A crisis can, if exploited correctly, kick-start a new phase of growth. And the growth can be quite explosive where financial muscle, innovation, volume demand and national security interests intersect.
No, this isn’t a spy novel: it’s a summary of the dynamics at play in the semiconductor industry.
The semiconductor industry has changed due to a number of interrelated factors – a change in the demand driver, rising R&D costs, high customer concentration putting pressure on ASPs and creating a risk of disintermediation (if the OEM can design its own chips) – reducing profit margins and encouraging consolidation (in turn giving rise to a new set of tough challenges).
What makes an interesting situation more adrenaline-pumping is the Chinese government. China is the world’s factory, and is aiming to move its economy to more value-add activities, so that the label will no longer say Made in China but will become Made by China.
China accounts for 40-50% of global semiconductor demand. Many of those semiconductor devices are re-exported in end products, although Chinese consumers are becoming increasingly enthusiastic and sophisticated buyers of electronic products. Following China’s last semiconductor policy – let’s call it China Semicon v1.0 – China still imports around 90% of its semiconductors. A large proportion of these come from Taiwan.
If China sneezes, Taiwan gets a bad cold.
To get a sense of scale, in 2013 China spent more on importing semiconductors than it did on importing oil. To China, semiconductors are not just a strategic industry; they are a matter of national security.
Therefore, in June 2014, China recognized the shortcomings of China Semicon v1.0 and released a more market-oriented policy: China Semicon v2.0. This takes the existing threads of Semicon v1.0 (the national projects for technology development) and bolsters them with new provincial initiatives, greater financial muscle (from the National IC Fund, various government-sponsored vehicles and public-private partnerships) and a new sense of anti-monopoly activism. China Semicon v2.0 is also supported by Made in China 2025 introduced in June 2015, which targets improvements in manufacturing capability.
So, what’s happening? Chinese government policy and financial firepower is directly driving a lot of consolidation, through outright acquisition and minority investments, and indirectly encouraging Intel, Qualcomm and others to financially participate in one way or another. This activity makes for terrific headlines on newswires and keeps bankers, lawyers and advisors busy. It can also make for one or two Chinese players of considerable scale, potentially rivalling MediaTek and Qualcomm and diverting orders from them.
That is a considerable threat.
But what of the Taiwanese semiconductor industry, created to help propel Taiwan’s economic miracle and shift the export-led economy from being labour-intensive to technology-intensive? It’s about 135 miles from Hsinchu to Fuzhou, but at times, it feels like it might as well be 100,000 miles.
Taiwan’s Ministry of Economic Affairs (MOEA) has relaxed restrictions on Taiwanese semiconductor firms investing in China. But the MOEA maintains restrictions on Chinese investment in Taiwanese semiconductor firms. The ministry, seeing an existential threat, is motivated to protect trade secrets, IP and jobs. But a look at the global semiconductor ecosystem of 2015 shows how its tactics can be no more than a futile gesture.
Design IP is available from Imagination Technologies, Synopsys and others. Intel and Qualcomm’s participation in the China ecosystem will help improve the country’s semiconductor capabilities. The National IC Fund and other entities will acquire non-Taiwanese companies, assets and IPR. And each day, Chinese design engineers will climb the learning curve that their American, European, Japanese, Taiwanese and Korean counterparts have previously climbed. In this context, government restrictions in Taiwan not only engender a false sense of security, but they disconnect the Taiwanese semiconductor industry from the profound shift taking place in the #1 semiconductor market that happens to be next door and speaks the same language.
Of course, industrial policy is affected by geopolitical concerns, in this case the Taiwan Strait. Some on the mainland desire reunification, and see economic dependency as a tool to achieve that. Some on the island desire good relations with the mainland, with Taiwan as a recognised sovereign state.
Becoming a key part of the so-called Red supply chain allows Taiwanese semiconductor firms to reap China’s financial largesse. Practical measures can protect trade secrets, IP and jobs (as much as these can ever be protected). Restrictions can be eased further rather than being totally removed.
But the clock is ticking. Taiwan elects its next President in January 2016. The election date potentially narrows the window of opportunity for changing the MOEA’s investment restrictions. This opportunity must be balanced with the risk of Taiwanese firms being supplanted by Chinese competitors. It’s better to be in the fight than to be excluded.
In practical terms, Taiwan has little to fear from easing or removing the restrictions currently posed by MOEA. It takes much more than money to make things happen in the semiconductor industry. Tacit knowledge, management smarts and sustainable competitive strategies (not the price-based competition, which created a race to the bottom and damaged many Chinese fabless firms) will be key. China’s massive investment does not guarantee success – just ask any venture capitalist who has ever invested in semiconductors. A lot of money can buy a lot of fresh problems in terms of outstripping the available experienced management bandwidth. By co-opting Chinese investment, Taiwanese semiconductor firms can reduce the risk of being supplanted, while investing further in their sources of competitive advantage. This would turn a potential existential crisis into a new phase of growth and innovation.Follow @ImaginationTech