ANCR looks like a great Spirit type play. Bounced hard off of 35 this morning. Competitor with BRCD. Maker of FC switches for SANs.
Some controversy whether this is truly a leading edge tech or a dead end, but it seems to be fast-growing. Very nice run at end of 99. Can't see why it wouldn't duplicate that run again... Has taken a haircut after disappointing on 1/10.
Best rundown is a Piper Jaffray report from November....
*** U.S. Bancorp Piper Jaffray *** U.S. Bancorp Piper Jaffray ***
Ancor Communications, Inc. (ANCR - 55 3/16) November 22, 1999
Ashok Kumar, CFA, 612-342-6461, akumar02@pjc.com
Paul H. Mansky, 612-342-6474, pmansky@pjc.com
Shifting Sans
Rating: Buy, Aggressive (#)
Price: $55 3/16 52-Week Range $73 3/8 - $2 9/16
FY End:December
Shares Out (Mil) 26.3
EPS 1998 1999E 2000E Market Cap (Mil) $1,451
Mar ($0.23) ($0.09)A ($0.02) Average Daily Vol (000) 756
Jun ($0.61) ($0.07)A $0.00 Book Value $2.68
Sept ($0.23) ($0.07)A $0.03 Dividend Nil
Dec ($0.28) ($0.05)E $0.07 Debt/Total Cap 0.2%
FY ($1.35) ($0.27)E $0.08 3-5 Year EPS Growth Rate 45%
P/E NM NM NM Mkt. Cap./FY99 Revenue 102x
Mkt. Cap./FY00 Revenue 32x
Revs ($Mil)1998 1999E 2000E 00 PE to Growth 15x
Mar $1.0 $1.5A $6.9
Jun $0.1 $3.6A $9.8
Sept $0.2 $4.0A $12.4
Dec $3.0 $5.0E $16.0 Quarter End December 31, 1999
FY $4.4 $14.1E $45.0 Reporting Date January 2000
*Fully taxed EPS.
Tables may not add due to rounding.
Highlights
The massive deployment of storage area networks (SANs) has been
instrumental in the explosive growth in Fibre Channel (FC) components. FC
is more of an extension to server storage than networking technology. Today
most of the FC deployments are at the front-end (host-to-subsystem) while
the back-end connection from the subsystem controller to the disc drive
remains UltraSCSI. This is because PCI bus bandwidth becomes a bottleneck
before limits of SCSI throughput are reached. But long-term, we expect FC
to displace SCSI at the back end as it offers better scalability,
manageability, and reliability. Once the interoperability issues at both
the hardware and SAN management levels are addressed, the SAN market should
explode.
A recent Gilder report indicated that the SAN is a fad and most if not all
the implementation will converge on GbE. GbE, which is a store and forward
architecture, exhibits 50x more latency than a FC switch. SAN is analogous
to building a superhighway network behind the network. E-commerce vendors
like E-Bay and Amazon are spending significant IT dollars on FC SAN versus
GbE. In FC implementation all the processing is in the hardware versus GbE
over IP, where the processing and decoding is in the software. Most of the
applications including data mining require virtual interfaces and cannot
tolerate the GbE latency.
In essence, the overhead caused by inserting a FC hub or switch in a data
path between server and storage subsystem is small compared to the overall
response time. A well-designed SAN topology will deliver the native
throughput of a storage subsystem. As such we forecast the FC switch
market, the largest opportunity in FC components, to grow from $150M this
year to over a billion
dollars by 2001 and exhibit a CAGR (98-03) of 160%. Forecasts by "seers"
predicting the demise of the SAN market have no better accuracy than
divining the future by reading animal entrails.
While ANCR was late in refocusing from LAN to SAN, on the strength of their
product offering they have established themselves as a credible competitor
to Brocade. ANCR would like to argue that their distributed memory
architecture offers lower latency and better scalability than Brocade's
central architecture. We believe that there is no measurable difference in
latencies (5-10% maximum) moreover, when the switches are cascaded, the
data does not stop at the F_port but flows through from one memory pool to
another. Bottom line is we expect Ancor and Brocade to control dominant
share in the mainstream switch market. The losers will be Vixel and
Gadzooks.
However, when we get to a director class product, 64-port solutions, a
single backplane (Ancor) is a superior solution to a cascade (Brocade). We
expect large installations to be aggregated on 64-port rather than 16 port
modules. Ancor recently announced that they will begin marketing Inrange's
64-port solution by summer. Furthermore the incumbent in this segment, EMC
(McData) has a director product that is a weak approximation to a FC
switch. It does not feature any E_ or FL_ports. Ancor is well positioned to
gain share at the high end. Furthermore, we expect the director class
product to represent over half the FC switch revenue market by
2003.
Fibre Channel Market
$M 1999E 2000E 2001E 2002E 2003E CAGR (98-
03)
FC $565 $1,075 $1,985 $2,680 $3,480 65%
HBA/NIC
FC Hubs $110 $190 $290 $395 $495 55%
FC $150 $355 $1,095 $3,020 $5,650 160%
Switches
Short-term, the biggest problem for Ancor is for Sun Microsystems to get
its act together. The issues converge around implementing the right adapter
board. Sun is adamant on using its own S-bus boards, which do not work
well. Also it would behoove Sun to utilize Emulex and JNI for the higher
end solutions rather then QLogic. Ancor's switches work well with Emulex
and JNI's boards. We believe, in aggregate, that these issues are
addressable.
FC SANs have barely penetrated the addressable compute and storage segment.
Switches, which feature the highest bandwidth/port will demonstrate the
fastest growth rate. A modest 15% share in 2001, would translate to $150M
which is 2x our current forecast for Ancor. Reference companies for
valuation trade at 15-65x projected CY00 revenues, with Ancor trading in
the middle. Applying a midpoint 40 multiple to the potential revenue
opportunity of $150M translates to a market cap of $6B. |
