Fahnestock's Bauer/Lee: "Is WinStar becoming a hybrid CLEC like NEXTLINK? We think so." (page 1 of 2)
Investment Opinion: We are reiterating our BUY rating on WinStar. Our year-end target price of $65
reflects a 30% public market discount to our revised 1999 net asset value of $92 per share and offers a 40%+
upside potential from the current price levels. Key points:
· What's in the air and on the ground at the same time? WinStar and NEXTLINK. Seemingly lost
amid the flurry of estimate revisions last week was the dramatic shift in WinStar's competitive positioning
vis-à-vis its terrestrial-only and wireless-only peers. By acquiring extensive fiber capacity (now in 50 of
its top 60 U.S. markets), the company is positioned to reach customers via two distinct access
technologies - wireless and fiber links. The implications are SWEEPING. Hybrid CLECs - those with
multiple access technologies - should ultimately post higher penetration levels, higher margins, and higher
returns on investment than their non-hybrid counterparts.
· NEXTLINK shares soared 240% after it acquired spectrum. In January 1999, NEXLINK
acknowledged its conviction in fixed wireless as an important access technology by acquiring spectrum
from FCC auction winners WNP and NEXTEL. Within three weeks of the announcement NXLK was up
over 50% as the market digested the implications – higher penetration, margins, and returns. In May of
this year, the company hit the road to sell stock, bonds, and to spread the hybrid gospel. By the
conclusion of the road show, NXLK shares had advanced another 60%. Following the road show, the
consensus net asset value for the company doubled to reflect the marked competitive (and financial)
advantages hybrid CLECs should enjoy over time. The rest, as they say, is history. NEXTLINK is now
trading at $60 - up 240% from its pre-hybrid valuation days.
· WinStar shares could soar as the implications of its fiber acquisitions are digested by the market.
WCII's portfolio of access technologies is starting to look very much like NEXTLINK's. Its valuation,
however, has yet to reflect this. The stock is still trading at a deep 50% discount to our year-end net asset
value of $92 per share. As the company's new strategic options are recognized, we think this discount
will continue to shrink and drive the stock closer to a more reasonable valuation in the $65 range.
Table 1
WinStar Communications
Revenue Segments and Growth Estimates
(in $000s)
Compounded
1999E % of Annual 2009E % of
Business Segment Revenue Business Growth Rate Revenue Business
CLEC 242,700 56% 28% 2,759,388 49%
Data 104,872 24% 38% 2,651,177 47%
Williams 5,600 1% 11% 16,000 0%
Other 25,305 6% NM 0 0%
Information Services 55,323 13% 16% 234,383 4%
TOTAL 433,800 100% 29% 5,660,949 100%
Company Description
WinStar is a wireless facilities-based provider of telecommunications services (primarily to
businesses) in more than 30 U.S. markets including Atlanta, Boston, Chicago, Dallas, Los Angeles, New
York City, San Diego, San Francisco and Washington, D.C. During the fourth quarter of 1998, management
announced plans to expand into 60 U.S. markets by the end of 2000. This expansion reflected (in part) the
impact of a $2 billion, 5-year vendor financing deal/strategic partnership with Lucent Technologies. As a
result, the company has a targeted addressable market covering roughly 80% of the entire U.S. business
market. The Lucent deal also accelerated the company's international expansion plans. WinStar plans to enter
up to 50 foreign markets by the end of 2003.
The company holds wireless licenses that provide it with the largest amount of 38 GHz radio spectrum
in the country, as well as a large amount of 28 GHz spectrum and various other spectrum rights. WinStar's
spectrum holdings cover markets encompassing more than 200 million people and as noted, approximately
80% of the business market in the United States.
The company offers customers a variety of individual and bundled services, including local and long
distance services, high-speed data transport, Internet access and other enhanced communications services. It
also provides Internet-based information content and services, such as online business resources. Additionally,
it markets information services in traditional media like television, video, cable and radio and through the
bundling of these services with the company's core products as an enhanced telecommunications service.
Over the past 18 months, the company has invested nearly $1 billion in local and long-haul fiber
capacity. As traffic is migrated onto this fiber (and off third-party vendor's networks), margins should
expand noticeably reflecting the cost savings. Although the company continues to justify its fiber acquisitions
on the basis of cost savings, – we know of no reason why the company couldn't sell CLEC services to
buildings directly connected to its fiber network. If the company were to pursue this strategy (as
NEXTLINK is aggressively doing), we would likely become materially more aggressive with our long-term
operating assumptions.
WinStar has two operations that could provide meaningful upside surprises. The first is the
company's Internet vertical community Office.Com. We believe management has explored the possibility of
a partial IPO of this subsidiary that offers Internet users a single destination for all their industry specific
needs. A successful IPO would highlight the value of this operation (management thinks it could worth $1
billion) to which we (and most others) attribute little to no value at the present time. What's a vertical
community? Say you're a
research chemist for a cheese
company – your site would
contain cheese trade journals,
chat rooms for other cheese
chemists, job listings,
conference dates – in short
everything a cheese chemist
needs to stay abreast of things.
Office.Com is expected to be
on-line this November. The
second area likely to see
surprises is the company's
international arm. At present
WinStar has secured licenses in
four international markets. The company hopes to have operations in 50 international markets within four to
five years – an aspiration not reflected in our model. Table 1 outlines the changing composition of the
company's revenues over the life of our discounted cash flow model (1999 –2009).
Wireless Licenses
WinStar holds licenses (with an aggregate average of 750 MHz or more) in each of the 60 markets it
intends to operate in by the end of 2000. In addition to these markets, WinStar holds licenses in 188
additional markets where its average spectrum approximates 230 MHz per market. Internationally, WinStar
holds a (400 MHz) nationwide license for 38 GHz point-to-multipoint spectrum in Argentina, individual and
“experimental” point-to-point licenses in Amsterdam, The Netherlands and a nationwide “preferred
allocation” for 38 GHz point-to-point spectrum in the United Kingdom. Through its 35% owned Japanese
affiliate WinStar owns 38 GHz spectrum rights covering Tokyo and Osaka. Finally, it also holds licenses (for
a limited amount of spectrum) in the 6 GHz, 10 GHz, 18 GHz, 23 GHz and 31 GHz portions of the radio
spectrum. Table 2 shows the 38 GHz and 28 GHz spectrum held by the company in each of its top-60 US
markets.
Investment Thesis
Wireless networks have huge advantages over fiber networks
Wireless networks can be built at a fraction of the cost and time typically associated with comparably
sized fiber networks. This stands to reason – it's a lot faster (and cheaper) to attach a 12” microwave antenna
to a customer's roof than dig a trench to his building to provide fiber access. As a result, wireless network
operators are de facto the low cost providers in their markets. Fiber, on the other hand, is typically the access
technology of choice for companies with huge bandwidth requirements. Other factors to consider:
· Wireless is cheaper to build. Construction outlays can represent up to 90% of the cost of building a fiber
ring. As a result, declining technology costs will have little impact on the overall expense of these
networks. WinStar's wireless network infrastructure consists largely of 12” dish antennas that it installs
on the roofs of commercial buildings. Once installed, traffic is transmitted to the nearest Winstar “hub”
where it is handed to a fiber network which, increasingly, is owned by WinStar.
· The wireless concept is now main stream. During the early 1990s, misconceptions regarding the
performance and reliability of wireless networks were rampant. The fact is these networks perform on
par with fiber networks in terms of error rates and reliability. NEXTLINK's $1 billion investment in
fixed wireless assets early this year followed by Sprint and WorldCom's buying spree of fixed wireless
licenses this spring finally put to rest any remaining questions about the value this technology.
· The wireless-only addressable market is potentially huge. Less than 3% of the 750,000 commercial
buildings in the U.S. have access to fiber. A portion of these buildings will never generate sufficient
revenues to justify fiber connectivity and a portion may never have DSL access over ILEC copper lines.
As a result, WinStar and other wireless CLECs could have this market entirely to themselves.
Hybrid (wireless/wireline) networks are even better than wireless networks
Fiber networks are expected to capture a majority of the high capacity traffic generated in central
business districts because that's what these networks were designed to do. Unfortunately they can't
economically reach as many customers as a wireless network can because their “spurs” (links to buildings)
typically extend less than a quarter mile from the rings to which they are attached. Wireless network on the
other hand can't meet the capacity requirements demanded by huge customers so they forgo this business. A
hybrid network captures customers in both market segments by using its wireless links to aggregate traffic
from buildings (from up to two and a half miles away) onto its downtown fiber rings. If these fiber rings are
connected into regional clusters, there's a good chance the entire call can be carried “on-net” (the
overwhelming majority of long distance traffic originates and terminates within a 600-mile radius).
WinStar is building a hybrid network
WinStar has paid nearly $1 billion over the past 18 months to acquire local and long haul fiber
capacity. This capacity was acquired via three separate deals the highlights of which are detailed below. By
acquiring local fiber, WinStar can use its spectrum to aggregate traffic at hub sites and use its fiber networks
to back haul this traffic to its switch centers and out to any other markets where it operates (via its long-haul
network.) This strategy matches the right technology with the right job and saves money. Moreover, as far
as we can tell, there is nothing stopping WinStar from selling CLEC services on its local fiber networks. If
the company were to pursue this strategy it would indeed look, smell, and probably be valued more like
NEXTLINK. Should this come to pass we think WinStar could double its current public market value.
WinStar has spent almost $1 billion on fiber in the last 18 months.
July 1998 - $40 million for local and regional fiber
On July 29, 1998, WinStar announced that it had obtained dark fiber capacity in and between a number of
major markets from Metromedia Fiber Network (NASDAQ- MFNX). The deal was the first major step in
the company's plan to build fiber networks in and between its top 40 U.S. markets. The 25-year agreement
(which commenced at the announcement) cost the company $40 million.
· Seven local rings: The agreement provides intra-city fiber rings, consisting of multiple fiber optic
strands, to connect WinStar's hub sites and central offices (which house the company's voice and data
switches) in Chicago, New York, Oakland, Philadelphia, San Francisco, San Jose and
Washington, D.C.
· Six regional rings: The agreement also included intercity fiber optic capacity to connect WinStar's
central offices, including its Class 4/5 switches, serving Baltimore, New York City, Newark,
Philadelphia, Stamford and Washington, D.C., thereby establishing an East Coast facilities-based
long distance network for WinStar.
December 1998 – $640 million for long-haul fiber
On December 17, 1998, WinStar announced the acquisition of 59,460 fiber miles (4 strands over
14,865 route miles) of nationwide dark fiber capacity from Williams Communications. The network will
deliver speeds of up to OC192 and connect most of the company's 60 local markets. Williams will deliver this
capacity as it finishes segments of its planned 32,000 route-mile nationwide network. The last segment should
be delivered in 2001.
As part of the deal, WinStar will pay a total of $640 million over seven years for this capacity and a
seven-year option to acquire two additional strands on its existing routes. Williams (which obtained 2% of
WinStar's long-term capacity in the deal) will pay $400 million for this capacity contingent upon WinStar
constructing 270 hubs by the end of 2001. Williams will pay this amount over four years.
October 1999 - $316 million for more local fiber.
On October 5, 1999, WinStar announced that it had entered into an agreement with Metromedia Fiber
Network to obtain dark fiber capacity in 38 major markets in the United States including Boston, Atlanta,
Dallas/Forth Worth, Los Angeles, Seattle and Houston. The deal also gives the company fiber capacity in
three international markets including: London, Amsterdam and Cologne. WinStar will pay approximately
$316 million over 20 years for an unspecified number of fiber miles in these markets. The deal materially
expands the company's local presence in the U.S by adding both metropolitan ring capacity and fiber-lit
buildings to WinStar's network in these markets. As a result, WinStar will have dark fiber rings in a total of
50 of the top 60 U.S. markets that it plans to serve. The agreement also provides that Metromedia will deploy
fiber into buildings designated by WinStar in each market, including WinStar hub sites and central offices
thereby adding another way for WinStar to expand its broadband connections to customers.
So far all this fiber buying has been justified on the basis of cost savings.
In fact, the company has given little airtime to the hybrid concept (specifically as it relates to higher
penetration levels). This is evident in the comments offered after the announcement of each of the deals
highlighted above. When WinStar announced its $316 million acquisition of local fiber, Chairman Bill
Rouhana noted “This agreement will reduce our operating costs and expand our on-net buildings, thereby
enhancing our EBITDA .” There is no question that the acquisition of fiber lowers costs, we just wonder if it
will be used to boost penetration as well.
Table 3 highlights NXLK's price action prior to and after becoming recognized as a hybrid CLEC.
During 1998 the company devoted a good deal of manpower testing its newly acquired wireless licensees but
gave little airtime to the hybrid concept. Over this period, consensus net asset value estimates for the company
were raised by a modest 17% (from $23 during the first half to $27 during the second half) reflecting stronger
than expected operating results (not related to developing hybrid status). However, in January 1999, the
company ended its silence and emerged from its testing phase a full-blown hybrid-network convert. The
impact of this conversion on the company's stock price and consensus net asset value is evident in the price
action above. Each of the transactions detailed below, contributed to the transformation, and each
announcement served to boost the stock higher. Today NXLK shares are trading at a 240% premium to their
pre-hybrid days.
February 1998 – first toe in
In February 1998, the FCC concluded its LMDS (wireless license) auctions. NEXTBAND (the 50-50
joint venture between Nextel and NEXTLINK) was the second largest bidder in these auctions paying $137
million for 42 licenses. NEXTLINK's portion of this obligation totaled only $67 million so the investment
was viewed (correctly) as an exploratory toe in the water. The company spent the better part of the next year
testing this spectrum at its research labs in Plano, Texas (where the notion of a hybrid CLEC remained
confined). It wasn't until the company announced the acquisition of WNP Communications nearly a year later
that the success of these tests was telegraphed to investors.
January 1999 – hook, line, and sinker
On January 14, 1999, NEXTLINK announced its intent to acquire Nextel's 50 percent interest in
NEXTBAND. The price ($137.7 million) was double the rate paid at the FCC auction just nine months
earlier. The company also announced it was buying WNP Communications for $695 million – four-and-a-half
times what that venture paid at the same auction. WNP was formed solely to bid on LMDS licenses and
emerged from the FCC's auction as the largest bidder having won 40 licenses (39 block A and one B block
license) in markets with a combined population count of 114 million. Because WNP received a 40%
“designated entity” (small business/minority etc.) credit from the FCC, it only paid $186.9 million (roughly
$5.7 million per license) for its licenses, which include the New York, Los Angeles, Chicago, Boston, Dallas,
Detroit and Philadelphia markets among others. NEXTLINK did not qualify as a designated entity and so
received no discount. When NEXTLINK announced the acquisition of WNP the FCC rules required that the
discount received by WNP be paid back to the agency. As a result NEXTLINK paid $152.9 million in
license charges to the Federal Communications Commission. This investment screamed conviction (as it was
warranted) and NXLK shares surged. Within two-and-a-half weeks of the announcement the stock had
advanced by over 50%
NEXTLINK has been extremely vocal about the benefits of its hybrid network structure
Fiber is more economical than wireless for carrying large amounts of traffic. Wireless can reach
customers that are uneconomical to reach with fiber. Supplement these two technologies with a DSL service
offering and you have a “triple threat” i.e., these three technologies allow a hybrid CLEC to serve large
downtown customers with fiber, their branch offices with wireless links and their work-at-home employees
with a digital subscriber line. Whereas wireless-only CLECs fail to capture the heavy downtown traffic and
fiber-only CLECs fail to capture suburban traffic,
the Hybrid captures both by matching the right
technology for the right kind of traffic. The
potential result is quantified in the table at the right.
NEXTLINK enjoys one of the highest 2009
penetration estimates among all CLECs we follow
except McLeodUSA - which is operating in much
smaller markets.
NEXTLINK and WinStar are amassing huge
amounts of spectrum and fiber
Over the last year-and-a-half NEXTLINK
and WinStar have each spent about $1 billion in the
diversifying into each other's “space”. As noted,
NEXTLINK has acquired roughly $1 billion of
spectrum in three transactions over the past 18
months. WinStar has spent roughly the same
amount over the same period for local and long haul
fiber.
If both companies are hybrids, how come the
forecasting assumptions for each are so
different?
Table 4
Comparison of Key Operating and Valuation Metrics
Market Assumptions WCII NXLK
2009 Assumptions
Domestic Markets 60 45
MHZ of Spectrum 45,450 50,900
Addressable Access Lines (000s) 69,635 45,066
Market Share 10.0% 13.7%
Access Lines Served (000s) 6,989 6,184
Monthly Revenue / Line $68 $70
CLEC (Local & LD) 2,759,388 $ 4,893,530 $
Data 2,651,177 $ 6,666,144 $
Other 250,383 $ 29,499 $
Total Revenues (in 000s) 5,660,948 $ 11,589,173 $
Gross Margin % 66% 70%
EBITDA Margin 42% 45%
NPV cash flows (2000-2009) 1,469,200 $ 1,129,196 $
NPV terminal value (2009) 6,413,424 $ 14,067,485 $
Other (1999 ending cash balance) 347,077 $ 1,394,875 $
Gross Asset Value (GAV) 8,229,701 $ 16,591,556 $
Long term debt (000s) (1,921,122) $ (3,655,881) $
Net Asset Value (NAV) 6,308,579 $ 12,935,675 $
Shares outstanding (000s) 68,806 156,400
NAV/Share $92 $83
Current stock price $46.06 $61.50
Current discount to NAV 50% 26%
The accompanying Table 4 highlights selected
year-end 2009 operating assumptions taken from
our 10 year discounted cash flow models for
WinStar and NEXTLINK. The differences in our
2009 estimates can be seen in the higher penetration
and EBIDTA margins we have for NEXTLINK.
An even bigger difference is the guidance from
these companies for data revenues reflected in
Table 4. Assuming WinStar decides to use its
growing mass of local fiber to go after the
downtown market data pie, our assumptions could
change meaningfully.
Management
Wireless technology played such a central role in MCI's business plan, that the company named itself
Microwave Communications Inc. With that said, is it any surprise that WinStar has drawn heavily from
MCI's management ranks to staff its key business units with seasoned telecom veterans who “get it.” As the
following table highlights, an MCI alumnus heads every telecom related business unit.
Management Position at WinStar Joined MCI Prior Position
MCI Alumni
Nathan Kantor President and COO 1994 10 President of Int'l Business
Charles T. Dickson EVP and CFO 1997 10 VP of Finance
Dave Ackerman COO - Winstar Network Services 1994 7 New Product Services
Robert K. McGuire COO - Large Accounts 1996 11 Developer of Marketing Plans
Jack Chidester COO - General Business 1998 17 Global Enterprise Management
Kathleen Flaherty COO - WinStar International 1998 10 SVP of Global Product
Non-MCI Executives
Timothy R. Graham General Counsel 1991 - Nixon,Hargrave, Devans & Doyle
Stuart B. Rekant President - WinStar New Media, Inc. 1994 - President- U.S. News
Richard J. Uhl COO - WinStar for Buildings 1997 - President- Chicago Holdings |
