Cross-border merger fever

After surveying the impact of the major events of 1997--large intra-industry mergers, pricing inflection, the siliconization of the data networking industry, and Cisco pulling farther ahead--I conclude that 1998 is likely to usher in yet another major restructuring of the networking industry. Most importantly, with market valuations of telecom equipment companies having risen dramatically over the last 12 months, many in the telecom space now have not only the strategic imperative, but the financial motivation, to be more acquisitive in the data networking industry.

The communications industry in general has gone through several waves of takeover activity, shown in the chart below, that have been defined by various product categories. Three years ago, it was switching technologies that were being scooped up. ATM properties followed. Then, a year ago, it was the desire to own remote access technologies that, in part, prompted the Ascend/Cascade and 3Com/US Robotics combinations.

Lately, companies with voice or telecommunications capabilities have been the favored acquisition targets.

Communications industry takeovers
Year	# of deals	Value (billions)
1997	42	$18.5
1996	42	$9.3
1995	30	$3.7
1994	10	$2.1
1993	10	$0.4

Most of the transactions followed a particular model and had certain characteristics. The majority were done for product breadth, in the hopes that synergy could be achieved via distribution. The model was simple: buy a small company with a hot product or a time-to-market advantage, and use your larger distribution channel to leverage sales by a factor of two to ten times what otherwise would be possible with a start-up. In order to take advantage of the larger distribution channel that already was tailored for data networking products, most transactions also were intra-industry. Examples of this include Bay/Rapid City or NetICs, Cisco/Grand Junction and Crescendo, 3Com/Chipcom, and BICC and Lucent/Prominet.

The next wave of growth is likely to break with the old model, in which distribution synergy was the primary motivation, and follow a model that is defined by market expansion, account control, and intellectual capital. With this in mind, what follows is my outline of factors that could result in cross-border transactions and some possible acquisition scenarios.

The urge to merge
Setting the stage for subsequent cross-border transactions in 1998 are three significant factors:

•  Lucent's handcuffs come off in 1998. Due to an IPO-related restriction that precludes Lucent from using the purchase accounting method until September 1998, the company has not yet consummated a "big" transaction. (In order to keep large amounts of goodwill off the books, virtually all large data communications transactions are done under the "pooling of interest" accounting method rather than the purchase accounting method.) While Lucent may or may not move quickly to pursue a large acquisition, an acquisition would be a logical step for a company seeking to expand its role in data networking. Perhaps more importantly, the mere threat of Lucent being unchained could spur a preemptive move by other large telecom companies that compete with Lucent. I believe there could be something of a domino effect if one of the large telecom companies make a significant move, such as purchasing Ascend, Bay, Cisco, Cabletron, or 3Com.

•  Telecom equipment valuations are at a relatively high level. For the first time in about three years, valuation levels for major telecom companies (such as Lucent, Northern Telecom, Alcatel, Ericsson, and Nokia) exceed the valuations of data networking companies. This reduces the dilution that telecom buyers would have to absorb, and alleviates one of the most common concerns that telecom companies have about acquiring data communications companies--asking price.

I have seen some evidence of cross-border merger fever already: Lucent has acquired three companies during the last year, Nokia has purchased Ipsilon, and Siemens/Newbridge jointly acquired RADNET. These transactions, however, are not large enough to impact the growth rate of the buyer, and are more of a "toe in the water" approach. I expect 1998 will be more of a big-splash year.

•  Data networking companies are vulnerable. With the exception of Cisco, every major networking company has missed expectations for one of its last four quarters, and expectations for all networking companies have been reduced as industry growth rates have moderated. Valuations have dropped accordingly. Bay, 3Com, and Cabletron currently trade at less than two times revenue, and Ascend's six-and-a-half times revenue multiple is only one-half to two-thirds of the company's historic levels. Combining these factors with the likelihood that data networking companies should recover at some point as pricing stabilizes and new product cycles begin, telecom companies likely will find it hard to pass up such valuations. If telecom companies truly believe the future is data (or data plus voice), they need these companies.

Where activity should occur
In the near term, as the telecom and datacom worlds collide, certain attributes are more valuable than others. The relative importance of these attributes, however, varies depending on the buyer. For example, telecom companies might have a stronger need for customer base than incumbent datacom companies seeking product breadth. The following attributes are particularly important:

•  WAN switching and/or access capabilities. Momentum in both the WAN switching and remote access markets already has begun to reemerge in 1998. Longer-term, the growth of the Internet, coupled with the circuit-to-packet transition, provides for some of the strongest growth attributes of any sector in the data networking space. Companies with relatively high exposure in these markets include Ascend, Shiva, and Newbridge, with Ascend being the most prized property.

•  Established enterprise customer base. While both Bay and Cabletron currently are having difficult financial times, both have large installed bases and strong data networking relationships that would be coveted by telecom suppliers looking for an entrée into the data networking business. Recent speculation that Nortel has made a bid for Bay only heightens the probability of consolidation.

•  ATM expertise. I expect ATM to be one of the fastest growth technologies in both WAN and LAN. ATM's demise, whether due to the threat from Gigabit Ethernet or to ATM's relatively high price, is greatly exaggerated. Most carriers, as well as many ISPs and CLECs, are moving to ATM at their core architectures, fueling strong usage of WAN. As for LAN, I believe there currently is no other technology that could be deployed as a stable, robust backbone, delivering quality of service. Companies with strong ATM products/expertise include Fore Systems, Xylan, and Newbridge, with Fore having the most expertise and Xylan having the most momentum.

Finally, I offer my thoughts on the role of the two largest companies--Cisco and Lucent--that are expected to shape the next wave of acquisition activity in the data networking space.

I have had numerous inquiries regarding the possibility of a Cisco/Lucent merger, and have scored the synergy for the combination high but the probability of such a merger actually taking place low. I look to several other scenarios as being more likely, and believe that the data networking companies with the most valuable near-term franchises--ones that could be blended with the franchise of a large telecom buyer--are Ascend, Bay, Xylan, and Fore.

Paul J. Weinstein is a director for Deutsche Morgan Grenfell. Based in San Francisco, he specializes in analyzing the data networking industry.