I don't think that we should be too quick to assume that a) the McDATA shares will be distributed according to the current ratio of McDATA Holdings and EMC; and, b) McDATA will be the same company with the same addressable markets in 6-12 months given the opportunities that will open up as the broad multi-year move from server-centricity to storage-centricity accelerates. Already, one can begin to see the kind of growth EMC is capable of generating with new products for its current installed base. McDATA's growth is inextricably linked to EMC's growth as well as IBM's own growth.
With a market capitalization of around $200 billion, EMC will eventually wrestle with the law of large numbers and that will affect its ability to recruit and retain scarce technical talent. With a market capitalization of around $9 billion, McDATA is the ideal vehicle for EMC to attract the talent that will keep its already robust growth going and to build shareholders value.
Already, we have seen EMC decide to keep its latest acquisition - Avalon (high performance archiving software) -- based in Denver. It just so happens that McDATA is flushed with about $380 to $400 million from the IPO and is in the process of buying a 90-acre parcel of land ($12M) and deciding on the appropriate vehicle for constructing a $50-$70M campus headquarters that they will most likely use to showcase fibre channel and how it enables novel networking architectures that employ the dynamic point-to-point connections of a SAN alongside a congestible general purpose LAN/WAN. I don't think the Avalon (current personnel: 10) will have any problems fitting into those plans especially since Avalon's best of breed software fits nicely into EMC's Rich Media initiatives.
McDATA also has a development lab in Toronto, Canada where Nortel's lead in optical networking is creating all sorts of research and start-up activity. They acquired this lab in 1996 from HWP and have continued to invest in it. That's interesting considering the white paper on optical networking and enterprise storage that EMC and Nortel recently published.
Colorado has a very strong storage industry talent pool that doesn't migrate easily to other parts of the country so McDATA allows EMC to deepen its presence there. Keep in mind that the top 3 executives at EMC - Egan (mid-70s Intel), Ruettgers (late-80s EMC) and Tucci (90's Wang) - all have major turnaround experiences so they're very familiar with the failure gene related to a company's ability to attract and retain talent.
The bottomline is that if EMC is a core position for you then McDATA is a core position as well, or at the very least, worthy of intense scrutiny as a starting point for other investing ideas. Some numbers to consider:
McDATA's Revenue Estimates :
IBM EMC Others
1994 $ 24M NA NA
1995 146M $ 0 $ 11.9M
1996 175M 0.5M 5.4M
1997 184M 1.3M 4.0M
1998 205M 5.6M 5.2M
1999 230M 48.1M 29.5M
1H2000 - 68.8M 26.5M
2H2000 * - -
2000 (est) $ 300M - -
* Major IBM mainframe upgrade cycle. IBM invented ESCON in 1991. McDATA was signed as IBM's exclusive ESCON (200 Mbps, 7 km) Director supplier in 1994. McDATA developed FICON (1 Gbps, 10-100km) bridge cards in 1998 with IBM as its exclusive customer. McDATA still controls over 85% of ESCON/FICON director market. Not surprisingly, McDATA controlled 99% of 1999 FC-based Director switch market. The ESCON installed base consists of more than 6,000 sites in 65 countries. Since 1994, McDATA has booked over $1 billion in revenues with over 800,000 ESCON/FICON ports shipped. That loosely translates into an average of 133 ports per proprietary ESCON site or an average lifetime revenue of about $1,250 per port. ESCON is proprietary and single-sourced. Fibre Channel is standards-based and multi-sourced. |
