Hi Richard:
There are several reasons for my skepticism.
a) The biggest danger facing the telco/ISP space
is overinvestment. In the last 6 months, the enthusiasm
for long haul fiber and the huge market valuations
awarded to QWST, LVLT, GBLX, etc... can only be viewed
as a huge encouragement for laying more fiber (Project
Oxygen, ....). Persons on SI I respect, such as Frank
Coluccio, assert that ultimately the fiber will be needed
when the last Mile bottleneck will be resolved, but
no one at this stage can say what will be constitute too much
capacity. Also, one fact which is not widely known is that we are
still a long way from tapping the capacity of each fiber strand.
Theoretically, the capacity of a single strand of fiber
is phenomenal, due to the huge amount of bandwidth available.
Even with deep WDM technology, we are not even remotely
close to tapping the information carrying capacity of
a single strand of fiber. So, if progress in efficient fiber use
continues to be made, I see the potential for a lot of
idle fiber capacity, and very unhappy shareholders and
bondholders for marginal fiber plays.
b) What has happened already in the fiber area is about to
happen in the BBFW space. How many broadband wireless
networks do you see as being needed in each market? We
already have TGNT (24GHz), NXLK and LMDS license holders
at 28Ghz, WCII (38Ghz), and now the MMDS license holders
(WCOM, FON, BLS...). So this already adds up to four.
A long time ago on the Last Mile thread we had figured that
there would be a need for one BBFW company for each IXC,
but with its cable holdings and its existing 38GHz licenses,
T may actually not need any. This leaves one for WCOM,
and one for FON (or its successor if FON is gobbled by
a European company). Of the existing BBFW play, TGNT is
clearly built to be sold (no long haul fiber) so in my
opinion it will be the first to go. NXLK and WCII may try
to evolve in end to end telcos (like WCOM in some sense),
but they have a long way to go before getting there.
I know some will say that QWST or LVLT, or GBLX could
ultimately take them over, but as we deduced on the Last
Mile Thread, QWST is being built to be bought by a Baby Bell,
so it won't really need a nationwide BBFW partner. LVLT's
fate may be similar. GBLX is harder to predict because
of the huge scope of Annunziata's vision. So, to the
above 4 BBFW operators, how many would you add: ARTT, plus
easily 1 or 2 based on the 39GHz licenses to be awarded.
This would give us 6 or 7 BBFW operators per market. Frankly,
this would be the kiss of death for the entire sector.
Each market has one ILEC, one cable incumbent, and the
picture of a fragmented group of wireless contenders should
send shivers through the spine of shareholders and bondholders
in this sector. In this respect, bondholders are the last
barrier to keeping the BBFW sector sane. We already know
that the TGNT, WCII, NXLK and MMDS wireless networks will
be built. At some point, a lender will have to say no.
May be ARTT will be the last one in the room before the
door closes, but it is cutting it very close.
c) ARTT followers are always emphasizing that it is
not a CLEC, in the sense that it is data only. However,
it is plain for everyone to see that TGNT, WCII and
NXLK will ultimately have both voice and data. Also, if
you think about it carefully, the small businesses
which are served by broadband wireless almost certainly
would love to have both ISP and telephony services
bundled together. So, sooner or later ARTT, WCII, TGNT
and NXLK will live in the same space (in fact NXLK
will cover a bigger market with its wired and wireless
offerings).
The bottom line is that the potential for overinvestment
suggests caution, which in my eyes means investing in the
leaders, and possibly even taking some money off the
table if valuations are excessive (we are already there with
TGNT and NXLK).
Best regards,
Bernard Levy |
