Seven Questions and Comments for Japan's Semiconductor Industry

FUKUDA Hidetaka
Senior Visiting Fellow, RIETI

The bursting of the global IT bubble a year and a half ago has traumatized Japan's semiconductor industry, which not only leads the country's manufacturing sector but is literally its bread and butter. There is a deepening sense of crisis over the possibility of the semiconductor industry, which is thought to hold the highest promise for growth among Japan's knowledge-intensive industries, losing its competitive edge in the international marketplace. Could this happen or is it merely a groundless apprehension? To explore the issue further, I pose questions in seven selected areas to the top managers of Japan's electronic device manufacturers. To these questions, I have taken the liberty to append some views of my own.

1. Top Management Leadership

Does your semiconductor business occupy your thoughts around the clock? Are you in regular contact with the top management of major semiconductor users and of rival companies around the world? Undoubtedly, top managers of such companies as Intel, Infineon, TSMC and Xilinx are, without taking time to relax, absorbed in their semiconductor businesses seven days a week. What, after all, is required of top management is the kind of intensive leadership that concentrates on the semiconductor business 14 hours a day, 365 days a year; and, taking responsibility for all aspects of the business, is capable of making resolute decisions. In going about doing this, reliance on internal company systems has, I believe, its limitations.

2. High Earnings Model

In evaluating the performance of your semiconductor business, do you have as a benchmark a target ratio of profit to sales? Why aren't Japanese semiconductor makers able to achieve, even at best, a 20% profit rate on their sales? Why is it that top managers do not view seriously the stock market's appraisal of their company's business performance? Samsung has a profit margin of nearly 40%, and other leading overseas makers enjoy margins of around 30%. With them, not only the appraisal of top management but also of the market is given a direct bearing on operational decisions. That is, a key indicator that they use in evaluating the performance of their company is its ability to fluidly raise funds from capital markets.

3. Bold Investment Resolve

Amidst the radical vicissitudes of the semiconductor business, do you pride yourself on having the resolve to invest and divest with optimum timing? Why is it that you have not made an aggressive effort to attract capital from capital markets? If all you can do is make investments within depreciable limits, I doubt whether your semiconductor business will be able to survive international competition in the years ahead. Samsung, which aggressively procures capital from US capital markets, has by taking the lead in investment pulled off a victory in the DRAM market.

4. Semiconductor Manufacture Cost

Why is the cost of semiconductor manufacturing in Japan considerably higher than in Taiwan? Is it merely because of the difference in labor costs? Or might there be other factors involved?

There is a need to cut both sales overhead and production-related R&D costs. The Japanese doctrine of in-house technology development should be reconsidered in favor of a more efficient paradigm of jointly developing "pre-competitive" technology. Within a company's product line, it may be necessary to rethink the manufacture of those items that are of superfluous quality. For example, the cause of freezing in personal computers is not the DRAM, but the Windows software. An immediate task, therefore, will be to lower the procurement cost of DRAM materials. Such retrenchment also involves rethinking in-house technology development practices. Furthermore, small- and medium-scale factories scattered about outlying regions must be consolidated. As for models on how this can be done, one only needs to look at Korea's Hwaesong or to Taiwan's Xin Zhu and Tai Nan. This will unavoidably entail the adoption of occupationally differentiated wage scales and the possible introduction of stock options for employees. As you must have heard somewhere before, reform has no sanctuaries!

5. Technology Marketing

Are you aware of the significance of technology marketing? In the past, you regrettably failed to recognize the need to make a transition from mainframes to servers; and clinging to ATM technology, you were fatally slow in moving to router technology. Furthermore, abiding by DOCOMO, you went back on your commitment to non-proprietary GSM (Global System for Mobile communications). Having brought these and other bitter experiences upon yourselves, you need now to create a marketing strategy that is independent from your internal systems and departments, and to strengthen your cooperative relations with the world's frontrunners in your field. You should avoid getting tied to the Japanese market's first de Facto (ERP-solution software), as it is important for you to participate in the world's various application technologies.

6. Product Development Capability

Despite an absence of clamor over what it will be important for you to make next, why is it that you lack strategic products? You clearly need to discard the doctrine of doing everything across the board yourself. Rather, why not try to create an open design and manufacturing environment that utilizes knowledge and wisdom from around the world? By giving yourselves the option to form linkages with the world's more than 1,000 fabless companies (companies without a foundry that contract out production stages), you can dramatically increase your speed of product development. To develop large-scale strategic products will require both fluidity and flexibility in creating your marketing and development systems and in building your external collaboration frameworks. This will call for bold rethinking of your large corporate organizations, management schemes, and evaluation systems. In IT utilization including Internet use, Japanese companies are currently playing second fiddle to their counterparts in South Korea and Taiwan.

7. Industrial Reorganization and M&A

Is it your view that it won't be long before there are too many integrated device manufacturers (IDMs) in the semiconductor industry? It might be a good idea to seek the perspectives of the world's leading investment bankers and overseas investors on this topic. Needless to say, mergers and consolidations that only to serve to create a larger corporate scale will not improve profitability, neither will they be appraised positively by capital markets. A similar reorganization rationale is needed in the banking sector of guess what country?

For you sagacious top managers and semiconductor business executives, I believe the answers to the above questions are self-evident. Of course, I am resolved to receiving criticism and rebuttal from you regarding them and my opinions. However, when taking together such factors as the next peak of the silicon cycle and the entry of Chinese makers into the market, at the longest there is probably only 15 months left. That is, left to act, not wait any longer. You top managers should exercise the kind of responsibility and possess the sort of vision that will impel you to exhort the younger generations of technicians and engineers who will shoulder the future of Japan's semiconductor industry, saying "It must be you (!) who continue to make semiconductors of a quality that will ensure the place of Japanese manufacturers at the top of the global industry ten years hence." It's time to recapture the challenger spirit.

January 29, 2002

January 29, 2002

Article(s) by this author