I have been a reader of SI columns for some time-- without comment. And will continue to avoid comment. I really have nothing to offer. My interest is the discussions that move markets, some here are pretty good. Recently here there have been comments about Reynen Readware. How surprized I was to see his name here-- it has to be the same "Readware". I have never seen him post here. I am responsible for a reasonably large stock portfolio (which has Loral), for whom he and his firm have given research on satellite companies. Here is a paper he gave yesterday, which appears in its entirety below. Let it speak for itself. I doubt that he knows it will appear here (he never posts here as far as I can tell), but how could he mind? From all the recent comments on Loral on this SI board, it seemed some constructive (knowledgeable?) comments were in order. Anyone can of course disagree with what is written in what appears below-- perhaps he/she will need in this case though some strong proof for the disagreement.
And for the record, contrary to what some poster wrote here, Reynen Readware, I do not think, is even 40.
"Be My Sky"
Alcatel-Loral's Skybridge
This is our fourth and final seminar on satellite broadband, as LEO realization of byte transmission at "the speed of light" dawns on the telecommunications age. How so very appreciative we are, then, to have you attend here under the cerulean blue of this vast Catalina sky. I love the quiet of the waves here so much more than the parched clay, however beautiful, of Indian Wells. Let it not be said that the environments for our presentations lack poetry or tone.
Advances in the past five years in LEO spectrum efficiencies and payload, we believe, now make satellite systems competitive with high speed ground networks on a bit error (BER) basis, and economically superior in 25% of the globe where terrestrial transmission buildout is either economically prohibitive or geographically/topographically impossible. Travel the globe, as we have, to Sumatra, the Federated States of Micronesia-- visit Abeokta in Nigeria, Franceville in Gabon, and Guyana-- go to Salto in Uraguay, Iasi in Romania, Belo Horizonte in Brazil-- and see this opportunity for satellite life. We see a material place for satcom broadband in the national rural areas of the developed countries in competition with landline ADSL and xDSL, and economic hegemony for it in populated areas where the satellite terminal is the standout economic preference to terrestrial buildout. Northern Canada, vast areas of mainland China, long stretches of India and Australia immediately come to mind in addition to the areas we named above. The US dollar market for satellite bandwidth delivery worldwide is currently an untapped $8 billion annually, and growing. Others tell us their work shows the market is a factor higher than ours. And for the fiber under the raging oceans-- the Global Crossing platforms-- surely beam spotting at their hubs is a complement to their data delivery systems. Internet competition between the two is not the case.
Electromagnetic spread spectrum enhancements have brought LEO constellations on a technological par with ground transmissions, antiquating the debate about the place of GEOs in the race to provide bandwidth in the 21st century.
In our four prior papers (one on narrowband) we had detailed our expectations for satcom telephony through CDMA satellite power conserving accomplishments for the Globalstar constellation. Path diversity and CDMA (which gives the LEO a 4-fold compression to make the Globalstar capacity the most economic and cost flexible) we saw as its lead over other proposed mobile systems. Its marketing structure through local regional telcos that bill the customer directly we believed was the most certain means of identifying and projecting the addressable market. We believe that the model will prove our forward looking pricing of the company's common as justified. How different from the failed satcom telephony efforts of the 80's.
We then focused on the entrance of GEOs for point-to-multipoint internet transmission, the advantages of a LEO-GEO bandwidth architecture to "feed the global community", and the Alcatel-Loral Skybridge system as the most cost efficient technology towards that end. You will recall our separate paper under the title "Skybridge 1998". The discussion further on will presume the detailed engineering of the system as laid out in that paper is kept in mind.
Discussions on latency, "spoofing", open standards (to at least integrate SONET with others) for universal interface began our study. Deregulation, the so-called "aperture shift", compression advances, the transformation from large receiving antennas to the ultrasats underway at Motorola, Samsung, and Sharp, digital traffic and packet switching improvements continued it. Steerable beam technologies, electronic signal shaping to precision pointing, beam spread reduction efficiencies, beam switching and beam "hopping", correctives of rain attenuation brought us to why satcom systems are technologically sufficient for mass services. While today we look forward towards cellular grid (high-gain multi-beam) super speed (multi-gigabit/second [will we get to terabit?]) on board signal processing in just the next few years, what do we see further out were the other points in our last seminar. On board nano-second processing, compliant encryption, higher wattage systems with less mass, new energy systems to reduce solar paneling dependence; then advanced on board and intra-orbit robotics with AI software to diminish labor intensivity at TT&C, Japan's ETS optical laser links (and its Oicets), and the dramatic progress in frequency re-use as in the Network 28 project (a re-use of 1 Ghz band 1,000 times)-- these issues have given us a far more forward look into the first decade of the next century. Teledesic (a bypass connectivity system [the so-called "mesh network"]), as everyone knows, has been the most vocal on frequency re-use. And with good reason, because of the results seen in signal regeneration improvements. This is all for the sake of economies of comparison with ground systems, of course: direct dollar per bit of throughput comparison will inarguably show sat systems at an economic advantage in currently unserved regions of the world, and for Teledesic's proponents, even in suburban areas of the developed countries. We are not prepared yet to pass judgment on the enthusiasm of the latter. Yet, as we think of frequency re-use gains, recall that the first 30 years of satellite communications has seen an approximate gain of some 500 times in terms of net-cost efficiency for transponder throughput per year of lifetime in orbit. There are no physical limits to increase that efficiency some 500-1,000 fold over the next 30 years. This would mean a frequency re-use capability upwards of 10,000 times or more.
Exciting Milstar developments in floating point processing are on the horizon also, as are satellite superconducting processors that are "cold"-- the Josephson junctions. Microsats, pixel sats, and nano sats-- these bring us to the end of the first decade of the next century.
So "telephysics" is not our subject today for you as potential institutional risk-takers in this new world of "information for everyone". And we all know that improvements in "telephysics" can make for current unimagined "tele-prototype" systems in communications for the future. We believe, having had our prior presentations in which you have participated, that we need to move to the entrepreneurial and risk-taker level-- to the economics of satellite broadband. We can sum up the economics in one phrase: the marginal cost to the satellite provider of the delivery of bytes per second to the end user. Bytes as "streaming video", as teleconferencing, as rich graphics-- eventually will each home have its own internet web page, and will each user terminal have a (Sharp) crystal to give it internet everywhere? This is the vision of the satellite companies, but what are the markets, the demand, and the costs? Again, it all comes down to the cost of delivering the bytes.
In our opinion, the most cost effective system for shareholders in satcom systems will be Skybridge. With it marginal ROI increases, more than in others we think, with frequency re-use dynamics/improvements. This allows it claim of being the supplier of "instantaneos feed" to a market waiting on T1, ADSL, VADSL, xDSL, and beyond.
We attended the Skybridge Washington conference in November of 1998, as well as Satellite '99 in Washington just three weeks ago, and here set forth a summary of what we believe to be the salient investment points for the satcom investor in 1999. Will he realize a return commensurate with the risk? High risk means potentially high returns, and in some instance that return has yet to be realized-- in others it has. In the case of Loral SpaceCom, a discount for "execution risk" is deeply embedded in its present share price-- the Zenit-2 launch failure may be the explanation. Impatience another. We see its February 1999 share price trading at 26 times trailing EBITDA before start-up costs and yet an EBITDA going forward after start-up growing at some 40% normalized (as estimated). The price it currently possesses does not reflect that growth in our view. Yet, if one looks at Echostar, there does appear to be some premium given to its forward EBITDA-- justifiably or not. In the case of Loral, pricing models for developing satcom companies require in their compass timely execution of satellite launches: the satellites are the revenue producers. Those launches were delayed as greater care was taken in the inspection of the satellites, and more demand placed upon ELVs. Quality control is responsible management.
Skybridge is a consortium proposal led by Alcatel S.A. to provide a global broadband system of 80 LEOs and 3 GEOs for high speed access to interactive broadband services and narrowband services such as voice and data. It is an in-orbit solution to the issue of bandwidth limitations in the "local loop" that serves end users. Images, voice, and data need high transmission bit rates for "user friendly" access in a market that has been doubling every six months, and is expected to continue exponentially over the next few years. This means affordable and ease of use ultrasats at the PC level, which is the objective of Mitsubishi, a partner in Skybridge. High reliability of uninterrupted transmission at 20Mbps capacity for residences, and 100Mbps for enterprise clients is the objective of its Ku-band proprietary wave form methodology. This wave form technology enables Skybridge LEOs to optimize the frequency spectrum operating in the Ku band while sharing spectrum on a co-primary basis with GEOs. This is an important issue within the United States, and recent FCC approval of the ITU rules for Skybridge in the United States assures its presence in our land.
Satellite transparency and switching intelligence kept on the ground instead of in-orbit, with modularity of (the Skybridge partner) Mitsubishi built gateways (and ground components) is the further technological safeguard for absolute dependability of the system to deliver "each and every time" to the customer. Its objective of global access to all information highways will be sold by telecom providers at $.03 per megabyte transmitted (= $.04 for a three minute telephone call at 64kbps). These telecom providers will be the major regional ISPs and telecom providers who will be the ground operators for Skybridge. This spreads the business risk, gives incentive for increased business for the providers, and enables Skybridge to meet the regional content demands of the end user quickly and efficiently. In a sentence, Skybridge will leverage terrestrial partnerships to fill the communication gaps in its markets, with the aim of ubiquitous seamless coverage to the world. Unlike other proposals, Skybridge will be a complement to ground systems then (even though, if we may, we do see some perfections in it not available in the DSLs). It will not circumvent the local operator to reach the end-user, but will be the wholesaler of access "at the speed of light". Intersatellite linking is completely absent from the Skybridge protocol. With Skybridge, the technical risk in implementation is low then.
Skybridge will overcome the internet bandwidth shortage through its broadband communications infrastructure of intelligent ground stations and (should we call them "dumb"?) "bent pipe" LEOs. In this way it will be a totally complementary way of delivering text, audio, video, and multimedia stream to a bandwidth depleted world. Narrowcasting (internet broadcasting) will be provided on TV through Skybridge: the internet video image will be totally indistinguishable from video broadcasting of today. Skybridge's Sharp Manufacturing partner will make in volume the consumer video-oriented devices for Skybridge in this regard. "Tele-education" is one principal application for this new technology.
And with Skybridge partner Loral Cyberstar providing GEO point-to-multipoint capability on the Ka band for Skybridge worldwide, enterprise customers receive a distribution/reception advantage of their data that we do not see available in any other proposed system. We see, too, a synergy between the Loral owned Globalstar customer base and the Skybridge proposed customer base. Skybridge believes its customer potential is 20 million users, but the mobile telephony user base is not part of that assumed potential. We can see a faster ramp-up to that FY 2004-05 number with "a Globalstar advantage". This would coincide with the view of satellite bandwidth accounting for 8-9% of all bandwidth revenues in FY 2006 (just short of $20 billion for [we estimate] eight satcoms systems, with 3 majors taking 60%).
The question to the risk-taker now inevitably arises, however: should he believe any of this? Wonderful frequency re-use advances have made the cost of sat broadband delivery profitable for the providers, and further advances will permit an increased customer base and higher profit margins-- according to the laws of productivity. But is the market there for a Skybridge? Can one ask if the DBS market of the early 80's is the model now for the risk-taker's guidance? DBS took a number of arduous years for EBITDA and cash positive flows to surface. "Purchasing" customers was a longer than desired strain on balance sheets, although it is without argument that the satellite industry made the correct decision. It had to first learn how to "market", and those in it who did succeeded. Regional ISPs (and they are hundreds) will, with Skybridge, now have a product where connectivity was impossible, but now with Skybridge at cost parity with fiber connectivity will no longer be an issue. If we see how quickly ISPs "pushed the copper and fiber" in this country to get customers, is there any reason to believe that they will not do the same once Skybridge becomes available in areas now unserved?
"How long before the investment returns match the risk taken?" is the question the risk-taker will ask of a system like Skybridge. 80 LEOs in orbit is no easy ELV task. Iridium launches took a year. And a number of their satellites underwent (premature) in-orbit failures, said by some due to perhaps inefficient on ground thermovac testing. The cause is not our issue: the question is what "guarantees" this will not be the case with Skybridge?
Because it is a "bent pipe" system, Skybridge can gradually ramp up service through the world. Unlike ISL (intersatellite linking), Skybridge does not need 80 sats in orbit for its mission. Nor does it need 60. It can start with as little as 40-- which it intends to do. Launching 40 LEOs and implementing the system will take it 18 months we believe, or to late FY 2001. It will target the more lucrative markets first, we believe, demonstrating the validity and serviceability of its system. It can then phase in more of the system as the markets allow. We envision not a fixed price per month for the Skybridge customer, but payment only on the actual bytes to which he has used Skybridge for access. Along with the normal laws of supply and demand, one can surmise the customer who uses the system the most will receive the greatest discount per bytes accessed.
So what we do think? We believe the system ultimately will cost close to $5 billion to implement. That is not Skybridge's guidance, but we will see. We see a user base, net of churn, of some 6.6-7 million people by FY 2004 covered by 80 LEOs. We think $300 per year (1999 dollar) per home user is a reasonable cost estimate for his use of the constellations's services, yielding some $1.9 billion in annual revenues in FY 2004. By the fourth year (FY 2006) the system should have paid for itself, and cash flow thereafter will permit preparations for a second generation.
By FY 2004 Loral Cyberstar (a 20% owner in Skybridge currently, we believe) should have launched, if Loral's corporate plans are a guide, 3 GEOs (the first in FY 2002) with Ka band capacity for corporate use. Some internet products are not latency sensitive, and it makes sense to use the Cyberstar GEO in this case. The corporate market for Skybridge we see ramping up at a pace of some $100 million in FY 2003, and moving to a range of some $300 million in FY 2004. However, there is caution that is necessary here. We do not know what the market for teleconferencing, telemedicine, and tele-education will be by then. How richly will it develop? When I was in Latin America in January, tele-education was a repeated theme, for example. In the United States, it is never mentioned. Till the wave breaks we are best advised to remain cautious on the corporate market use of Skybridge-Cyberstar. Other GEO systems will have emerged at that time competing for the same market and offering similar services for data carry, videoconferencing, and the like. Our greater confidence in the individual user market is based on our confidence in Skybridge offering superior transmission capability than other providers to the home end user.
Where will the market be? We believe that Latin America can provide some 1.8 million customers, South East Asia and China just under 2 million, Eastern Europe, Turkey, India, Russia, South Africa, and Australia 1 million. The rest of the industrialized world will bring us to the remainder for a 7 million user base.
This, then, is the challenge of the Alcatel-Loral Skybridge constellation that will appear early in the 21st century. Launches of 80 LEOs by 2003 will bring a communications system that can reasonably be called a rapid growth company. We believe the market is there for Skybridge, although we are currently willing to envision only 7 million, not 20 million, users for its system. These are all preliminary calculations requiring caution and constant revision. We are only offering a suggested possibility now, a cautious outlook that will be revised as the market moves forward. And there is time for that.
To us today, however, the sky does look bright for Skyrbridge, a project that offers an opportunity for investors seeking profit commensurate with the risk taken (and the risk we cannot minimize-- for that matter we should not downplay the reward that comes with beating the risk either), and willing to participate in a technology that we believe brings promise to a world soon to be discovered, a world just over the horizon.
For the satellite investor, on that day it will be his sky. |
