To: Crossy who wrote (3157 ) 4/11/2002 7:35:32 PM From: Crossy Read Replies (2) | Respond to of 37387 re: New Pick - MLTC $ NasdaqNM: 1.95 (Multilink Technologies) Description: Multilink Technology Corp. (Multilink) designs, develops and markets advanced integrated circuits, modules and higher-level assemblies that enable next generation optical networking systems. The Company's products address the markets for DWDM and SONET/SDH optical transport equipment. Multilink focuses exclusively on the fastest commercially available speeds of OC-192, or 10 gigabits, or billions of bits per second, or higher and are in the early stages of developing products designed to address future systems that may operate at speeds of OC-768, or 40 gigabits per second. The Company seeks to be first to develop components that allow communications equipment manufactures to rapidly build and deliver high performance fiber optic systems more quickly and with more functionality and greater performance than their competitors. Historically, a relatively small number of customers have accounted for a majority of the Company's revenues. The Company's three largest customers accounted for approximately 73% of its revenues for the six months ended June 30, 2001, 73% of its revenues in 2000 and 74% of its revenues in 1999. The Company's top three customers for the six months ended June 30, 2001 were Alcatel, TyCom and Marconi, representing approximately 39%, 23% and 11% of the Company's revenues, respectively. The Company's top three customers in 2000 were Lucent, Alcatel and Cisco, representing approximately 34%, 28% and 11% of its revenues, respectively. The Company's top three customers in 1999 were Lucent, Alcatel and TyCom, representing approximately 36%, 20% and 18% of its revenues, respectively. The Company competes primarily against Agere, Applied Micro Circuits, Conexant, Giga (acquired by Intel), Infineon, JDS Uniphase, Maxim, Nortel (microelectronics division), NTT Electronics, Philips, PMC-Sierra and Vitesse. Notable: Company obtained a very advanced SiGe (Silicon Germanium) process in a codevelopment effort with IBM and recently conducted a similar agreement on InP with a German Upstart. Financial Data Price & Performance 52-Week Low on 1-Mar-2002 $1.88 Recent Price $1.95 52-Week High on 8-Aug-2001 $18.40 Daily Volume (3-month avg) 281.9K Daily Volume (10-day avg) 100.0K Performance relative to S&P500 -68.0% Share-Related Items Market Capitalization $146.6M Shares Outstanding 75.2M Float 20.3M Per-Share Data Book Value (mrq) $2.15 Earnings (ttm) -$0.28 Earnings (mrq) -$0.12 Sales (ttm) $2.32 Cash (mrq) $1.31 Valuation Ratios Price/Book (mrq) 0.91 Price/Earnings N/A Price/Sales (ttm) 0.84 Income Statements Sales (ttm) $130.5M EBITDA (ttm) -$12.9M Income available to common (ttm) -$12.4M Financial Strength Current Ratio (mrq) 3.97 Debt/Equity (mrq) 0.03 Total Cash (mrq) $92.7M Short Interest as of 8-Mar-2002 Shares Short 386.0K Percent of Float 1.9% Shares Short (Prior Month) 631.0K Short Ratio 0.74 Daily Volume 521.0K My take: This valuation is insane and under book. Optoelectronic whiz-kid shop MLTC on sale for PSR of 0.90. What a steal. Like a close out.. Many posters on the Yahoo board have been hammering the ULH/LH theme for MLTC. That should be more indicative of the past then the present. I noticed their most recent quarterly PR. They describe OC192 products for METRO AND Long Haul.. 2) Now "specialization" on the wrong niche was cited for MLTC indirectly implying MLTC "bet on the wrong horse". First of all the only viable market in the PAST for OC-192 (!) was long haul and ultra long haul particualrly submarine chips had extra premium (due to the even longer spans required). However you should see that MLTC uses SiGe for it 10Gbit chipsets (OC-192). That should be way cheaper to produce than InP or GaAs designs plus the fact that yield should be higher. I have all enough reasons to assume MLTC could sell to metrocore /normal long haul system designers at competitive pricing also and still making good prodits, just because they have an edge with their SiGe process.. 3) There was a nice hint in the recent quarterly PR on how MLTC is about to acquire InP technology from a tiny German outfit for OC-768. That's real frugal. Focus on design and try to bring advanced processes inhouse thru partnerships (like IBM for OC-192 SiGe). But OC-768 and >10Gbit/s speeds are at least 2 years off. Even on submarine. Advanced TDM equipment is way too expensive for commercial deployment to justify its cost yet.. 4) You guys here seem to forget the NEXT FORCE of "convergence" btw. datacom and telecom: Gigabit Ethernet. While telecom capex spending by carriers might be in the crapper literally, enterprise network spending by bigger firms for their company backbones is definitely not. Many recent PR's on 10Gbit (OC-192) has direct implciations thru applicability for Gigabit Ethernet also. MLTC potentially can supply to all transceiver manufacturers. Short haul interconnects will be important (even parallel optics need a SERDES chip 4x2.5Gbit= 10Gbit). So this will be a market for MLTC too. And don'T count out Line-card interfaces for routers. I think MLTC could supply this sector also or at least provide some chipset suppor to line card manufacturers.. 5) The latter brings another piece of insight: read Lightwave (recent issues) and you will see while total DWDM system level shipments (the "BOXES") are down from last year (20-40% as to statistic cited), line card shipments are up in volume (not in $$ though). MLTC supplies optoelectronic "boundary" and support chips so they are affected by the strength in the linecard segment not the "boxes" shipped. If you look at revenues, it's holding nicely at $20m+ per quarter and I don't think first quarter levels will be lower.. 6) Finally last quarter's numbers were impacted by unsually high R&D spending : $14m and a massive inventory writedown : $4m. If not for R&D (a DISCRETIONARY item) they would have been earnings positive even with the inventory charge. In fact annual R&D in 2001 was twice the Fy-2000 level at $55m. New process technologies (SiGe, InP) and a handful of metro & Gigabit Ethernet datacom products will be the result of these initiatives. 7) If you calculate a PRR - a Price to Research number you will arrive at an almost ABSURD valuation of a P/RR 3. Normally any figure under 10 is good even for a techstock. Ken Fisher was using this ratio alongside Price/Sales to find the next superstocks.. best rgrds CROSSY