SI
SI
discoversearch

We've detected that you're using an ad content blocking browser plug-in or feature. Ads provide a critical source of revenue to the continued operation of Silicon Investor.  We ask that you disable ad blocking while on Silicon Investor in the best interests of our community.  If you are not using an ad blocker but are still receiving this message, make sure your browser's tracking protection is set to the 'standard' level.
Strategies & Market Trends : Technical analysis for shorts & longs -- Ignore unavailable to you. Want to Upgrade?


To: Clint E. who wrote (23436)9/4/1999 9:44:00 PM
From: drsvelte  Read Replies (1) | Respond to of 68145
 
Clint, great post!



To: Clint E. who wrote (23436)9/5/1999 12:15:00 AM
From: Clint E.  Respond to of 68145
 
----9/3/99 July SIA Report-----SUMMARY by SSB:----

*Semi shipments continue to improve. Earlier today, the Semiconductor
Industry Association released its July sales data showing worldwide
semiconductor shipments grew 19% year-over-year, another solid pickup from
June's 14% growth and May's 12% number. On a month-to-month basis,
worldwide shipments increased 3%, above of the industry's ten-year average
growth rate of 1%.*On a geographic basis, Asia-Pacific continued to
accelerate, growing 29% year-over-year in July, compared to 23% in June and
18% in May, while Japan improved substantially as well, growing 25%
year-over-year, up from 20% in June. The North American market also
accelerated further, growing 18% year-over-year in July, up from 12% in
June. The tail of the industry's recovery has finally begun to whip through
Europe as sales grew 6% year over year, compared to 2% in June.

------------ OPINION: ------------
Riding the industry wave. Though the data is somewhat of a trailing
indicator, it once again illustrates our thesis that the industry is in
the midst of a broad-based recovery. The breadth of the recovery
continued to spread as virtually every segment and sub-segment of the
industry showed positive year-over-year growth. Once again in July, the
Analog sector continued to improve, growing 13% year-over-year, up from
8% in May. This is a positive for National Semiconductor (NSM, 1H). The
PLD segment started the September quarter with a huge month of July,
growing 42% year-over-year, up from 28% growth in June. This is
consistent with business updates we have received from Xilinx (XLNX, 1H)
and Altera (ALTR, 1H) and reiterates our belief in both company's
ability to post solid sequential upticks in revenue. Non-volatile memory
remained extremely hot with Flash shipments growing 64% year-over-year,
up from 51% in June and SRAM growing 18% year-over-year, up from 13% in
June. This is consistent with patterns we have seen in the channel over
the last several months and is a positive for Integrated Device
Technology (IDTI, 1H) and STMicroelectronics (STM, 1H). The DSP sector
also posted solid 30% year-over-year growth in July, a positive for Texas
Instruments (TXN, 1M).

PC components show beginning of 2H bounce. July's SIA update showed that
the PC industry began its seasonal uptick a little earlier than usual
this year. Indeed, the MPU sector increased 17% year-over-year in July,
up from 8% in June and the DRAM sector jumped 56% year-over-year, up from
51% in June. This is consistent with our channel checks over the last
several months and confirms our belief that the PC sector is minimally in
a seasonal upturn and perhaps even accelerating, with much of that
marginal strength coming from overseas markets. As such, we reiterate our
belief in Intel (INTC, 1M) and Micron Technology (MU, 1S).
--------------------------------------------------------------------
--------------------------------------------------------------------
--------------------------------------------------------------------
--------------------------------------------------------------------
--------------------------------------------------------------------
--------------------------------------------------------------------

The Semi Beat: A weekly report tracking trends in the semi industry
Salomon Smith Barney
Friday, August 27, 1999

--SUMMARY:----Semiconductors *Demand for high-end processors was very strong last week in response to Intel's August 22nd price cuts. BX chipsets tightened further as Intel warned customers there would be no relief this quarter. *Trading of Celeron processors was firm last week, though the momentum appears to have shifted toward the Pentium family, which should help Intel ASPs this quarter. Interest in the K6-2 was stronger. *Pricing for bellwether 64Mb DRAM jumped 10% to about $8.15; inventories remain very tight. *The broad-based industry steamed ahead last week as Flash, SRAM and EEPROM memory remained very strong and power MOSFETS continued to percolate in the channel.
------------ OPINION: ------------
Demand for high-end processors picked up following the August 22nd price cuts. Our look at the processor landscape last week revealed a solid uptick in demand following Intel's (INTC, 2M) price cuts on August 22nd. The cuts, which averaged 27%, were on the high side of historical norms and dropped the average gray market discount to list pricing from 16% to 4%. Though the price action was scheduled for some time, it is also likely Intel was making a pre-emptive strike at AMD (AMD, 3H) before its upcoming launch of the new Athlon chip. This is a strategy that Intel pursued very successfully earlier this year with the Celeron. Intel may announce at this week's Developer Forum that the market's conversion from the Pentium II to P-III is about three months ahead of schedule and may even announce the end of P-II wafer starts. All of this positive activity at the high end should help Intel's ASPs this quarter. Most brokers continued to note strong gray market interest in the Athlon. We heard last week that Gateway should announce their support of the Athlon in the coming months, joining both IBM and Compaq as the lead tier-one OEMs to support AMD's high-end chip. However, a shortage of compatible motherboards continues to inhibit volume shipments. BX chipsets and motherboards remain very tight. Once again last week, Intel's P-II/III compatible 440BX chipset remained extremely tight in the spot, with some prices spiking to over $70, compared to a list price of about $26. We attribute the shortfall to a combination factors, including a stronger than forecast pickup in demand this fall, Intel's pushout of the Rambus-ready Camino 820, and lack of interest in the Pentium class, for the Whitney 810 chipset, which includes graphics. In addition, Intel was informing customers last week that there would be no improvement in the supply of BX chipsets in the third quarter, which could be a boost for SiS and VIA. Also, look for more confirmation of Intel's plans to support the PC-133 memory interface at the Forum, though management is expected to reiterate its long-term support for Rambus. Trading of low-end processors remained firm last week. We found demand for the Celeron family good last week as gray market prices remained flat. Broker sentiment was reasonably upbeat, though most recognize recent momentum has shifted to the Pentium III, especially following Intel's recent price moves. Intel appears to be gaining further market share as Emachines announced last week it would convert all of its SKUs to Celeron: $399 gets you a 366MHz Celeron system with a 4GB hard drive. Activity levels for AMD's K6-2 remained strong last week in the gray market. DRAM pricing jumped early last week and settled in for a solid week-over-week increase. DRAM spot pricing for the 64Mb jumped to the $8.30 range early last week and then settled in at $8.15 late in the week, up from $7.40 the previous week. Product remains very tight in the spot with most brokers forecasting pricing will remain north of $7.50 for some time. A few in the channel even alluded to a possible shortage developing over the next several months. Output, which has been weak in recent months should improve this fall, so further tightening of supply is unlikely. However, pricing should remain firm for several reasons: (1) OEM demand, which picked up strongly in early June, continues to strengthen, driven particularly by a strong Asian market, an early rebound in Europe, and solid U.S. corporate demand. Witness the recent tightness in PC components such as Intel's 440BX chipsets and compatible motherboards. The market has yet to hit its seasonal peak, which should come in late October. Even so, we believe the PC market may have entered an acceleration mode from the low-double digit growth generated through most of the late 1990s. (2) Many suppliers have had a tough time moving die shrinks from the 0.28 micron to 0.22 micron process. Hyundai, for example, reports unit output in the past couple of months has been flat because of lower yields at the smaller geometry. The company may not improve output until late this year. (3) Memory per box appears to be accelerating. On the high-end, 128Mb modules over the last few months have gotten tighter as many major OEMs have increased the memory per box from 64MB to 128MB and even 256MB. A recent survey of several "free PC" homepages found that, aside from the absolute cheapest offerings, all of these low-end machines now included 64MB of memory, up from 32MB several months ago. In addition, AMD's Athlon ramp should eat up proportionately more DRAM. (4) Inventories at all vendors remain very tight, which is not allowing for much buffer supply as demand picks up. Micron (MU, 1H) too is running short of product and we see no evidence of quarter end dumping by the company to adjust inventories. Though Micron only services about 30% of their output to the spot, we believe there is a positive bias to contract pricing. As such, we raised our estimates and price target last week from $90 to $110. Broad-based devices continue to build momentum. The semiconductor industry continues to improve as checks with brokers and distributors last week were met with a certain air of ebullience. Once again Flash memory led the way with many quoting 4Mb pricing in the $6.75 range, up from $4.25 just two months ago, and 16Mb pricing in the $18-20 range, up from $15 a month ago. High density Flash has remained extremely strong with evidence that 32Mb pricing has doubled over the last several weeks. The SRAM market also continues to percolate as more and more brokers speak of rising volume and pricing. Indeed, we have found some varieties of 1Mb SRAM pricing rising over 75% over the last few months. Despite the rebound, the segment has yet to attract much new capacity, which should result in firmer pricing for some time. The EEPROM market also picked up once again last week with several brokers noting pricing increases of roughly 40% over the last few months. Demand continues to rise for power MOSFETS as well. Similar to SRAM, the increasing prevalence over the last several weeks of MOSFET requirements in the spot channel bodes well for pricing going forward. We remain strong believers in Integrated Device Technology (IDT, 1H), STMicroelectronics (STM, 1M) and International Rectifier (IRF, 1H) on the heels of the industry's broader strength.
--------------------------------------------------------------------
--------------------------------------------------------------------
--------------------------------------------------------------------
--------------------------------------------------------------------
--------------------------------------------------------------------
--------------------------------------------------------------------

Perspective on Intel's Announcements
Salomon Smith Barney
Monday, August 30, 1999

--SUMMARY:----Communications Components * As part of the its multi-day Developers' Forum, Intel (INTC, 1M) will highlight a wide-ranging set of initiatives in the communications component space. Based on conversations with industry contacts, we expect Intel's ambitions will touch just about every communications market segment, although specifics will be lacking in many areas. * Since Intel's communications component efforts are unlikely to have a meaningful impact upon its own top-line growth anytime within the investment horizon, the main role of the communications thrust in the eyes of the investment community will be that of a potential 'spoiler'. Given the current evidence and Intel's past track record in diversification (e.g., graphics chipsets), we continue to see its impact upon other companies in the communications component market as relatively minor.
------------ OPINION: ------------
Some of the issues which are likely to be topical this week: The race is on: Porsche versus Maverick. Through its Level One Communications acquisition, Intel is putting the finishing touches on its Ethernet switch chip code named 'Porsche', although it is unclear whether Porsche is ready in time for the IDF. Combining 24 ports of Fast Ethernet with two ports of Gigabit Ethernet switch capacity (known as 24+2 in industry parlance), the Porsche chip will also include the Layer 2/3/4 intelligence needed in router-type equipment. In many respects, Porsche is identical to the Maverick BCM5600 StrataSwitch chip announced by Broadcom (BRCM, 2S) two weeks ago. While datacom OEMs we spoke with have not seen silicon from either vendor, the initial reaction is that both are interesting as point products but are not flexible enough to be used in multiple configurations. Both are being actively considered, however, particularly by second-tier or private label manufacturers such as D-Link and Accton. The merchant switching silicon market should be roughly $150 million in size next year, and the 24+2 sub-segment is probably about a third of that amount. We believe the relative success of Porsche versus Maverick will be largely defined by the rapidity with which production-worthy silicon (i.e., more than just sample devices) is available in the market. Based on recent history, we would give the edge to Broadcom. Internet Exchange Processor (IXP). Using the StrongARM processor technology acquired from DEC in 1997, Intel has developed a chip employing multiple processor cores for handling network protocols. By addressing in software many of the functions normally performed in hardware, the IXP is similar in concept to the Maker Communications (MAKR, 2S) MXT4400 and the architecture recently announced by C-Port, a private company. Since silicon and development tools are still not available, one communications original equipment manufacturer (OEM) we spoke with was taking a 'wait-and-see' attitude about the IXP. By any measure, we believe the 'network processor' market (including chips such as the IXP and MXT4400) is likely to remain a very small (albeit interesting) niche for the next several years. For example, we estimate that the MXT4400 (already sampling and designed-in with Lucent and Xylan/Alcatel) will generate only $3 million in sales for Maker in 2000. If successful, Intel could hope for similar volume in 2000-2001 since its development effort lags Maker's by around six-to-twelve months. Cornucopia of efforts. While its product presence in most communications chip markets is negligible, we expect Intel will not rule out any communications market opportunity at this point in time. Intel also has been continuing Level One's development of Gigabit Ethernet and dense Fast Ethernet transceivers, although the merger appears to have knocked those products off previously announced schedules. Level One was also a fading star in T/E and DSL chip markets, and Intel is hoping to revive its fortunes in those areas. Intel's recent purchase of Softcom Microsystems shows its commitment to the wide area networking (WAN) chip market, although its plans in WAN applications such as SONET and ATM remain very sketchy. In a somewhat surprising move, Intel is also expected to position the x86 architecture into network equipment. Given the extensive use of reduced instruction set computing (RISC) chips such as MIPS and PowerPC in network gear, Intel's push in this direction seems very late, in our opinion. Many of Intel's communications efforts such as wireless, home networking and enterprise networking gear are reliant upon outside vendors. Its new Fast Ethernet switch equipment, for example, is made by D-Link from a combination of Galileo Technology (GALT, 2S) switch fabrics and Broadcom transceivers. We expect Intel's equipment business will remain dependent upon third parties for the foreseeable future. Many questions remain. By folding multiple far-flung groups together such as DEC (StrongARM), Level One, and Softcom, Intel faces the challenges inherent in most mergers. For example, we have heard reports of infighting among various communications units as the internal reporting relationships and product roadmaps are developed. Despite Intel's efforts to depict a coherent master plan this week, the industry contacts we spoke with see the company's message as muddled. We continue to believe the number one success factor in the communications chip market is time-to-market - i.e., getting high caliber products to market ahead of the competition. Inherent attributes that Intel brings to the table such as brand name, balance sheet, and vertically-integrated manufacturing have little meaning when compared to TTM. Given the merger imbroglio and Intel's track record outside the PC microprocessor market, we remain somewhat skeptical about Intel's communications efforts.



To: Clint E. who wrote (23436)9/5/1999 2:26:00 PM
From: d. alexander  Read Replies (2) | Respond to of 68145
 
>>>I read those rules today and thought they were good.

Terrific; & thanks for posting. Also, the reference to President Roosevelt in #39 (roughly-don't let your liberal convictions cloud your market judgement) indicates they may have been written 55 or more years ago! The mm's were after stops then too <g>

My favorites (meaning the ones I should keep on top of my list) #4, 14, 40, 41, 47, 49, 50. Nos. 33 & 36 are new & glad to have them. (Patience is not my forte, but am working on it. So i don't mention those involving exercise of patience.)

I think you can develop a trading system around these earnings-less or low-float/volume stocks as long as you don't watch them tick by tick

Precisely, I do watch them tick by tick. So, I have to tweak that. And the approach outlined sounds good. Thank you. As someone posted recently - you are a wonderful teacher. Have never traded that way & would like to give it a try.

Also, examples from the recent past are very good to get. Like, here it was & you might have done this or that.

EFNT, specifically, I can't quite untangle.

It went public July 15; tripled 1st day then settled back & range traded, as you mention from Aug 4 to 24

The stock was trading between 35-41 for weeks with ~200k-300k average daily volume.

Oh, wait, I'm beginning to get that part of it. Yes,

if you had entered a limit buy GTC order @ 36 and once filled, a limit sell order @ 39, you would have made ~10% in the stock while the stock was range-bound.

Absolutely. The part that's confusing me is the parenthetical

(stop at 34, below recent spike-low)...

because the spike-low to 40.625 was 4 sessions ago, & how could that affect choices made earlier in August? Well, obviously, it doesn't, but there's some short-hand connection in your mind after the fact that isn't apparent.

Likewise,

in anticipation of breakout, you could have also left a limit buy-stop order at 43...

and done very well as it soared to 65 with limit-sells in place (that 2nd day 8/26, incidentally is exactly what i thot was happening to CCRD Thursday! famous last), but how does that connect with

the recent spike-high?

now, case-study ongoing. I might suspect another breakout, but it isn't exactly range-trading. A limit-buy at 52 might be a way to go, but i would probably just alarm the volume & Friday's high (purely hypothetical) & act in the event. It also has a small gap to (probably) fill because it wasn't breakaway on Friday.

Sorry this is so long!

I haven't read Nicholas Darvas "How I Made $2M.." Amazon doesn't have it, even in willing to search for out-of-print. Will do some further checking. I ordered the other one you mentioned, also 500 pages of Adam Smith.

Dorothy