 |
 |
25 Sep 2006
Tom Hausken of Strategies Unlimited believes that the photonics industry is fundamentally fragmented. Here, he outlines the problems facing a market full of niche players and offers his opinions on the future.
How many times have you heard that the photonics industry needs to consolidate? Certainly you will have heard it more often than calls for further fragmentation. The call for consolidation is repeated so often, it is automatically taken for granted. Yet I wonder, what is so good about it?
Myths of economies of scale
Industry consolidation usually means reducing the number of suppliers as companies merge or exit the market. Revenues are spread over fewer suppliers, who spread their fixed costs over larger volumes. This means better margins, which in turn could make an unprofitable company profitable.
This would be just the solution for the telecoms components market, for example. There, revenues are growing but suppliers are still losing money, because their costs still exceed their revenues.
Unfortunately, however, much of the photonics industry is fundamentally fragmented. Even a small manufacturer is able to produce enough chips to feed the entire industry, but the volumes are simply too small to achieve any economies of scale. It is hard to make a profit if the market is made up of niches, but your company is positioned for high volumes.
This fragmentation actually works in favour of many companies that have reasonable expectations. They get by with careful control of their finances, providing specialized products to narrow markets or to partners, often charging high margins for them. There are many companies like this. Applied Optoelectronics, Hamamatsu Photonics, Lasertel, Barr Associates, DALSA, VLOC (II–VI) and OCP are just a few examples of survivors that operate using this approach.
The "standard" solution
If consolidation alone is not the answer, then what is? Some say that the answer is to generate standards that can reduce a supplier's manufacturing costs. Some or all of the savings would then be passed onto the customers, who (the thinking goes) could create new applications, generate more volume and so on. This isn't the case.
What happens is that the standard is the first step that allows the customer to commoditize the parts that do not differentiate their own systems. Customers love standards because they can play the suppliers off against each other, squeeze the margin from them and keep it for themselves.
This is what has happened in high-volume consumer applications. LEDs, diode lasers for CD and DVD players, image sensors for camera phones, and the killer application of them all – liquid crystal and other displays – are all manufactured in truly volume quantities. The very nature of these large-volume consumer markets means that the price of the basic device must be low, margins are thin and a few large players eventually dominate these segments, with much or all of their production in Asia.
But industrial applications are different. They are much more diverse, supporting a wide array of sophisticated and conservative customers, who assemble the parts for expensive capital equipment. Volumes are low and product cycles are long. These customers are interested in standards as a way to lower the prices that they pay for components, but they are also willing to pay for customized performance to differentiate their systems.
Myths of economies of scope
A few years ago there was a great deal of talk in the telecoms sector about the "one-stop shop". The idea was that a company that offered a broad product line could bundle customer orders and win a large part of the business. In return, some economies would be gained from consolidating sales and marketing. Customers would also benefit from having fewer suppliers to qualify and manage.
The trouble is that the products are diverse and technical. There may only be a few suppliers in any one niche that are shipping products. Each company can therefore proudly point to a few areas where it has leading market share. But there are so many small markets. It is unlikely that one company has the technical wherewithal to dominate many niches. It is as likely that there is a niche player that excels in each category and spoils the market for a more diversified company. Major shows like Photonics West and OFC are full of these niche survivors, who show up year in, year out.
What the niche player has to avoid is putting all of its eggs in one basket and spending carelessly. The large diversified player has deeper pockets and can recover from a few falls – the niche player cannot.
Fragmentation can be good
One can argue that the fragmented market has been good for many photonics companies, compared with the alternative. For example, the industrial laser segment is growing at a healthy 14% per year and the market has doubled since 1997. New applications appear that support steady growth. Prices decline gradually but steadily as a result of the normal maturingmarket process. The highly fragmented nature of the business means that suppliers can get reasonable margins. Consequently, it supports hundreds of players. For example, we know of over 20 suppliers that fabricate high-power laser diodes.
Could consolidation gain econ-omies to reduce the costs enough to increase volume in these markets substantially? In our surveys we have not identified a plausible example where it could, apart from in consumer applications. Instead, slow but steady wins the race.
Now consider the telecoms industry. I challenge anyone to name a telecoms application that is not happening because the price of the photonics components is too high. Of course, we have all heard that the components must come down in cost, but there is much more to it than that.
First, there is the cost of the electronics, software and labour (such as digging trenches and making truck rolls). Even more important is the uncertain demand for the services that photonics can enable. There must be demand for the services to justify the cost. Despite what they say "if you build it, they will come" it takes time for that demand to appear.
Slow path to consolidation
Over time, it is likely that the industry will drift towards some consolidation, as any maturing industry does. For example, the collapse of the telecoms market accelerated the consolidation of the carriers, and that consolidation is now reaching the system vendors. The recent merger of Alcatel and Lucent Technologies is a prominent example. One would expect the component suppliers to consolidate next. Up until now they have mostly consolidated only their internal operations.
The consolidation is likely to happen first along more standardized volume products. The larger players will then narrow the definition of the market to include only these prominent products. Or, they will diversify to greener pastures, mutate, or disappear altogether from the market.
Meanwhile, the many hopelessly low-volume products get marginalized to the niche players, many of whom will survive and even prosper from the fragmentation that remains fundamental to the industry. And that can be a good thing.
 |  |  |  |  |  |  |
| © 2024 SPIE Europe |
|
