ground, much less
operational, than was thought to be the case 6-12 months ago. If, in fact,
there are fire sales of capacity, it will go to lower the cost structure of
the existing players, as opposed to creating new players.
CAPACITY SCHEDULE Capacity (Gbps) Capacity
(Gbps)
Year-end 2000 Today
Ready-for-Service YE 2000 Fully Fully
or
Europe-Africa-Asia Date (RFS) RFS Upgraded Upgraded
capacity
Africa ONE 4Q 2002 NA 640.0 640.0
FLAG Europe-Asia November 1997 10.0 10.0 10.0
SAT-3/WASC/SAFE October 2001 40.0 80.0 80.0
SeaMeWe-3 September 1999 20.0 40.0 40.0
Pan-Asian
APCN February 1997 10.0 20.0 20.0
APCN-2 September 2001 NA 2,560.0 2,560.0
Australia-Japan July 2001 80.0 640.0 640.0
C2C Cable Network December 2001 NA 7,680.0 7,680.0
East Asia Crossing (EAC) December 2000 80.0 2,560.0 2,560.0
Guam-Philippines (G-P) March 1999 5.0 20.0 20.0
North Asia Cable (Level 3) June 2001 320.0 2,560.0 2,560.0
Trans-American
360americas 2Q 2001 NA 1,280.0 1,280.0
Americas-2 May 2000 80.0 80.0 80.0
ARCOS 1 September 2000 15.0 2,560.0 2,560.0
Maya-1 May 2000 15.0 15.0 15.0
Mercus-1 December 2001 80.0 2,560.0 2,560.0
Mid-Atlantic Crossing (MAC) February 2000 20.0 320.0 320.0
Pan-American Crossing (PAC) March 2000 10.0 80.0 80.0
SAm-1 March 2001 40.0 1,920.0 1,920.0
South American Crossing (SAC) September 2000 40.0 1,280.0 1,280.0
Trans-Atlantic
360atlantic March 2001 160.0 1,920.0 1,920.0
Atlantic Crossing-1 (AC-1) May 1998 140.0 160.0 160.0
Atlantis-2 February 2000 5.0 20.0 20.0
Columbus-III December 1999 10.0 40.0 40.0
FLAG Atlantic-1 / a March 2001 160.0 2,400.0 2,400.0
Gemini Cable System February 1998 30.0 30.0 30.0
TAT 12/13 September 1996 30.0 30.0 30.0
TAT 14 / b April/May 2001 640.0 640.0 640.0
TyCom Trans-Atlantic Cable December 2001 160.0 2,560.0 2,560.0
Yellow/AC-2 / c December 2000 320.0 1,280.0 1,280.0
Trans-Pacific
360pacific / d 1H'02 NA 4,800.0 0.0
China-US Cable System January 2000 80.0 80.0 80.0
Japan-US Cable Network (JUS) October 2000 80.0 640.0 640.0
Pacific Crossing-1 (PC-1) December 1999 20.0 640.0 640.0
Southern Cross Cable Network October 2000 NA 120.0 120.0
TPC 5 January 1997 20.0 20.0 20.0
TyCom Trans-Pacific Q2 2002 5,120.0 0.0
Flag Pacific Q2 2002 5,120.0 0.0
Flag Pacific One (Joint w/TCM) Q2 2002 160.0 0.0 7,680.0
/ e
Total 52,525.0 45,165.0
Notes: Capacity for cables currently in service shows bandwidth as of April
2000. Capacity for planned cables shows initial bandwidth at RFS. a/
Expected upgrade to 280 Gbps by Q1'02, a year ahead of schedule due to demand
for larger capacity purchases. b/ In service on unprotected basis from US to
France; fully protected loop expect to by completed by year-end '01. c/
Expect upgrade to 560 Gbps in 1H'02, a year ahead of schedule due to demand
for larger bandwidth increments. d/360 announced it will terminate its
Pacific build. e/ Initial capacity of 160 Gbps expected by Q2'02. Source:
Telegeography 2001 and Salomon Smith Barney.
As far as economic conditions and growth in demand is concerned, obviously,
the macro-economic climate is challenging. There is no question that certain
carriers who are expecting a disproportionate amount of growth from dot.com
and other IP-only sources of demand (i.e. collocation, soft-switch, etc.) are
seeing tremendous slowdown. However, the reality is GX's overall demand for
data services is growing at a triple-digit rate on a year-over-year basis.
We continue to get evidence in the traditional data marketplace that demand
is continuing to grow at a healthy double-digit rate (e.g. we believe that
WCOM will, in fact, achieve north of 12% revenue growth in Q2'01 despite
carnage in the Brazilian currency which is costing the company several
hundred basis points of growth) despite a clear slowdown in dot.com related
IP-transit, collocation and soft-switched services.
Clearly, the implosion of dot.coms has jolted IP growth, but the biggest
casualties there were companies that built a business model only to serve IP
transit traffic. GX has built a business model to layer on services on top
of its global transport network to serve corporate enterprises and carrier-
class customers for global wholesale bandwidth offerings.
Finally, there seems to be a "brouhaha" surrounding the alleged signal of a
shift to lease capacity vs. selling IRU's. First, GX, in no meeting that we
were at or in any meeting that we know of, said that it was changing its
business model to a leased vs. IRU business. The company threw out the
notion that, over time, there might be advantages financially and, more
importantly, strategically, to do so.
The reality is that big buyers of capacity, especially subsea capacity,
namely the Deutsche Telekom's or WorldCom's of the world tend to prefer IRU
vs. spot market lease capacity. Around the world there are probably 300-400
buyers of network capacity, of which 40 or 50 drive the vast majority of
demand. These entities are primarily PTTs around the world, ultimately
RBOCs, wireless companies and big media companies. These companies need to
have predictability of capacity for network services that they are layering
on that have terms, conditions and service level agreements that allow them
to then put things like core director routers and switches on top of these
wavelengths to deliver whatever service they are delivering to their end
users, whether those end users are buying cellular service, cable TV service
or data services from a telco. This level of network specification and
predictability is not doable on a short term lease basis because if one is
selling capacity on a month-by-month or even a year-by-year basis, the level
of SLAs and predictability are not there.
Thus while we agree that dark fiber is not likely to be a robust market, we
tend to believe that big buyers of capacity will continue to want IRU rather
than spot capacity type of arrangements. The point is this: if and when GX
chooses to pro-actively change its business model, we will adjust our
estimates accordingly to the near-term and long-term financial impacts.
However, as we lay out in this table in the text, we thought that for
illustrative purposes, it might be important to demonstrate the financial
impact of such a decision.
Current SSB Estimate SSB Estimate Assuming
Assuming 100% IRUs Half IRUs, Half 3 Year
Leases (a)
2001E Capacity
Revenues
GAAP (a) $ 216.9 $ 424.9
Cash (b) 2,216.9 1,477.9
(a) Assumes IRU GAAP revenues are recognized over a 20 year life, and Lease
GAAP
revenues are recognized equally over three years.
(b) Cash revenues assume 100% of IRU revenues are recognized up front, and
Lease
revenues are recognized equally over three years.
The punchline is this: GX has roughly $2 billion of IRU revenue. Even if
they chose to consciously go to a leased model, our guess is that no more of
50% of that revenue is represented by customers who would want to enter into
short term leases vs. buying IRUs. That gets us to $1 billion of IRU
revenues that could shift to short term leases. Then, one has to understand
the math. In an IRU, you get 100% of the revenue up front. In a leased
model, revenue is recognized and cash is received evenly over the lease term-
--typically a 3-year lease period. Thus, you get two-thirds of the original
cash in year one even with a shift to short term leases. The remainder comes
over the course of the remaining two years of the lease.
More importantly, from a life cycle perspective, on an IRU, if you lease that
IRU for $1 million for a 20-year period, you get $1 million and that's it.
If you lease that IRU for $1 million on a 3-year lease the capacity pricing
on a shorter-term lease would probably be higher than on a 20-year lease,
because obviously, you are not going to factor in 20-year forward cost
reductions. The fact of the matter is that on a 3 year lease you will get $1
million after 3 years; then, you get to re-lease the circuit again. Even if
you assume a 50% price reduction every 3 years, the NPV of that circuit is
much higher on a leasing basis than an IRU basis. In fact, the knock against
the GX model, from the time of the IPO, was that an IRU business was more of
an annuity business than a perpetual business and people questioned what type
of terminal multiple one should put on out year cashflows. If this went to a
leased model formula, you would have a recurring perpetual business, more so
than an one-off, annuity business. We think that would add predictability
and stability to the revenue flows and probably result in a higher
theoretical terminal value calculation than under an IRU lease arrangement.
Capacity Sales Comparison, IRU Sale vs. Three Year Lease
$ in millions
2001E Capacity 1 2 3 4 5 6 7 8
Revenues
Unit Price (per $ 1,000.0 $ 500.0 $ 250.0
circuit)
GAAP Revenues
IRU Sale $ 50.0 50.0 50.0 50.0 50.0 50.0 50.0
50.0
Three Year Lease (a) 333.3 333.3 333.3 166.7 166.7 166.7 83.3 83.3
Cash Revenues
IRU Sale $ 1,000.0 - - - - - - -
Three Year Lease (a) 333.3 333.3 333.3 166.7 166.7 166.7 83.3 83.3
9 10 11 12 13 14 15 16 17 18 19 20 Total
$ $ 62.5 $ 31.3 $ 15.6
125.0
50.0 50.0 50.0 50.0 50.0 50.0 50.0 50.0 50.0 50.0 50.0 50.0 $
1,000.0
83.3 41.7 41.7 41.7 20.8 20.8 20.8 10.4 10.4 10.4 5.2 5.2 $
1,979.2
- - - - - - - - - - - - $
1,000.0
83.3 41.7 41.7 41.7 20.8 20.8 20.8 10.4 10.4 10.4 5.2 5.2 $
1,979.2
(a) Assumes unit price per circuit declines 50% every three years.
From a funding perspective, GX is fully-funded. Embedded in that assumption
was that it had $1 billion of cash taxes to pay on the proceeds it will
receive from the sale of the ILEC business. The reality is its cash taxes
will probably be closer to $500 million. They have also financed
receivables, which we have never factored in. So the reality is that even if
next year, they chose to go to a lease, vs. IRU, model, any shortfall
relative to plan of upfront cash of maybe $600-700 million would be more than
offset, in our model, by the fact that they have less cash taxes and they
have actually done receivable financing that we have never counted on. In
addition, of course, we haven't factored in any incremental demand coming
from a broader market that might be willing to buy short term leases, and not
willing to buy IRU's. Moreover, in a lease vs. IRU model, CapEx would be
less and would help cushion lower upfront cash.
The bottom line is that GX has built and paid for a global network. They are
moving up the value chain in serving large corporate enterprises with over
1,000 sales and support people calling on large multi-national customers. In
fact, a major financial institution in midtown Manhattan, which is building a
new global headquarters, chose WCOM, VZ and GX as its three telecom
providers. GX is trading at below book value and at a extremely low
multiple, relative to its 5 year EBITDA growth rate of 26.5%, of 7.7x of 2001
adjusted EBITDA and 5.8x 2002 adjusted EBITDA.
The bottom line is this: the macro environment of the industry is actually
getting better, not worse. Less capacity is coming on. Pricing is, at the
margin, getting better, and demand is continuing to grow at a healthy rate.
An IRU sales to lease type of transition would be done pro-actively by GX,
not reactively to the market. If that occurs, the NPV of its assets actually
goes up. The cash impact is minimal and offset by other factors that people
have not baked into their models. In addition, you probably have a more
predictable stable business model where GAAP revenues are closer to cash
revenues. We would take advantage of this misperception in the market place
to be very aggressive buyers of GX stock.
NET/ NET:
In the current market environment, any remark that hits at something bad gets
huge disproportionate impact. We believe that there are historic
opportunities to buy world-class assets such as GX that are evolving into
world-class operating businesses at unbelievable compelling value.
COMPANIES MENTIONED
Flag Telecom Holdings Ltd. (FTHL#, $5.54, 1H)
TyCom Ltd. (TCM#, $15.00, 1S)
Verizon Communications (VZ, $52.45, 1M)
WorldCom Group (WCOM#, $15.80, 1M) |
