The Official Website of the Long-Term Investor
|
|
<COMMUNITY>
This special presents Dale Wettaufer's (MF Raleigh's) interview
of Texas Instruments' Chief Economist Vladi Catto. This probing interview
explores everything from projected worldwide growth in semiconductor demand
to fab equipment needs to the history of semiconductor revenue flucations.
Enjoy!
Dale Wettaufer (MF Raleigh) has brought you an interview with Texas Instruments Chief Economist Vladi Catto. This probing interview explores everything from projected worldwide growth in semiconductor demand to the history of semiconductor revenue flucations. We bring together the best of the Motley Fool and AOL's Personal Finance area to deliver this collection to you. Enjoy, Randy Befumo (MF Templar), a Fool
Dr. Catto (there's an accent egout above the "o") has been TI's Chief Economist since 1976. He was an international economist from 1973 to 1976, and a financial analyst in 1970 when he joined the company. He has been a visiting lecturer in International Finance at the University of Texas at Dallas and at UCLA. Dr Catto's publications include "A Suggested New Leading Indicator for the U.S. Economy," Business Economics (September 1979); "Market Share and Profits," Business Economics (May 1980); "Productivity and Growth," Business Economics (May 1979 and January 1980); "Technological Innovation: Key to Productivity," Research Management (July 1982); and "A Few Words on Windows," Business Economics (July 1993). He was a national award recipient for best professional paper on an economic subject in 1979. He assisted in the preparation of two Presidential Commissions' reports: one on the National Agenda for the Eighties in 1980 and the second on Industrial Competitiveness in 1985. Dr. Catto has served on the Texas Comptroller's Panel of Economic Advisors and on the Board of Directors of Head Start. Presently, he serves on the Dallas Morning News Board of Economists, is a technical consultant to the Business Council, and is a member of the board for the Child Care Group in Dallas. Dr. Catto lives in Plano with his wife, Judy. They have two children. Ph.D. Economics, University of Virginia. F. Dale Wettlaufer is from Buffalo, NY though he eats and sleeps in East
Aurora. He is a peripatetic student, working on a BA in History at this time.
A former broker, he manages a small amount of money and intends on growing
this as his main vocation in life. True to the legend of Sir Walter, he is
also interested in treading in the footsteps of his forefathers who walked
the goldfields of northern Ontario. In his spare time, he enjoys reading,
golfing, motorcycles, and critically studying the Reagan presidency.
DW: I was reading that you are going to be the moderator for the Economic Forecast Panel at the IC Expo later this year. In the article I read, you talked about the semiconductor industry being a $143 billion industry in 1995. Are you still on track with the 20% growth out to the end of the century?
VC: We are still forecasting 20% growth to the end of the century and actually beyond that. That would be the _average_ rate of growth on an annual basis, although clearly this year, 1996, is going to be a very weak year for the semiconductor industry, something that happened rather unexpectedly, I would say. We knew the industry would weaken; we didn't know it would weaken as much as it did. But even with the weakness in 1996, yes, the statement of 20% average growth into the future still holds.
DW: Was that a figure you were talking about in 1995 and before that also? Was 1995 an aberration for the industry?
VC: The 42% growth of the industry in 1995, yes, it was an aberration in the sense that it was well above trend, but it's not really unusual for the semiconductor industry to grow at rates between 30 and 40 percent. It did happen several times in the past. We have experienced now a very prolonged period of positive rates of growth for the industry, like 10 consecutive years of positive rates of growth, which is unprecedented and represents a demonstration of the fact that the industry is more stable now than it used to be. We still have probably a little bit too much cyclicality.
DW: Over the last ten consecutive years of positive growth, has it averaged above 20% or just about 20%?
VC: If you take the growth of the industry between 1985 and 1995, I believe the growth of the industry averaged almost exactly 20%. If you take the rate of growth of the industry over a longer period of time, like 1975 through 1995, then the growth of the industry averaged 15% per year. So, what we have seen, actually, is an acceleration in the average rate of growth in the more recent past.
DW: And that's just talking about sales growth and not unit growth.
VC: Yes, that's correct. That is growth of revenues, semiconductor dollars, semiconductor sales, semiconductor shipments, not semiconductor units.
DW: What sort of unit growth would that work out to be? Does that have relevance?
VC: Yes and no. In the sense that, to measure unit growth across the semiconductor industry, you would need to come up with a very accurate price index or price deflator for the industry, and that's something which has not been done yet. We don't have a good price index across all semiconductor devices. On the other hand, your question does acquire relevance if you focus, for example, on a specific segment of the semiconductor industry like the memory segment. In that segment, we do talk about the growth in bits, which is the equivalent of units, and there, the growth of semiconductor memory bits has been close to 70% per year. It's not unusual for bit growth to exceed 100% per year at times, but on average, 70% has been a good rate of growth. In fact, 70% is what we are experiencing this year, in 1996.
DW: How should the regular investor and the rest of the world outside of your industry look at the book-to-bill? In the last three months, it's dropped below unitary growth for the first time since, I think, 1985. To some people, that's somewhat startling, but if you look at the price factor in there, it's understandable.
VC: I think that too much emphasis , really, has been placed on the book-to-bill ratio, an indicator which really has not been very accurate in the past. It's not a leading indicator. At best, it's a coincident indicator of what's going on in the industry. The problem with the book-to-bill ratio is two or three-fold. One problem is, it's a ratio. You can never tell too much just by looking at the ratio of two numbers. For example, you may have a book-to-bill ratio which is going down simply because the denominator is growing faster than the numerator, but both of them are still growing. So, you can have bookings and billings both growing at a good rate, but the book-to-bill ratio is falling nonetheless simply because billings are growing faster. Well, is that significant? I don't know, maybe over the long term it can be significant but from month to month, it's not really all that significant. So, it's better to get away from the ratio concept perhaps, and look at the individual components, that is, bookings and billings separately.
Another issue has to do with, if you really want to look at the ratio, the fact that the book-to-bill ratio in general is only for the U.S.. I think if people want to continue to worship the book-to-bill ratio as an indicator of something, they probably would be better served by a global book-to-bill ratio. I think this what the Semiconductor Industry Association is trying to do. They're trying to come up with a measure that would be global rather than a regional, geographic measure of bookings and billings.
If you look at the book-to-bill statistically, measuring the correlation coefficient of the book-to-bill ratio relative to semiconductor shipments, the correlation coefficient is very low at 0.3 (the best correlation coefficient being 1.0). So, 0.3 means that, essentially, the book-to-bill ratio explains less than one-third of the fluctuations in semiconductor shipments. [over at least a decade of data, according to Dr. Catto].
DW: It's also somewhat of a conundrum that as the industry becomes more efficient at delivering the parts, that the ratio would drop. Does that explain some of the recent drop or is it almost all explained by DRAM pricing?
VC: Looking at 1996, the drop that we have observed is mostly explained by DRAM prices, yes. In and of itself, this drop has been unprecedented. DRAM prices have dropped from $12-13, looking at the 4 Meg, all the way to $3. That's a 75% drop in just a few months, and I've never seen that before. Interestingly enough, underlying demand is still quite good. The economy is doing fine, demand for computers is holding up, it's just that there was an accumulation of DRAM inventories toward the end of 1995, and then those inventories were liquidated all at once. They were just dumped on the market, so that really caused the problem. I guess what caused it, really, was the expectation of an unrealistically fast rate of growth in the computer industry. Like I said, the computer industry is doing fine, it's just not growing as fast as some people expected it to grow. DW: What do you think of market signals for people who are outside the industry? I know that within the industry, [DRAM pricing] information is fairly perfect, but for the capital markets, it seems like there's a sizeable time-lag between when pricing changes and when the market actually figures that out. When you're talking about such an important commodity, the information for outsiders is not that great--it has to come through an analyst somewhere, whereas with oil, gold, or orange juice, you can look at the daily trading of those commodities.
VC: I live in the industry and I deal with people who are familiar with the industry, so the people who surround me usually are people who are very well aware of what's happening to the book-to-bill ratio. I see the book-to-bill reported in publications which are not necessarily industry publications. You read about it in the Wall Street Journal, the Financial Times, and the New York Times. When the book-to-bill ratio comes out every month, it is widely communicated. The price of DRAM devices is not widely communicated, I grant you that; that's more difficult to find. If you're interested to know what it is, I guess you wouldn't have too much difficulty finding what the spot price of DRAM is. Do you?
DW: The information doesn't flow as readily as you might think.
VC: That comes as a surprise to me. I thought information was more widely available.
DW: From an economist's viewpoint, do you think there's a capital market inefficiency in the way the markets price companies like Texas Instruments, Micron, Applied Materials, and any other company involved in the IC industry?
VC: No, I really don't think so, despite what we just talked about, the fact that information concerning certain prices may not be free-flowing, or as free-flowing as I thought it was. I still think that the semiconductor industry, certainly one of the most global industries, anywhere you go, you find people interested in semiconductors. The industry is everywhere and the technology is at the cutting edge. Some countries would steal or kill to have their own semiconductor industry, and the fact is, that all of that is comprehended by capital markets when they price the companies in the industry. In fact, I think sometimes companies in the semiconductor industry are penalized by the fact that investors know, quote-unquote too much, about the industry; they know things that don't even exist. I think, for example, that the beta which is attached to semiconductor stocks right now is still way too high, and I think that P/E ratios are way too low because they still reflect a very high level of risk, which is not necessarily present. Indeed, as I look at the future, one of the major challenges for the industry will be to change the perception of investors and make them understand that risk is lower now than it used to be. We're not going to go through periods where the industry will grow by plus 40% one year and then minus 40% the next year. A plus 40 followed by a zero may be possible, but a plus 40 followed by a minus 40 is _highly_ unlikely. The betas and P/E ratios that currently exist unfortunately still reflect the structure of the industry as it was 10-15 years ago. The increased stability of the industry and the longer-term relationships are still not being fully understood or fully included in the financial ratios.
DW: So what you're saying, actually, is that there is inefficiency in the capital markets?
VC: I suppose there is [laughs]. Right.
DW: What I find interesting is, you look at alot of semiconductor makers and capital equipment suppliers to the industry, and they're trading at what a private-market acquirer would see as reasonable multiples. If you look across the whole span of the financial markets right now, that's not really common at all. I find it strange.
VC: It is strange and I think ultimately you're right. We still have inefficiencies in the capital markets with regard to a global industry like the semiconductor industry.
DW: I read an Economist article that cited VLSI Research's estimates of the industry's revenues, that 60% comes from the computer industry, 10% from mobile phones, and the rest of IC consumption coming from a diffuse lot of industries. Is that accurate?
VC: It's not inaccurate. If you look at the major customers of the semiconductor industry, clearly the computer industry continues to be the major customer. I would say maybe not 60%, but certainly close to 50% of all semiconductor devices end up in the computer industry. Communications, I think, is a little bit larger than the numbers you cite, probably 20 or 30%.
DW: When you say "communications," does that include networking equipment?
VC: Yes. What we're leaving out, essentially, is what we call industrial instrumentation and we're also leaving out consumer electronics. We're leaving out semiconductor devices that end up in your automobile, which incidentally, will continue to grow quite fast. Really, distinctions between computers and communications continue to blur. At the same time, it is possible that consumer electronics will absorb more and more semiconductors. We mentioned the automobile; I see the automobile as a major consumer of semiconductors in the future. I think for now that's at a relatively low level. In the future, for reasons having to do with safety, entertainment, or performance of the car, you will see more and more semis being used. You will have semiconductors that will keep the proper distance between you and the car in front of you, you'll have semiconductors that will tell you where to go and how to get there, you will have semiconductors that will regulate your tire pressure, and so forth.
DW: You've talked about U.S. electronics end-equipment sales being 6% of 1995 GDP, 3% of 1990, and 1% in 1970. Before I go on to the question, can you define these terms?
VC: What I am referring to there is the capital portion of electronic end-equipment, so it's computers, communications, and industrial electronics. So, then the total of those would be like 6% of GDP. The point I was making is that we're still at the beginning of the information age.
DW: We're talking about capital spending, not consumption [as defined in nation income accounting equations]?
VC: Correct. Software is not included. [This is treated, wrongly in the opinion of many economists, as consumption by the Dep't of Commerce]. The bottom line here is that we're just at the beginning of the Information Age, and in fact if you look at electronics as a percentage of GDP, we're still at 6%. My point, "how much further is there to go?" I was reminded that almost a century ago, railroads represented 15% of GDP. If railroads were 15% of GDP, then high-tech electronics, which are so much more pervasive, and to a large extent so much more useful than railroads, clearly we'll also someday get up to 15% of GDP. So, then, we're not even halfway there yet. When we think about the length of the information age, we should really compare it to the length of the Renaissance, which lasted 200 years, or the Industrial Revolution, which lasted 150 years. We have probably more than 100 years to go in terms of the information age, so we have only just begun. My focus now is "high-tech capital equipment," that is the high-tech portion of capital equipment, which is a little bit more limited than the 6% of GDP that I gave you there. In fact high-tech capital equipment is probably only 3-4% of U.S. GDP right now. So that more clearly shows that we have a long-way to go.
DW: However, quite significantly, in the capital accounts, high-tech capital spending has gone from, what?, 25% of non-fixed capital expenditures in the economy in 1980, to something like 50% now?
VC: About 50% is right. Eventually, high-tech will become 60-70% of the total, but it will take a while for it to get there. More traditional forms of capital spending, heavy machinery and so on, will continue to diminish as a percentage of the total. The trend is well established. If you look at the absolute level, if you look at the installed base of high-tech vs. more traditional forms of equipment, we are close to 50-50 right now.
DW: On semiconductor fab investment, has the rate of return on investment increased even if the cost of a fab has increased a thousand-fold since 1970?
VC: Generally speaking, rates of return have increased, really as the result of many different factors. One of those is the increased productivity of the industry, the other is cost reduction, and the third, is the lower cost of capital. Fact is, our cost of capital has now declined to a level which is comparable to Japan's, so the return on investment has increased as a result of that.
DW: It helps that the Nikkei is not trading at 60,000.
VC: Right.
DW: TI has instituted a number of joint ventures with Kobe Steel, Acer, H-P, Canon, and Hitachi, among other. Does that change the risk-reward mix for TI? Do you cap the upside while limiting the downside on something like that?
VC: Yes, as a result volatility is reduced as well as the risk. The joint-venture strategy, the way TI is doing it in the semiconductor industry, is something that will be very useful from the standpoint of not only risk reduction, but also from the standpoint of being able to finance the kind of expense required to build wafer fabs in the future. It's rather difficult to think that a company by itself would be able to finance a wafer fab that will cost $1.5 billion two or three years from now, and not too long from now, $2 billion. Joint ventures will be a way of life in the industry.
DW: Is Moore's Law, or the revised law, of chip densities doubling every 18 months, still in effect?
VC: The revised law is the one that holds right now, for the next decade or so. Eventually, you reach the point where the laws of physics take over or you'll have to manufacture semiconductors on the moon.
DW: Over the last couple of years, the cash flows have been so huge where the regular disinflation in some sectors of the industry has not been in effect. Since 1970, the supply-side effect of decreasing prices and increasing output has been very successful. Have the last couple of years, that cash flow priming, pushed the industry forward faster or changed the model in any way?
VC: We touched upon it when we talked about returns on investment. One thing that's happening in the semiconductor industry is that more of the value-added that used to accrue to computer manufacturers is now accruing to semiconductor manufacturers. The reason is that computer manufacturers are assuming the physiognomy that semiconductor manufacturers had a few years ago when semiconductors were known as commodity makers. The major differentiation will be in terms of price. Semiconductors, on the other hand, more and more are assuming the characteristics of tailor-made products. There is a higher value-added which is built into a semiconductor device. Sure, some portions of semiconductors, memory comes to mind, are still basically commodity products. But the rest of semiconductor devices, I'm talking about 70%, are really custom made. This has been a tremendous plus for the industry because it has become more profitable as a result. Even within the memory segment, you are beginning to see differentiation. For example you have the so-called synchronous memory chips which work best with certain microprocessors, because it works in synchronously with the speed of the microprocessor that you have in the computer. This has been really one of the primary changes in the industry. Money that used to go to the computer manufacturers now goes to semiconductor manufacturers because they are able to differentiate their products, moreso now than they could in the past.
DW: And also there's a substitution effect into the intellectual output of the software industry.
VC: Right. DW: Since much of the value-added in the cutting-edge processors is derived from intellectual inputs, does that favor the growth of natural resource-poor nations that have the engineering capital but not necessarily the financial resources or infrastructure to build the fabs?
VC: Possibly. In that sense, a country like India would have a great advantage over other countries because they have some of the best engineers. They still don't have wafer fabs, and many of those engineers that India has to offer are currently being used by companies like U.S. semiconductor companies. But they're also being used by software companies to perhaps not only manufacture or design software, but also to design tools that make the production of software more efficient. And they do all of this long-distance and can do this sort of job by satellite. If you have engineering talent then you certainly have an advantage, particularly when the kind of talent you have is relatively low-priced.
DW: Because of the importance of intellectual input, the irreplaceable nature of such an input, is that why TI is looking toward custom ICs such as digital signal processors?
VC: TI is looking at DSPs because they clearly represent the fastest-growing segment within the semiconductor industry and because they are so closely tied to improvements in communications, to the growth of the internet, to the growth of the networking industry. So, you expect DSP and DSP solutions to be the fastest-growing segment for a long time. You want to grow where growth is and of course, you want to go where profits and differentiation are, and DSPs offer all of that.
DW: Is this the result of telecom reform here and around the world, as well as privatization?
VC: I'm not an expert in that area, so all that I can say is yes, the more deregulation we have, the better it will be for private industry and really, for the consumer in general.
DW: Looking at the macro situation across the world, how are China, India, Russia and the eastern European nations, and Latin America going to play into the industry both on the demand and the supply sides?
VC: China appears to be well positioned, better positioned than other countries. If you just look at the major developing countries, China, Russia, and India, China probably has a lead on the other two simply because they are trying desperately to attract high-tech companies from all over the world. To a certain extent, they are succeeding in doing so. Intel is going to set up a wafer fab close to Shanghai in the next two years, Motorola has been in China for a long time, and TI eventually will be in China, so China is opening up to investment from abroad, something that India has not done quite yet, not to the same extent. India is still making it very difficult for foreign companies to operate there; there's more regulation and more bureaucracy in India than there is in China. Russia is probably in last place right now among those countries. Russia is sort of fumbling--I mean Russia doesn't have a stable political environment, economic environment, or really a strategy concerning what to do next. I personally am not too optimistic on Russia or eastern Europe in general. As I look at Asia, I see much more probability for success there.
DW: Those are all supply-side factors. What about the demand side?
VC: Based on my trips to China over the last decade or so, I think China has the edge there, too. This is partially because the Chinese want to catch up with the U.S. in terms of use of computers, but also because of cultural reasons, the Chinese parents, for example, care about the education of their children, they don't mind spending money on the purchase of a PC if it's going to be used by their children in school or at home. This kind of thing is not taking place in India or Russia at this time.
DW: Recently, an officer with AT&T's semiconductor division spoke about the depreciation schedules mandated by the IRS being restrictive to cash flow, relative to fiscal policy treatment in Japan, Taiwan, and Korea. Is fiscal policy restraining semiconductor industry cash-flow?
VC: To some extent. The depreciation issue is something the semiconductor industry pushed in the past, in the sense that, yes, we wanted to get the same kind of accelerated depreciation available to the Japanese. But ultimately, when the numbers were calculated, when the benefits of accelerated depreciation were calculated, I just don't think that we were talking about very significant benefits. Recently, there has been a bit of de-emphasis on the depreciation issue. Clearly, things like an investment tax credit, for example, that kind of policy action would have a very favorable impact on the semiconductor industry, or any industry that uses alot of capital equipment. In general, an investment tax credit would be more beneficial to us than accelerated depreciation. Along the same lines, a flat tax, the kind of tax that would allow for the deduction of money spent on capital equipment, would be beneficial to the semiconductor industry and high-tech industries.
In the past, we've talked about R & D tax credits being beneficial to the industry. But it would have to be the kind of R & D tax credit which goes well beyond what we have in now. Currently we have something equivalent to a 20% incremental R & D tax credit, so the additional money that you spend on R & D every year would benefit from a 20% tax credit. What would really benefit the industry would be an R & D tax credit on total R & D spending, as opposed to incremental spending. If we had a 25% tax credit on total R & D spending, of course the benefits to the industry would be significant, and that kind of R & D tax credit would be comparable to what's available to Japanese companies right now.
DW: Overall, do you think that the U.S. industry is at a competitive disadvantage to other countries in the realm of fiscal policy?
VC: I think so. We talked yesterday about the cost of capital, the gap between Japan and the U.S. closing, but that gap closing has nothing to do with fiscal policy changes in the U.S. There is alot that the U.S. government could do to boost spending on capital equipment in the U.S. Japan still has the most favorable fiscal-policy environment. In Korea, though, the government still literally subsidizes the industry, which is still true in Taiwan. Bottom line, purely from a tax standpoint, I think Japan is still the country that favors high-tech more than any other country. In a more comprehensive manner, if you look at everything government does as "fiscal policy," then Korea and Taiwan are still favored moreso than most other countries in the world in terms of assistance to high-tech industries. This is the kind of assistance that Chinese corporations will probably receive in the future. I'm sure that China will want to model the growth of its high-tech industries after Taiwan or Korea.
DW: Does the U.S. industry worry more about Japan these days or does it look over its shoulder more at Korea?
VC: It depends. Korea is now number one in the memory business, so if you happen to be a U.S. semiconductor company like Micron, which is totally in the memory business, then you'd better worry about the Koreans first. In the differentiated semiconductors markets, you'd worry more about the Japanese.
[On the issue of tariffs and quotas in the 1980s, I asked Dr. Catto about the supply-side effects causing DRAM price stability in the 1990s. He declined to give a specific chronology of the tariff and quota issues per se, but did have this to say about it:] VC: You know, what happened really, is that there was a huge semiconductor recession that started in October of 1984 and went on through 1985 and only in 1986 did we start to see the industry start to come back. As a result of that recession, many of the companies, particularly U.S. companies, but also Japanese, pulled back their capital investment projects and some simply cancelled their capital investment plans. So, it was the recession that caused the supply shortage in the following years; quite simply not enough wafer fabs were built to take care of demand. Events which you mentioned, concerning tariffs and so on, may have also been responsible for the supply shortage. But I still think that the recession had the primary role and tariffs played a secondary role.
DW: Are we in any danger of a semiconductor recession right now?
VC: I don't think so. It depends on how you segment the semiconductor industry. In the memory business, we may have supply in excess of supply over demand right now and obviously this is the reason why prices are falling. On the other hand, we have segments of the industry which continue to suffer, so to speak, from excess demand. Not enough supply is being manufactured with some devices. All in all, as you look at the next few years, the semiconductor industry will face more of a problem in terms of excess demand than in excess supply. Demand I think will continue to be strong, as a result of the emergence of new consumers like China, India, Russia, and eastern Europe, and also as a result of the fact that more and more semiconductors are being used in products that had previously not used them. Demand growth may remain close to 20% per year and it's not going to be easy for supply to grow at 20% per year in view of the fact that it's becoming more and more expensive to manufacture semiconductor devices. So we have to find the money to do so and we have to find the skilled people to do so. Many people fear that an excess supply situation will be a chronic problem for the industry. I fear the opposite perhaps. I fear that the industry may not be able to generate enough supply, although you will always run into situations, like the one we are experiencing in 1996, where a particular device experiences price drops because of excess supply.
DW: Technologically speaking, with the advent of 300 mm. wafers and other processes widening and deepening the cost curve, I was wondering if that will make the peaks of profitability and the dropouts in profitability any sharper. Will that increase the cyclicality of the industry?
VC: I have the opposite view. The semiconductor industry is made up of players who learn from past mistakes. The reason for cyclicality in the past has been partially the result of industry mistakes like excessive accumulation of inventories and lack of proper planning. Some of it has been due to some ruthlessness at times on the part of Japanese and Korean companies. The Japanese are beginning to behave in a way which pays more attention to profits and less to market share. Eventually the Koreans will behave in the same way. U.S. companies [semiconductor end-users] are engaging in longer-term relationships with their suppliers, which in a sense is mitigating some of the inventory problems which we've had in the past. The profitability itself of the semiconductor industry in the future will be more stable. If you make revenues more stable, then eventually you will also make profits more stable. Sure, you can still have changes in revenues in costs, but I'm a strong believer in the fact that geographic diversification and product diversification will make the semiconductor industry more stable in the future.
DW: Do you see this as somewhat of a triumph of the American capitalist system over the Japanese mercantilist-keiretsu system?
VC: Not the reduction in cyclicality. I think the reduction in the cost of capital in the U.S. may have been evidence of the fact that the keiretsu arrangement in Japan could not go on forever, and eventually we saw cost of capital increasing in Japan. I think it's evidence that the semiconductor industry is maturing and is becoming more diversified.
DW: I mention that because you spoke about the change in the way the Japanese do business. They're looking more toward profitability than building market and share and are getting away from their ruthlessness in pricing.
VC: They are, and they have to really. For one thing, it's because of the change in the value of the yen over the last decade, its strength. That in itself would really cause a change in behavior. That, in addition to their stock market weakness and the problems with the Japanese banking system, has brought on the death of the keiretsu arrangement and the birth of a bond market in Japan, which is beginning to look like a US-style bond market. Companies in Japan, instead of being able to borrow at insignificant interest rates simply because they are part of the keiretsu, now have to go out in the open market and pay the going rate of capital in the market. These structural changes are an indication that Japan is maturing. The country itself is developing the kind of financial system that has been with us in the U.S., and is also typical of European markets, for quite some time. It's also evidence that more and more of the stockholders in Japanese companies may be Westerners, and these are people that are interested in profits.
DW: Is the rebound in the Japanese economy good for the U.S. semiconductor industry?
VC: Yes, I think so. The Japanese consumers are becoming consumers of imports, and not just the conspicuous kinds of imports like Gucci clothing and shoes or BMWs and other brand-name goods. They're becoming consumers of any kind of imported good that has a decent quality and competes and compares favorably to the price of local goods. The Japanese are even buying Korean imports despite the acrimony that still exists between Japan and Korea. So, yes, the fact that Japan is recovering and consumer spending is gaining strength is a plus for the U.S. and any other country which exports to Japan.
DW: I just had one follow-up from yesterday. We talked about the book-to-bill yesterday; what do you use as forward-looking indicators, or what can the individual investor use as forward-looking indicators for the semiconductor industry, instead of the book-to-bill?
VC: I wouldn't throw away the book-to-bill completely. If we can arrive at a global book-to-bill ratio, something that will be available in a few months, then global book-to-bill indices may be used as an indicator of the health of the industry. Some additional indicators that may be used are bookings. Instead of looking at the ratio of bookings to billings, just look at bookings themselves, and perhaps a moving average of bookings. There are measures of backlog which also appear to provide a good indication of what will happen in the future for semiconductor shipments. There are clearly indications provided by the computer industry; there is a very high correlation between PC sales and semiconductor sales. Not from month to month, but over periods of time like one to two years. The correlation coefficient is very high between PC sales and semiconductor sales simply because most semis end up in PC units. Plus, you always want to look at the strength of the economy and the semiconductor capital equipment industry.
DW: Dr. Catto, thank you very much.
|
|||||
|
|||||||
