Monday, January 15, 1996
MARKET CLOSE

INDEX:

I. Market News: Nasdaq Continues Trailing the Dow
II. Heroes: Checkfree, Maybelline, True North Comm., Rob't Mondavi, Oakley, Gemstar
III. Goats: Gen'l Acceptance, W.H. Brady, Watts Inds., Failure Grp., Best Buy, Softkey, Minn. Educational Computing, Internet Stocks
IV. Investment Perspective: Rick Whittington Feature, Part Three
V. Calendar: Tuesday's Economic Events

[Today's Issue of the Evening News concludes MF Templar's three-part series on the downgrade of semiconductor and semiconductor equipment stocks by SoundView Financial analyst Rick Whittington. If you missed Parts I or II, please hit the "Old News" button. If you found the recent series compelling and want to ensure that you never miss an issue of the News again, simply sign up for a subscription! The Evening News can now be delivered to any Internet mail address daily around 6:00 PM; simply go to Keyword:Foolmart and order a $26-per-year subscription. Or, for only a dime a day, give the gift of the Evening News to a friend or relative who hasn't become Foolish yet.]

MARKET CLOSE

DJIA: 5043.78 -17.34 (-0.34%)
S&P 500: 599.82 - 1.99 (-0.33%)
NASDAQ: 988.57 -19.66 (-1.95%)

MARKET NEWS

The bond market was closed for Dr. King's birthday, and in the absence of any government economic data, it looked like we were in for a quiet session on the stock markets. . . Wrong! In the middle of the afternoon, the bottom dropped out for semiconductor stocks again and the Nasdaq Composite picked up where it left off last week, dropping another 2 percent.

HEROES

Online Transaction Services concern Checkfree Corp. <% if gsSubBrand = "aolsnapshot" then Response.Write("(NASDAQ:CKFR)") else Response.Write("(NASDAQ:CKFR)") end if %> fought the downward trend today, rising $1 5/8 to $21 1/2 after it announced an acquisition. Checkfree will take control of privately-held Servantis Systems Holdings for between $165 and $173 million, with $43 million in cash and the balance in stock. Servantis Systems provides electronic banking and electronic commerce services to banks and corporations. Servantis posted revenues of $63 million in FY '95, which is about 50% more than Checkfree had for the same period. The combined companies suddenly make Checkfree's $663 million market capitalization look remotely reasonable---*only* six and a half times revenues as opposed to 17 times revenues.

Cosmetics giant Maybelline <% if gsSubBrand = "aolsnapshot" then Response.Write("(NYSE:MAY)") else Response.Write("(NYSE:MAY)") end if %> may have become the object of a bidding war between French consumer products king L'Oreal and German cosmetics and detergent company Joh. A. Benckiser GmbH. In the first territorial battle between Germany and France since World War II, L'Oreal might be forced to up its current $660 million bid to keep the American prize (Maybelline). Benckiser has said publicly that it is prepared to top L'Oreal's $36.75-per-share bid and is rumored to be considering a price in the $45 region. For all of this hubbub, Maybelline was up $3 1/4 to $40.

True North Communications <% if gsSubBrand = "aolsnapshot" then Response.Write("(NYSE:TNO)") else Response.Write("(NYSE:TNO)") end if %> started to head in that direction today, rising $1 1/8 to $18 5/8 when Merrill Lynch upgraded the stock from "neutral" to "above average." No other information was released from Merrill on its upgrade of the advertising and direct-marketing concern. Robert Mondavi Winery <% if gsSubBrand = "aolsnapshot" then Response.Write("(NASDAQ:MOND)") else Response.Write("(NASDAQ:MOND)") end if %> received a similar boost today, up $2 1/2 to $29 1/2 when Goldman Sachs raised its rating on the stock to "recommend" from "moderate outperform."

Oakley <% if gsSubBrand = "aolsnapshot" then Response.Write("(NYSE:OO)") else Response.Write("(NYSE:OO)") end if %> surprised the Street when it said that revenues rose to $43 million in the fourth quarter, up from $31.2 million in the previous quarter, with domestic sales up 33% in the same period. This boosted shares of the sunglass manufacturer $1 1/4 to $33 1/4. The company also announced it had an order backlog of $12.5 million as of December 31, 1995.

Shares of Gemstar International Group <% if gsSubBrand = "aolsnapshot" then Response.Write("(NASDAQ:GMSTF)") else Response.Write("(NASDAQ:GMSTF)") end if %>, maker of the "VCR Plus" remote control cassette recording systems, got a break today, rising $3 to $24 1/4. The company has been falling down since it hit a $43 post-Initial Public Offering (IPO) high a few weeks ago. The company called today's bounce a "technical correction" and a "non-event," stating that the company had gotten a little ahead of itself in the post-offering euphoria.

GOATS

William Blair dealt a near-fatal blow to General Acceptance Corp. <% if gsSubBrand = "aolsnapshot" then Response.Write("(NASDAQ:GACC)") else Response.Write("(NASDAQ:GACC)") end if %> today, causing the originator and reseller of auto installment loans to break down, falling $5 to $7. With General Acceptance ramping up loan-loss provisions while slowing down overall loan growth, Joseph LaManna of William Blair cut his estimates for 1995 to $0.73 a share from $0.95 and slashed his 1996 view to $1.00 from $1.40. The company has seen loan delinquencies rise from 1% at the end of 1994 to 2.4% at the end of 1995---a sizable gain, but within normal limits for a company that specializes in high- risk customers. After the market reacted to his cut, however, LaManna raised the company to a "strong buy." "At this level, I'd strongly recommend that people buy it,'' LaManna said. ''The company has stumbled, but I think they'll very much be back on track in 1996.'' Consumer Acceptance Corp. <% if gsSubBrand = "aolsnapshot" then Response.Write("(NASDAQ:CACC)") else Response.Write("(NASDAQ:CACC)") end if %> was down in sympathy with the news, losing $2 1/4 to close at $15 1/4.

Robert Baird pounded on W.H. Brady <% if gsSubBrand = "aolsnapshot" then Response.Write("(NASDAQ:BRCOA)") else Response.Write("(NASDAQ:BRCOA)") end if %>, pushing them down $6 to $21.The manufacturer of coated films and industrial-identification products refused to comment on the cut, only to say that they believe it was the reason for the stock being down today. Go figure. Kim Stulo of Baird cut her 1996 estimates to $1.23 per share from $1.43 and reduced next year's number to $1.72 over concerns that sales would be weaker than previously estimated. But what industrial product manufacturer W.H. Brady had an analyst do to it, its peer Watts Industries <% if gsSubBrand = "aolsnapshot" then Response.Write("(NYSE:WTS)") else Response.Write("(NYSE:WTS)") end if %> did to itself. The value manufacturer fell $2 7/8 to $19 3/4 today after it told the Street that earnings estimates for the second quarter would not be in the pipeline. Weaker-than-expected market conditions will bring the net down to $0.36 to $0.38 a share versus $0.38 in the year-ago period. Analysts had been expecting somewhere in the $0.42 range.

The Failure Group <% if gsSubBrand = "aolsnapshot" then Response.Write("(NASDAQ:FAIL)") else Response.Write("(NASDAQ:FAIL)") end if %> did just that today, sliding $1/1/4 to $5 1/8 when it reported that it would not meet analyst expectations for the fourth quarter. The company expects $0.02 to $0.04 a share on revenues of about $13.5 million compared to $0.10 on $15.4 million in revenues a year ago. Analysts were looking for $0.09. The Failure Group is an independent consulting firm that investigates, analyzes and prevents failures of an engineering or scientific nature.

Best Buy's <% if gsSubBrand = "aolsnapshot" then Response.Write("(NYSE:BBY)") else Response.Write("(NYSE:BBY)") end if %> inability to meet its old debt obligations are still weighing on the shares, tugging them $7/8 to $12 5/8 lower. Despite the fact that the company has worked out new agreements with two large suppliers, the fact that this has been achieved by simply postponing payment for 60 days has not assuaged the Street's concerns. Standard & Poor's credit-rating service has put Best Buy on CreditWatch with negative implications as a result of the company's rapid growth and ramp up of inventory levels which cannot be supported by its current cash flow.

The eroding value of Softkey International <% if gsSubBrand = "aolsnapshot" then Response.Write("(NASDAQ:SKEY)") else Response.Write("(NASDAQ:SKEY)") end if %> shares caused acquisition target Minnesota Educational Computing Corp. <% if gsSubBrand = "aolsnapshot" then Response.Write("(NASDAQ:MECC)") else Response.Write("(NASDAQ:MECC)") end if %> to slide in tandem, down $3 to $17 today. As the deal is a stock-for-stock swap, the plunging share price of Softkey, down $2 1/2 to $15 1/2 today, could potentially torpedo the deal. The exchange ratio was fixed with SoftKey trading between $35 and $45 per share, a range it left behind long ago.

The "Internet" was not a good place for investment dollars today. Internet Infrastructure (both hardware and software), Internet Service Providers (ISPs), Online Service Providers (OSPs), Online Transaction Providers and Venture-Capital firms all felt the Street's heat today. Hardware companies like Cisco Systems <% if gsSubBrand = "aolsnapshot" then Response.Write("(NASDAQ:CSCO)") else Response.Write("(NASDAQ:CSCO)") end if %>, down $3 3/8 to $67 1/4, Cascade Communications <% if gsSubBrand = "aolsnapshot" then Response.Write("(NASDAQ:CSCC)") else Response.Write("(NASDAQ:CSCC)") end if %>, losing $6 3/4 to $65 1/2 and Ascend Communications <% if gsSubBrand = "aolsnapshot" then Response.Write("(NASDAQ:ASND)") else Response.Write("(NASDAQ:ASND)") end if %>, falling $1 3/4 to $36 1/4, all had profits shaken free. The news that IBM and Sears are trying to dump Prodigy must have weighed on OSPs like America Online <% if gsSubBrand = "aolsnapshot" then Response.Write("(NASDAQ:AMER)") else Response.Write("(NASDAQ:AMER)") end if %>, which was off $3 7/8 to $32 7/8. The ever-so-questionable venture capital outfit CMG Information Systems <% if gsSubBrand = "aolsnapshot" then Response.Write("(NASDAQ:CMGI)") else Response.Write("(NASDAQ:CMGI)") end if %>, which sold America Online one mediocre browser, lost $19 to $60 today. MF Templar is currently at work on a multi-part series that will explore the prospects for this diverse bunch of companies. Look for it the week of the January 22nd-Febuary 2nd.

INVESTMENT PERSPECTIVE: Part Three of Three

When the Chips Are Down, Or Rick Whittington: Not a Chip Off the Old Block, but the Old Block Itself

Part III. What the Heck is a "Specialty Chip"?

"Profitless prosperity." These two words sum up Rick Whittington's vision of what is in store for the semiconductor industry over the next year. Lumping all of the semiconductor manufacturers in one large category without any regard for product mix or margins, Mr. Whittington has essentially doomed them all to the same fate---increasing revenues and flat profits, meaning decreasing margins. The spurious nature of talking about the semiconductor manufacturers en masse seems to escape the Street right now, however, plunging investors into the red in their cherished semiconductor positions.

After a day like today, it becomes increasingly important for investors in semiconductors to differentiate between semiconductor manufacturers. The technology sell-off driven by concerns over Intel's earnings (to be released tomorrow) as well as continued downgrades of former star Micron Technology (today from Robertson Stephens) demonstrate how much power the perceived profit streams of chip companies have on their market valuations. The reason why investors want to save their ammo for only the most specialty of specialty chip companies is because otherwise they will be caught in the up and down slam shares experienced by Integrated Silicon Solutions, Micron Technology or Integrated Device Technologies recently, and they won't be aware why it's happening. But what is a "specialty" chip company and how can you tell whether yours is or not?

Semiconductor manufacturers, like any other cyclical companies, can be classified into two groups---commodity and specialty. The reason why analysts ignore this distinction in their quips in the convention financial press is because what separates the commodity and the specialty chip maker is too complicated to gloss over in their one-paragraph analyses.

On Thursday I suggested one of the criteria might be analog and mixed signal chips versus digital chips in discussing National Semiconductor <% if gsSubBrand = "aolsnapshot" then Response.Write("(NYSE:NSM)") else Response.Write("(NYSE:NSM)") end if %>, a company whose fortunes Whittington has projected onto the entire industry. MF Chips and DonL0001 were quick to correct me on this point, both asserting that the distinction did not fall along analog and digital lines.

MF Chips wrote "I would argue that [National Semiconductor's] analog and mixed signal products are of two types. Some are quite proprietary and profitable. Others are less so. There are many simple analog parts such as amplifiers, converters, regulators, etc. that are true commodities. On the other hand, NSC's Ethernet mixed signal parts, optical interface chips, high performance converters, many telecom parts, are pretty proprietary. But it's a broad mix between proprietary and commodity."

DonL0001 offered one of the more concise explanations of the real difference between commodity and specialty chips: "I can see no justification for believing that analog and mixed-signal chips inherently have lower margins than digital products. Margins have more to do with how a company builds and markets its products and where its added value comes from, rather than what it builds. To the best of my knowledge, Linear Technology <% if gsSubBrand = "aolsnapshot" then Response.Write("(NASDAQ:LLTC)") else Response.Write("(NASDAQ:LLTC)") end if %> has the highest margins of any semiconductor company, and it produces nothing but analog and mixed signal chips, and supports them. (about 71%, 52% and 36% gross, pretax and net margins )."

The difference is that "NSM produces a lot of simple parts that LLTC would not consider." This goes to the core distinction---how a company adds value to the products that it builds. For non-commodity chips, MF Chips continued, "the contract is negotiated prior to the design being completed." He offered an extended example of the process: "Cisco Systems <% if gsSubBrand = "aolsnapshot" then Response.Write("(NASDAQ:CSCO)") else Response.Write("(NASDAQ:CSCO)") end if %> designs a new product. For each component, they carefully select a vendor. They may choose to develop some key ASIC's with VLSI and LSI Logic, a couple of nice communication chips from Seeq, Micro Linear, or whatever. These are competitively bid. There are usually a few select suppliers for each technology requirement."

This is a stark difference from how commodity chip makers are contracted. "DRAM and other commodities are selected by part number. The key to this selection is to be sure that each commodity has many sources. A given memory chip may have 20 suppliers. The component engineering group will then 'qualify' 3-5 suppliers' parts, after the contract pricing is done, and add them to a 'qualified' list." No commodity chip supplier is the sole supplier for any one product line. Much of the confusion as to the value of Micron Technology <% if gsSubBrand = "aolsnapshot" then Response.Write("(NYSE:MU)") else Response.Write("(NYSE:MU)") end if %> has rested on inadequate research by buyers that determined falsely that Micron was a dominant supplier of DRAM. Micron, by contrast, holds one of the smallest market shares and is stacked up against giants like Samsung, IBM and Texas Instruments. The only reason this commodity supplier has survived has been because it has focused on being the lowest-cost producer, making money at prices where competitors are breaking even.

The difference between specialty and commodity chip manufacturers comes out over time. "Let's assume that. . . VLSI Technology <% if gsSubBrand = "aolsnapshot" then Response.Write("(NASDAQ:VLSI)") else Response.Write("(NASDAQ:VLSI)") end if %> won a [contract to supply a] key Application Specific Integrated Circuit (ASIC) for $30 and co-develop[s] this chip with Cisco. [They contract] to buy a certain amount of DRAM at $25/Mbyte from 5 qualified [suppliers.] The supply of DRAM begins to become constrained. . . [and] prices are raised. Cisco pays [up] and struggles to get supply. [Companies like Micron] make huge profit gains. VLSI, enjoying stronger demand, does a little better, but is tied MORALLY to $30. It was negotiated, and as a proprietary product, they are bound to it for ASIC products." Right here is the difference between the specialty and commodity when demand tightens. Commodity suppliers, because of the very way they are contracted, are free to raise prices where specialty companies are obligated to honor their original price.

The trade-off, however, is that the specialty semiconductor manufacturer has a regular and predictable revenue stream whereas the commodity chip company has limited visibility of forward estimates. In MF Chips' example, "[t]wo years later supply loosens and DRAM starts dropping. [Micron's] profits wane. VLSI keeps chugging along at $30. The price does not fluctuate with 'supply/demand' like the commodities. Better yet, the proprietary designs booking in '95 were at higher margins than what was shipping in '95 (generally). These designs should start to ship in '96." The one-year delay is important because it means that specialty chip manufacturers can have big booms even when commodity chips profits are crumbling.

"The reality is that the price has increased for proprietary products. . . where Micron enjoyed instantaneous price increases, VLSI and others are licking their chops, awaiting ever-increasing higher margin programs to kick in [because of the one-year lag effect]," MF Chips concludes. The prices of "true" commodity chips vary daily whereas proprietary products "are negotiated for long periods of time," normally measured in years. True commodity chips are "pin-compatible," meaning they can be replaced by the same product from other manufacturers.

Some of the basic core logic audio, video and graphics chips are "de-facto" commodities, as they are similar in function and fairly easy to redesign to replace one another. The "de-facto" commodity chips have a short product life cycle "because the computer specifications are public, meaning that there is little room to differentiate and competition is fierce," asserts MF Chips. "So although Compaq <% if gsSubBrand = "aolsnapshot" then Response.Write("(NYSE:CPQ)") else Response.Write("(NYSE:CPQ)") end if %> may choose VLSI to do a design with, they can constantly shop for better pricing, threatening to design you back out in a moment's notice (3 months) if your pricing is not right where they want you." A peripheral product that has been relegated to commodity status as a result of all of this is the personal computer, according to MF Chips. "The PC has become so 'low margin' that there is no room for profits. Intel <% if gsSubBrand = "aolsnapshot" then Response.Write("(NASDAQ:INTC)") else Response.Write("(NASDAQ:INTC)") end if %> ate them all."

So where can you find non-commodity chip companies? MF Chips suggests the portable PC market. "I do believe the portable PC market is non-commodity still, but on the hairy edge. It still takes longer than 3 months to re-design a portable PC motherboard due to power management issues." He favors telecommunications applications due to their long design cycles, long life cycles and the possible differentiation. "[Telecommunications companies] will qualify additional DRAM sources but they won't switch key chips that would require software, firmware, and circuit board changes," meaning that the chip manufacturers profit stream is secure.

There is a "moat" that can be built around a proprietary product developed with the customer that requires supporting software and hardware. Those relationships cannot be erased even if some Taiwanese competitor has a couple of billion burning a whole in their pocket and the yen to build a fab. MF Chips sent a list of companies to consider as specialty chip makers as opposed to the commodity bunch that currently dominates investor speculation.

Carl Johnson at Infrastructure (http//:www.infras.com) sent a short list of semiconductor manufacturers that made the "specialty" cut: Analog Devices, Burr Brown and International Rectifier. "I have not considered these for our [Model Portfolio] but they are the companies that will not feel the pinch of a huge capacity expansion as much as memory makers, etc."

Johnson uses size as a distinguishing factor between commodity and specialty chip companies. "'Leading Edge' is. . . defined as manufacturing on 200mm wafers with minimum feature sizes below 0.35 micron" with 'Trailing Edge' being everything larger. Until recently, he had considered only smaller chips "Leading Edge" but has recently reconsidered because in some cases, "moving to smaller features is impossible because they are analog devices and the area of the capacitor is fixed---60-70% of the area of an analog device is capacitors. It [made] no sense to [require that they shrink] the other 30-40% of the device. Particularly when you would have to buy all new equipment to do it."

MF Chips has a much longer list of "specialty" chip makers. I would only add that you *ensure* that these companies make 75% or more specialty chips and that you consider the x586 and below a commodity at this juncture.

Analog and mixed signal: Exar, Burr Brown, Linear Tech, Analog Devices, Micro Linear, ICT

Communications: Seeq, Sierra Semi, Cirrus (PCSI), VLSI, ATT, Motorola, Texas Instruments, Level One, Brooktree, LSI Logic

Digital Signal Processing (DSP): Texas Instruments, AT&T, Analog Devices

Graphics: Chips & Technologies, S3, Trident, Cirrus, Tseng Labs, Brooktree

Programmable Gate Arrays (FPGA): Altera, Xilinx, Actel, Cypress

Application Specific Integrated Circuits (ASIC): LSI Logic, VLSI Tech

Audio: ESS Technology, Opti, AMD, Yamaha

Central Processing Units (CPU): Intel, AMD, Cyrix

Modems: Rockwell, Sierra

TV Encoding: Zilog

Digital Video (MPEG and transmission): CUBE, VLSI, LSI, SGS Thomson, IBM

NOTE: This three-part series would not have been possible without assistance from MF Chips (at Keyword:Sector, Semiconductors) and Carl Johnson and Ron Leckie at Infrastructure (at http//:www.infras.com). If you enjoyed this series, by all means check both spots out. We should have the Infrastructure Website linked in our own Semiconductors area in short order, along with a juicy list of other Websites to check out.

CALENDAR: Tuesday's Economic Events

---Mitsubishi Bank & Schroder Wertheim Weekly US Chain-Store Sales (9:00)
---November Housing Starts (1:30)
---Treasury Announces Results of 13- & 26-Week Bill Auction (about 1:30)
---Treasury Announces Size of 13- & 26-Week Bills (2:30)
---Johnson Redbook Weekly Survey of US Retail Sales (2:55)
---American Petroleum Institute's Weekly US Oil Statistics (after 4:00)
---Weekly Commercial/Industrial Loans (4:15), tentative

Byline: Randy Befumo (MF Templar)