To All:
Here's Kurlak's interview.
Search Archives:
Advanced Search
ÿSections
ÿNewsÿ
ÿStocks/Financeÿ
ÿWeb Developmentÿ
ÿProduct Reviewsÿ
ÿActiveX/Javaÿ
ÿCareersÿ
ÿTrade Showsÿ
ÿOpinionÿ
ÿHumorÿ
Advertisement
ÿSupport
ÿEncyclopediaÿ
ÿFeedbackÿ
ÿE-mail Newslettersÿ
ÿFeatured Sites
ÿNet Insiderÿ
ÿNetGuideÿ
ÿSharewareÿ
ÿGamesÿ
ÿHoroscopesÿ
ÿHomeÿ
ÿSite Mapÿ
ÿSearchÿ
ÿAd Infoÿ
ÿSponsor Indexÿ
ÿAffiliatesÿ
TechInvestor Home All Stories Mavens Stock Quotes
Newsletters & Links TechWeb 250 IPO Watch Company Releases
Get Stock Quote -- Enter Symbol ÿ
Q&A: Merrill Lynch's Thomas Kurlak
(10/24/97; 5:00 p.m. EDT)
Semiconductor Business News
Almost nothing these days is more important to a semiconductor company than keeping its stock price moving higher. A host of benefits -- from making affordable acquisitions to keeping key employees happy with stock options -- can come from a rising stock price. But whether any semiconductor stock goes up or down often depends a lot on what a small band of financial analysts is telling investors.
Certainly the most visible of this elite group of semiconductor industry specialists is Thomas Kurlak, first vice president at Merrill Lynch in New York. The veteran chip industry analyst has never been shy about calling market and company turns, but in recent weeks, he has actually become notorious. It all stems from his Aug. 13 note to clients slashing in half his 1998 estimates for Micron Technology. What bugged some investors was that little more than a week prior to his gloomy note, Kurlak had issued a 20 percent move up for semiconductor stocks.
The Merrill Lynch analyst is developing something of a reputation on Wall Street for breathtaking flip-flops, according to The Wall Street Journal. Even rumors about a change in recommendation from Kurlak can send a stock down, it said.
While some investors are getting fed up with his sudden shifts, many investors pay attention to Kurlak because of his reputation for bold semiconductor recommendations. Rob Lineback, editor-in-chief of Semiconductor Business News, talked to the veteran semiconductor analyst in mid-September to find out more about his views on the industry and what's new in evaluating chip stocks.
Has Wall Street fundamentally changed the way it views the semiconductor industry and chip company stocks?
I have been covering the semiconductor industry for about 20 years now. From what I've seen, there is much greater emphasis on the end-equipment markets [for semiconductors] and a lack of focus on the analysis of the industry's internal workings -- the manufacturing processes, the inventory cycles, the concept of yield, and all the individual things that contribute to making money in the semiconductor business.
There is much less microanalysis [among analysts] and much more attention to end markets, products, and the applications of chips. As a result, people are being surprised by sudden downturns caused by inventory corrections. This happened pretty hard in 1996, and it happened back in 1990 and 1991.
I tend to focus less on the end markets and concentrate on trying to get the cycle right, because stocks will follow earnings trends. What we are trying to do is to get in and out of the stocks at opportune times. We do it this way rather than make long-term fundamental cases about the applications of microcircuits, because I think those potentials are obvious to almost everyone.
What caused this lack of focus on semiconductor fundamentals and greater attention to be placed on end applications?
It was the result of a 10-year bull market sucking in a lot of new analysts, [less experienced ones] who concentrate more on the end markets for these products. But oftentimes, the end markets and the chip industry go off in different directions because chip companies build too many factories, or they don't build enough, or they don't know how to make a new process yield. These are the things involved in making money in the semiconductor industry.
They are the nuts and bolts, the blocking and tackling. That is what I'm interested in.
Many large U.S. chip companies have suggested the semiconductor industry has entered into an era of stability without the severe boom-bust cycles of the past decades. Do you buy that concept?
No. My whole purpose in being here is to predict the next [up or down cycle], and to look for the next opportunity to take profits or the next opportunity to make big purchases in these stocks. The inventory cycles are getting worse, not better.
Cyclicality is getting more pronounced. It is becoming more pronounced because the industry is becoming more capital intensive. So manufacturing is less flexible. It is less able to adjust to subtle changes in demand. This makes the industry less flexible, and then you have greater swings in inventory cycles, both up and down.
When the industry had downturns 15 years ago, there would be significant layoffs in Silicon Valley and in Texas. Real estate prices would drop and there would be a recession-type situation. Today, when companies face downturns, there are no layoffs because there are no people [due to greater factory automation]. The investment is all in machines and you cannot lay off a machine. You cannot get rid of the expense. The depreciation keeps getting charged to the companies. And when the demand drops and machines go idle, the expense is still there.
So earnings are more volatile [now] and the cycles are more severe. The 1996 cycle was the most severe since 1985. Semiconductor stock prices dropped 60 percent to 90 percent, and Companies went from record profits to losses in just 12 months. I don't think there is any evidence that we entered an era of stability.
Have there been any changes in the way semiconductor stocks are traded in relationship to the rest of the market?
I have not seen a change. They usually do well at the beginnings of bull markets and at the end of bull markets. I think this pattern will continue, but the current bull market has not ended yet, so it's hard to say. I would say there is more institutional ownership [of semiconductor stocks] now than there was 10 years ago. There is a greater belief in technology in general as a way to obtain growth than [there was] 10 years ago. And there is still a high degree of confidence that semiconductor stocks will be a primary way to get technology growth in a portfolio. Even with the 1996 correction, there is still a higher degree of confidence in these stocks on the part of the institutions.
With regard to Wall Street, is it easier or more difficult for semiconductor suppliers to distinguish themselves from other chip companies?
I think it is easier for them to differentiate themselves today because there are a lot more niches within the semiconductor industry. Companies no longer need to own a factory. People are able to tap foundries in Taiwan, so it is easier to start up a company. You don't have [to come up with] the money to build a fab. All you need is a good design.
Should Intel be treated as a special case from the rest of the chip industry?
It should as long as the monopoly continues.
When something goes wrong at Intel, the rest of the chip industry usually gets hammered on Wall Street. Is that fair?
It is sort of fair because Intel is so big and the product it sells to -- PCs -- absorbs a big chunk of the industry's output. So if Intel does something that causes a change in the PC world, it causes dislocations for other chip companies, even by accident. This happens all the time. Take the change in mix between the Pentium and Pentium MMX chips that occurred in the second quarter. Intel could not make enough MMX chips, but they had plenty of the old Pentium. The PC market did not want the old one.
So the end result was a slowdown in the PC industry during the second quarter until PC suppliers could get enough chips. The slowdown caused softness in semiconductor bookings until it worked itself out.
The U.S. capital equipment industry has been trying to move out the shadows of the semiconductor industry as far as the financial community is concerned. From Wall Street's point of view, is that possible?
I don't see that they can. They are in the shadows of the chip industry and they cannot do much about it. They sell all their products to the chip industry. If there is some other market for plasma etchers, I don't know where it is.
Capital equipment makers believe their business cycles are different from that of the chip industry.
No, I don't think that's true at all. If you look at the last [semiconductor] downturn, capital equipment orders went down and the [equipment] companies' earnings went down too. Their orders are going to follow the cash flow of their customers, and that will follow the [chip market] cycle. A number of those [capital equipment] companies had pretty sharp downturns; even Applied Materials -- the best of that industry -- had a downturn in earnings for several quarters. I don't think they have got back to year-ago earnings [levels] yet. I don't see any evidence that they are not associated with [chip market cycles]. They follow the trend and that's it.
Does Wall Street miss the SIA monthly book-to-bill reports?
Probably. I think it was a mistake to get rid of the book-to-bill, because we now have one less piece of information. I have learned to live without it, but it was another piece of information I thought was useful. The argument for getting rid of it was a poor one. Texas Instruments and other companies argued that [the book-to-bill] misrepresented the true industry trends. If you go back to the time they said that -- in late 1995 or early 1996 -- the perception then was that the U.S. was having a problem, but the world market was still in good shape.
But looking at the data now, with 20/20 hindsight, you can see sales trends then in the United States and the world were identical. They both went down together. It was a big downswing. Annualized growth of shipments went from positive 40 percent to negative 20 percent in about 10 months. You cannot possibly have a downturn in the United States and not have one in the world market because it is a commodity industry. It is like the price of oil plunging in Japan but nowhere else. It cannot happen. There is nothing wrong with a sample the size of the United States. This is where the most innovation is going on in this industry.
What are your most important measurement tools today?
They are proprietary [techniques] I have developed over the years. In general, I would say I have a healthy respect for the repetition of human nature and human psychology. In this business, the cycles are going to reoccur over and over again. I don't expect them to level out. Order lead time. This is an important variable. It's probably a very good indicator among the many you would take. It is like human blood pressure -- you don't know everything from it, but it gives you a good piece of information. The rate of change annualized on a monthly basis for sales is an important variable. We track a lot of them. Cycle time. The nature of how the chips are made produces the cycle because it takes 12-to-16 week cycle time to make a chip. On top of that, you have a highly capital-intensive business, which takes billions of dollars to build a factory. It is not a very responsive industry to changes in demand. When demand picks up, the industry is not able to immediately respond because it takes 12 to 16 weeks to get a chip out the door. And when demand turns down, the opposite is true. Inventories. Semiconductor customers want to work hand-to-mouth and have no inventory. So how do you marry a 12-to-16 week cycle time to a guy who wants just-in-time inventory control? Somebody is going to have to carry some inventory to make that work. Therein lies the problem. The inventory either gets too big or too small, depending upon how the leadtimes are going. This fundamental instability will never go away. We [stock analysts] key off that.
The trouble with semiconductor management is they look at [Wall Street analysts] as being the conduits of information and flowing [that information] to the public on the nature of their business, their success, and their potential for success. That's not our job. My job is to help people make money in these stocks. So we are concentrating on the things that move the earnings up and down. I let someone else be the information conduit for the management team.
Recently, you switched suddenly from positive to negative in your outlook. One example was your position on Micron Technology. How does that happen?
When you are taking a journey -- and you go in one direction and then another -- there is always a point in time when you have to make that turn. The point is to get to the end of the distention. We bought [Micron] at $20 with an objective of [reaching] $55 to $60. We got there and we got out. The whole idea is to buy low and sell high. It's not that complicated. The stock hit the price objective that we had set for it.
Secondly, and coincidental to the stock hitting our price objective, it became apparent that Micron was not going to be able to continue expanding earnings because of price weakness in DRAMs. They had pretty much [reached the end of their] string on cost reduction -- their die size shrinks on the old 16 megs. So that story pretty much ran its course.
We took a hike, declared victory, and went home. It was not throwing in the towel. That happens when the stocks go from $50 to $20 and you get out. That's a give-up. If we are doing our jobs, we would be getting people in at low prices and out at high prices in this industry. If I followed a different industry there might be some other approach, but in the [semiconductor] industry, you have to do that.
What is immediately ahead for semiconductor stocks?
I'm cautious. I think we are going to have [another] mini-slowdown, starting maybe [about] now and going into early or mid-1998. We are not recommending purchases of these stocks any longer. They are all up between 50 percent to 150 percent, so their valuations are quite high relative to what we see happening coming forward.
It looks like [chip revenue] momentum is beginning to flatten out on sequential sales, according to the SIA data. One of the reasons is the economy is not getting stronger. Normally when you come out of one of these semiconductor downturns, the U.S. economy has also had a recession and the industry is recovering at the same time the economy is recovering. We have just had a big inventory correction, and the U.S. economy did not go down. We have a snapback in bookings to those consumption levels. Those levels are not accelerating, so we are starting to roll over in line with consumption, which is growing at a modest pace.
I don't know of any end markets that are accelerating. There [also] are a few pieces of information indicating the forward momentum for the chip markets is not accelerating; if anything, it is probably decelerating. Motorola said cellular phones and pagers are soft. Altera and Xilinx are indicating the advanced computing and telecommunications markets have softened. Gateway is indicating they have too much PC inventory.
The other thing is there is excess chip-making capacity. So customers don't want to rebuild their chip inventories; they want to go hand-to-mouth. Consequently, we are not going to get this strong inventory rebuild phase you usually get in a chip cycle when capacity is used up and order lead times stretch out. I think the acceleration off the bottom [of the down cycle] last summer is rolling over, and we are saying [to investors]: Take profits in the stocks and wait for a better buying opportunity later -- say sometime next winter or next spring.
Stock Lookup ÿÿÿ Search Archives
Quote
1-Minute Chart
5-Minute Chart
Daily Chart
Weekly Chart
Monthly Chart
Ticker Search
ÿÿÿ
All CMP Publications
TechInvestor
|
