Wednesday, January 24, 2007

Structural Separation and Golden Age of Network Centric Computing

[Here is a couple more of excellent reports from CIBC World Markets and Paul Budde's blog about why it is in interest of the telco's shareholders for them to be structurally separated companies in the last mile. The basic argument is the same - that new technologies such as web services, Web 2.0, P2P etc are allowing companies that are not linked to the network to offer the same services that were traditionally offered by the carriers including control and management of network equipment and facilities. Thanks to Timothy Horan and Paul Budde for these pointers. Some excerpts --- BSA]

CIBC World Markets Report

Paul Budde Blog

The industry food chain is now in place to enable Network Centric (NC) computing to become mainstream, or enter the golden age. NC computing is well described by Sun’s old adage that “The network is the computer.” • Among other things, new cost-effective technologies are making legacy telecom networks obsolete and enabling the broad convergence of applications in four separate industries—communications, computing, entertainment and consumer electronics. • In this network environment applications and services and are being provisioned by independent providers that are not linked to the underlying network. • NC computing is driving demand for basic communications services, but it is also rapidly changing legacy business models throughout the economy. • While growing vertical integration has been the industry trend in the last five years, which we believe should help earnings for the next few years, a shift to a more horizontally segmented industry structure is inevitable in the long term, in our view. • Emerging service providers are best positioned to take advantage of these trends. Incumbents will also likely adjust their business models to capitalize on these solid fundamentals.

With respect to the large telcos, we expect them to report good earnings growth over the next few years, partially aided by recent industry restructuring. While the long-term business fundamentals are uncertain, restructuring has helped pricing, improved service quality and lowered costs. However, to maximize long-term shareholder value, we believe integrated carriers should split themselves along business lines, such as transport-only wholesalers (selling to other service providers) and “customer-facing” service integrators (buying from other wholesalers to build a complete bundle). The alternative, in our view, is sub-optimal performance in one or both businesses.

If integrated carriers do end up separating their underlying “last mile” pipes from their services’ businesses, we believe it could be one of the greatest value maximizing events in the history of the telecom industry. This process will probably begin outside the U.S., because regulators cannot point to robust cable competition, and the only real way to generate long-term competition is through structural separation (horizontal segmentation), in our view.

Other than the last five years of vertical integration, the industry has been marching towards a more horizontally segmented structure for forty years—T (SO) spun out its PBX business, Local Exchange, Computing, Wireless and Equipment businesses. In this regard, we would note that the three major steps taken by the old AT&T towards horizontal segmentation were some of the greatest creations of shareholder value in its history: 1) allowing independent telcos to interconnect with its LD network in the 1920s; 2) allowing CPE equipment to be connected to its network, the 1968 Carter decision; and 3)spinning-out the RBOCs in 1984. These steps were all forced on the company by regulators, and interestingly all three were vehemently fought bymanagement. In many ways, this is exactly what the net neutrality principle is all about.

As we have been writing for the last seven years, we foresee an economy-wide shift to NC computing - driven by disruptive technologies. Technologies such as IP/Ethernet, softswitches, optronics, and wireless broadband, are driving traffic onto one multi-purpose IP network that enables new applications (e.g., IT to small businesses) to be purchased separately from network access (e.g., voice and video over IP). These technologies have also increased broadband speeds and reduced latency. Additionally, substantial improvements in computing power (Moore’s law), network security (authentication, intrusion detection, encryption, etc.), compression and higher layer protocols are setting the stage for the golden age of NC computing.

There are a lot of different names or segments of NC computing including the Internet itself, Grid Computing (which has quietly become very popular), Software as a Service (SaaS), Virtual Servers and Web 2.0. Regardless of the name used, it basically means networking together and sharing computer processing, storage and data, which can be accessed over the Internet using thin devices at the edge of the network. This increases demand for communications connectivity and boosts computer utilization and efficiencies.