Crystal method: an example of Japan losing its way

by SR on July 5, 2012

Crystal device supplier revenues (million USD)

  • Japanese once dominated this sector. The other day I came across an article on TXC, a Taiwanese supplier of crystal components (electronic parts) and a company I visited quite some time ago. Crystal devices (oscillators, resonators, SAW filters etc) are tiny but vital parts of modern electronic equipment, including smartphones and tablet PCs. A decade ago, Japanese companies like Nihon Dempa and Toyo Communications (now part of Seiko Epson) dominated the global market and even lesser players such as Tokyo Denpa were considered noteworthy. I remember one Japanese supplier literally giving a polite laugh when I asked them about the competition in Asia. “Not credible competitors”, the IR spokesman assured me. Maybe the Japanese still have their much-vaunted technical edge but compared to Nihon Dempa, which is the largest independently listed crystal device supplier, Taiwanese rival TXC has grown far more quickly over the past decade. The chart above shows revenues for both firms. (NB: For Nihon Dempa we use its financial year, so ‘2011’ indicates the 12-month period that ended on 31 March 2012.)
  • Not long now before TXC overtakes. Strictly speaking, the Japanese are still dominant. In US dollar terms, TXC reckons it has already overtaken the crystal device operations of electronic parts giant Kyocera, which bought an independent supplier called Kinseki a decade ago. That puts TXC in the No.4 spot globally but still leaves Seiko Epson, Nihon Dempa and Daishinku ahead of TXC. If the Taiwanese company can maintain this differential in growth rates, it is only a matter of time before it overtakes the Japanese suppliers. To make that differential clear I put together the following chart, which shows the change in US-dollar based revenues for TXC and Nihon Dempa. The latter’s revenues rose from $396m in 2002 to $615m in 2011 – so just 1.55x higher. TXC’s revenues rose 6x (six times) over the same period, from $57m to $340m.

TXC and Nihon Dempa sales indexed from 2002

  • Ex-growth on a yen basis. In fact, it’s arguably worse than this, because on a native currency basis Nihon Dempa’s sales haven’t grown at all. Revenues were Y49.6bn in 2002 and were Y50.8bn in 2011, as the chart below shows. I think we need to look at US dollar revenues because the exchange rate has made Japanese products less competitive and thus it does have a bearing on TXC’s ability to grow more rapidly. Nevertheless, on a yen-denominated basis, the fact remains that Nihon Dempa’s revenue is where it was 9 years ago.

TXC and Nihon Dempa native currency sales indexed from 2002

  • All this plus better profitability too. The argument usually deployed by Japanese companies is that they dominate the top end of the market. This is often true, but what leaves investors scratching their heads is that Japanese suppliers often do not show better profitability. If they’re operating in the premium segment of the market, surely they should have higher margins? The reality is that the opposite is often the case. Certainly it is the case here. The chart below shows that TXC did not record a consolidated operating profit margin below 10% between 2002 and 2011, while Nihon Dempa has only had an operating margin of 10% or more for 2 out of those 10 years and the median is 5.9% compared to a median OPM of 15.1% for TXC. (NB: in the below chart, we cut off the y-axis at zero, but Nihon Dempa recorded a loss in 2008.)

TXC and Nihon Dempa operating profit margin (%)

  • Why did it happen? We don’t have time here to discuss in detail what went wrong for Nihon Dempa and why it has lost so much ground to overseas competitors. We could easily fill a book with the story of the fall from grace of the Japanese technology space over the past 20 years and the electronic parts sector has performed well relative to the semiconductor industry in Japan. So, briefly, here are a few factors that I think contributed to this relative decline.
    1. Shift in customer base. The growth in the customer base was coming from outside Japan rather than within Japan, which diluted the company’s ability to tackle key accounts for new business. Taiwan became a tremendously important part of the technology food-chain in the way that Japan became important in the 1980s. Naturally that favoured Taiwanese suppliers.
    2. Misguided focus on quality. Like many Japanese companies Nihon Dempa did not understand or chose not to believe that many foreign customers for electronic components want a part with “good enough” quality and a low price. If the cellphone is only going to be used for three years, a product engineered to last five years is more expensive than it needs to be. Quality is a cost and an excessive emphasis on quality will price you out of the market. That may be what happened to Japan’s crystal device makers.
    3. Leisurely overseas expansion. Nihon Dempa was slow to begin moving production overseas to reduce costs and when it did start the shift it initially focused on North America (putting a crystal growing facility in Illinois) rather than making a stronger commitment to Asia. The US was the easy option but given that manufacturing was in full flight from North America it was a questionable decision.
    4. Lack of dynamism. Frankly, Nihon Dempa and its peers in the crystal device industry come across as companies that are not particularly well-managed. The impression is that they’re a bit sleepy, a bit old-fashioned, a bit too focused on traditional Japanese approaches that are less effective when your customer base is rapidly diversifying. You just don’t get the same feeling of quiet competence from Nihon Dempa that you get from capacitor supplier Murata Manufacturing or connector manufacturer Hirose Electric. Those two companies are Japanese – polite, conservative, given to understatement – but both seem far more ‘together’ than the crystal suppliers.
  • Could the same thing happen to Japanese auto OEMs and parts suppliers? This is the big question. Without doubt the Japanese auto industry is globally competitive. With the rise of the Korean auto OEMs (Hyundai Motor and KIA) apparently from nowhere over the past few years, it is tempting to assume – as many Asian investors apparently have – that the Koreans will supplant the Japanese in some markets. I think it is unlikely for a number of reasons.
    1. China is “foreign” for everybody. In the case of the Japanese crystal parts suppliers, Taiwan became the workshop of the world for electronic products and in doing so gifted an extraordinary opportunity to local Taiwanese parts suppliers. In the case of the auto market, the largest and most rapidly growing customer base is in China, which is a foreign country for both the Japanese and the Koreans, not to mention the Europeans and the US. So there is more of a level playing field than there was in Taiwan.
    2. Quality is the over-riding issue. If your smartphone fails, you will be annoyed but it’s unlikely to do more than ruin your day. If you’re buying at the cheap end of the market you may in any case accept a device failure as the risk of buying low-end products. Conversely, if a crucial part of the vehicle you are sitting in breaks, you – and others – can be injured or killed in seconds. However inexpensive the car is, that’s an unacceptable outcome. You cannot make a car that has a high probability of failing suddenly and catastrophically after (say) five years and still expect people – at least in developed markets – to buy it. That’s why cheap Chinese cars don’t sell in Europe and the US. Even in quasi-developing markets like China, consumers understand what they’re not getting when they buy low-end cars.
    3. Move overseas nearly done. Japanese auto OEMs established production facilities outside Japan decades ago and have significant manufacturing assets in every major market. Given the large size of many automotive components – for example seats, engines and exhaust systems – and the “just-in-time” techniques of modern assembly, auto parts suppliers usually have their own facilities close by. This is one area in which the auto and auto parts industry is well ahead of the electronic parts industry. If a company lacks significant overseas production it is less likely, in my view, to be able to realise the need for genuine localisation. Real localisation means the use of local management who can guide local designers and engineers to develop new low-cost products suited for the local market, ideally using local parts and materials to reduce costs. In the case of China, no foreign company has fully achieved this goal yet, but the race is on.
    4. World class management. In the case of Toyota and Honda and many of the Toyota group parts suppliers, I don’t think you can credibly argue that these are not world-class companies. Their distinctly non-Western approach to issues of capital structure and shareholder returns tend to obscure their operational excellence. We shouldn’t forget that this is a cluster of companies that rose from the post-WWII ashes to take on the biggest auto companies in the world and beat them at their own game. They have something going for them. We will return to some of these topics in a future post.

Previous post:

Next post: