Is the Business of Displays a Good One? The Answer Depends on Your Position in the Supply Chain
by Paul Semenza
Many in the electronic-information-display industry take it for granted that displays are a key technology and that the strong growth of the market is an indication that it is a good business. After all, the global display market is now approaching $100 billion; in the past few years, liquid-crystal-display (LCD) monitors and flat-panel televisions have become highly sought-after technology products.
So, what to make of companies and investors that do not put money into promising display technologies, or that exit the business, during what seems to be such a strong market? Are they short-sighted or have they accurately concluded that the display business is not a good one in which to compete? Comparing display manufacturing with semiconductor production, and examining the supply chain in its entirety, helps yield some answers.
Moore's Law vs. the 100-in. Panel
The debate over involvement in display manufacturing has been going on the longest in the U.S., where it reaches back more than three decades. As with semiconductors, much of the early theoretical work and initial laboratory demonstrations of LCDs, plasma-display panels (PDPs), and electroluminescent (EL) displays was conducted in U.S.-based corporate R&D labs and universities. Unlike semiconductor manufacturers Intel Corp., Texas Instruments, Inc., and Freescale Semiconductor, Inc. (newly spun out from Motorola, Inc.), this activity resulted in very little in the way of volume manufacturing in the U.S. American companies including RCA, Westinghouse, and IBM made pioneering developments in flat-panel technologies, but did not build manufacturing capacity in the U.S.
In Europe, the situation has been similar. However, companies such as Philips and Thomson, which had active programs in display development and manufacturing, recently have nearly completed their divestitures of display-manufacturing operations. While these and other large European companies also have divested their semiconductor businesses – NXP Semiconductors (from Philips), STMicroelectronics (from Thomson), and Infineon Technologies AG (from Siemens), for example – they have remained active in Europe in terms of design and manufacturing.
So, why have Western companies remained competitive in semiconductor manufacturing but not in displays? One reason is the difference in the technological paths of the two industries. For semiconductors, Moore's Law – the observation that the density of transistors on integrated circuits doubles every 2 years – has been a constant for at least three decades. This means that competition has focused on the capability to reduce geometries and to use increasingly sophisticated design tools to manage the complexity.
Therefore, both the design and manufacture of semiconductors are high-value-added activities. At large integrated-device manufacturers such as Intel Corp., these two activities are handled inside a single corporate entity; at the same time, the emergence of semiconductor foundries, such as Taiwan Semiconductor Manufacturing Co., United Microelectronics Corp., and Chartered Semiconductor Manufacturing, Ltd., have enabled hundreds of fabless design firms to thrive at a very small scale. Furthermore, the rising transistor density has meant that manufacturers can increase revenues faster than they expand manufacturing capacity.
For flat-panel displays, since the circuit density is scaled to the human-visual system, the number of transistors has grown relatively slowly and is in the tens of millions per square meter vs. tens of millions per square centimeter in advanced semiconductors. At the same time, the substrate size for displays has grown much more rapidly. Whereas at the beginning of this decade, a Gen 4 thin-film-transistor liquid-crystal-display (TFT-LCD) fab handled substrates roughly three times the size of the standard 150-mm wafer; today, a Gen 7 substrate is 59 times the area of a 300-mm-wafer fab.
Since all displays convert data from electronic to visual forms, there is much less value ascribed to design, in the sense of creating the device's functionality. Rather, success is determined by prowess in manufacturing, i.e., raising yields and lowering costs – and, increasingly, the willingness to invest ever-larger sums at an increasing frequency. To complete the analogy, since the product designs are similar, the main way of increasing revenues is to expand manufacturing capacity.
This is perhaps the most important factor behind the exit of many companies from display manufacturing; with the shift from manufacturing to services among American and European technology companies, the notion of competing on the basis of capital expenditures has not been an attractive one. Some point to the rapid swings in profitability of display makers, the low level of cumulative profits, and the high level of capital costs in supporting such decisions. Others note that many of the largest display manufacturers view the display as a strategic technology that is too important a component to simply purchase. The evolution of smaller, more-focused manufacturers, particularly in Taiwan, can be viewed as a challenge to this business model.
Smile If You Have a Competitive Advantage
So far, this analysis has focused on the key activity of panel manufacturing. But another area of difference between displays and semiconductors is the higher level of complexity of the display supply chain. The high share of materials in the value of displays means that there are many upstream suppliers to panel makers, and the fact that displays are often the main part of systems, such as monitors and televisions, means that there are opportunities for the downstream enhancement and integration of displays.
In the case of upstream suppliers of materials and equipment, there tends to be a higher level of product differentiation, achieved through trade secrets and patents, and thus the competitive positions of the suppliers are more defensible than at the panel-manufacturing stage. In these cases, manufacturing prowess is important, but not always the most crucial determinant of success. In this sector, American companies such as Corning, Inc., and 3M, and European firms such as Merck & Co., Inc., have demonstrated staying power.
Looking downstream, companies such as Dell, Inc., and Hewlett-Packard Co. have been leading branded vendors in the desktop-monitor and notebook-PC markets. Retailers such as Best Buy and Wal-Mart have seen flat-panel displays contribute directly to increased profits. Many smaller companies have made good businesses out of adding value to reselling and installing displays in the medical, industrial, financial, signage, and home-theater markets worldwide.
Perhaps most notably, the wide availability of standardized flat panels from numerous suppliers, combined with reference designs and chipsets for digital TV, along with the eagerness of original design manufacturers (ODMs) to expand from assembling monitors and notebooks to putting together TVs, has enabled the entry of literally dozens of new players into what had been a fairly concentrated and vertically oriented market for branded TV sets. It is too soon to know whether this will be a lasting model, but it is suggestive of a fabless model in displays, although at the final product – rather than the component – stage.
The success of companies such as 3M at one end of the display supply chain and Dell at the other, while the display manufacturers struggle with high investments and uneven profitability, has led to the notion of the "smile" curve, in which profit levels, plotted from materials, through panel making, to end markets, trace out a smile shape. While certainly an oversimplification, it captures the notion that just as a smile is more attractive than a frown, the location in the display value chain can be a determining factor in the activeness of the business.
Paul Semenza is Vice President, Display and Consumer Research, at the market-research-firm iSuppli Corp., 2901 Tasman Dr., Santa Clara, CA 95054-1136; telephone 408/654-1720, e-mail: psemenza@isuppli. com.