After ignoring NN throughout the discussion, it gets mentioned as a BUY in the closing comment...
Seems as if Lutz still has a lot more PR to do...
Ian.
++++++++++++++++++++++++++++++++
Updated: 22-Feb-99
StreetBeat is designed to provide you with additional insights on the market from recognized financial experts on (and off) Wall Street. Please note that the views and opinions expressed by the panelist below are not necessarily those of Briefing.com.
This week's topic: Networking
Panelist
Seth Spalding, Broadband Communications Analyst at C.E. Unterberg Towbin
Patrick Houghton, Senior Vice President and Senior Telecommunications Equipment Analyst at Sutro & Co
Q&A
Briefing: How soon before we see true convergence between the "old world" telecommunications infrastructure and "new world" packet-based architectures?
Seth Spalding: The current telecommunications architecture is essentially a circuit switch architecture that was designed for voice. The "new world" architecture is moving from analog circuit switched to a digital packet-based or IP (Internet Protocol) architecture in order to accommodate the massive amounts of data traffic resulting from the rise of the Internet. This presents an interesting situation because while there is a huge installed base of circuit switched architecture, the industry is moving in the direction of a packet-based format which will carry voice, video, and data in a unified format. Packet-based technology is currently in place for data traffic, however, the technology necessary for an end-to-end packet-based architecture that will allow reliable, real-time voice transmission is still about four to five years away. In other words, the hype surrounding end-to-end packet-based technology is ahead of the reality. I think that we will see an organic migration to packet-switched networks versus a binary shift. The industry, right now, is developing around gateway applications which is the technology between the circuit switch and the packet switch worlds. Cisco (CSCO) and Nortel (NT) are the leaders in gateway applications.
Patrick Houghton: The convergence has already happened on the transport level, however, the convergence of circuits into IP will take longer. I think that we will see a cap and build approach to this issue. Circuits won't be torn out but rather left in place to work in conjunction with the new IP technology. The important tenant here is "follow the money" and the money right now is in voice. Nearly half of the bits that are transported are data but of the total worldwide telecommunications services revenues of about $700 billion, Sutro estimates that only about $50 billion are data. Therefore, the rate of convergence will follow the rate of revenue growth in data. I think that in the next five years, revenues in voice will stay relatively flat at about $650 billion but data revenues will grow at around a 50% compound annual rate putting datacom (data will include some voice at that point) revenues at roughly $400-$500 billion. Telco equipment revenues will follow suit. Total equipment revenues are about $200 billion with $25 billion in data equipment and $175 billion in voice equipment. In five years, we will see about $300 billion in telco equipment and about half of those equipment revenues will be data kinds of equipment. Again, Telcos will follow the money and the pace at which service revenues migrate to data will dictate the pace of the convergence.
Briefing: Who is best positioned to succeed in the coming age of converged communications?
Seth Spalding: Unquestionably, the two prominent players are Lucent (LU) and Cisco (CSCO). Lucent dominates in the circuit switched world and has a huge installed base of customers which will continue to be significant asset going forward. Their acquisition of Ascend Communications (ASND) will establish Lucent as a player in the packet-based world. Cisco is coming from the packet world and through acquisitions, like SummaFour, is merging into the telco world. Cisco has a knack for making strategic acquisitions and making the integration a success. They are well positioned in the gateway application arena and will emerge as one of the dominant players in the world of packet-based telecommunications
Patrick Houghton: Simply put, Tellabs (TLAB) and Cisco (CSCO). Tellabs dominates the core of the network with 60-65% of the cross connect market and Cisco dominates the edge with over 70% market share in routers. Tellabs' products straddle the old world architecture and the new. Its growth with cross connects from 1990 to today has come primarily from data. The basic architecture of the data network is still cross connects at the core and routers at the edge. Tellabs' cross connects are morphing into higher level routing/switching devices with ATM fabric at the core. Similarly, Cisco's routers are morphing into IP switches. Tellabs sells to the telcos such as Qwest (QWST), AT&T (T), and MCI Worldcom (WCOM) and Cisco sells to the ISPs like America OnLine (AOL) and UUNet (division of WCOM). In fact I think that merger of Tellabs and Cisco would make for a formidable company, but neither entity has expressed any interest in the idea.
Lucent (LU) and Nortel (NT) are the other two players that are gunning for business at the core and edge of the network. They stand to benefit from any growth in bits as does Uniphase (UNPH) and Ciena (CIEN).
Briefing: What are the applications and the underlying technologies that will drive the networking market in 1999 and who are the leaders in those respective areas?
Seth Spalding: There are four major applications of importance in 1999 and going forward. Gateway Applications: again, this is the marriage of the circuit world and the packet world. Cicso (CSCO) and Nortel (NT) are currently the dominant players. Cable Telephony: using the cable infrastructure for voice. ANTEC Corporation (ANTC) is the strongest player in this market at the moment. Down the road, I think that Terayon Communication Sys (TERN) and Com21 (CMTO) will be well positioned in the cable telephony space. Metro DWDM: this is a growth area that requires that the equipment be much cheaper than existing DWDM gear in long-distance networks. Uniphase Corp (UNPH) stands to benefit from the growth because while they don't make DWDM, they are a significant supplier to companies manufacturing the equipment. Broadband Wireless: this technology allows CLECs (MCI/Worldcom, WinStar, Teligent, Williams , etc) to bypass the local loop and not have to lease facilities from the BOCs. The players in this area include, P-COM (PCMS), Nortel (NT), and Stanford Telecom (STII), and California Microwave (CMIC). These companies stand to benefit from an international buildout as well.
Patrick Houghton: The most important applications driving the network are applications that increase bandwidth usage. E commerce, business intranets, video conferencing, interactive video games - all of these push more bits across the network thereby increasing the need for bandwidth. Telecommunications, after all, is really just moving bits from point A to point B and the appetite for bandwidth is almost unlimited. As the need for bandwidth expands, Tellabs will benefit at the core, Cisco at the edge and Uniphase will benefit as one of the significant players in building the blocks for the transport infrastructure.
Voice doesn't really use much bandwidth per minute of usage. Consequently, increases in voice minutes don't have much of an impact on total bandwidth utilization. What increases in voice traffic on data networks will do, however, is force convergence of the circuit and data communications worlds at a higher level - from the transport layer to the switching layer. There are no clear winners in the arena now, but Cisco, Lucent (through its acquisition of Ascend), and Nortel (through its acquisition of Bay Networks) have demonstrated the strategic leadership necessary to winning in this emerging market.
Briefing: Which stocks are you recommending and/or avoiding?
Seth Spalding: My current favorites are Uniphase Corp (UNPH), Cisco (CSCO), P-COM (PCMS), Corning Incorporated (GLW), and ANTEC Corporation (ANTC). Among the service providers, I like Global Crossing Ltd (GBLX) for their involvement in undersea fiber networks.
Patrick Houghton: My top pick at the moment is Tellabs (TLAB). I have a STRONG BUY on the stock. Relative to CSCO, I think that TLAB is undervalued. I also like CISCO (CSCO) very much but cut my rating from STRONG BUY to ACCUMULATE on price when the stock hit $112 per share. I think that you need to own CSCO if you don't own it, and if you do own it, hold on to it. I have a BUY on Newbridge Networks (NN) as one of the few independent ATM players and I have a BUY on Nortel (NT). |
