Why settle for 1,000 chip companies when you can have 10,000?

There is a thread going around Chinese Technology social media circles now about a report that China now has almost 10,000 companies in the “semiconductor business”. A few months back we wrote a note admiring the fact that China had 1,300 chip companies up from almost none a decade ago. However, the rapid move to 10,000 is almost certainly something not so admirable.

Here is a report about this from a reasonable source (Google translate does a good enough job with this). This article takes a fairly negative tone on the matter, with the angle that we have seen peek semis investment. We found the statistics about the geography of these companies particularly interesting with many of the fastest growing provinces on the list….let’s say not known for their technology industries. The article also highlight the way that many non-tech companies have been piling in as well, another important warning sign. We could parse the number of companies going into high-end ‘design’ and ‘manufacture’ versus low-margin “packaging” and “distribution”, but even without this, the point is already clear.

How should we in the tech industry think about this?

The first thing to understand is that China has been investing heavily in semiconductors for a decade. The government has a comprehensive plan for reducing the country’s dependence on foreign (i.e. US) semis. In many ways that has been hugely successful, hence the 1,300 chip companies we noted earlier.

More recently, it seems that the government opened up the spigot. Our guess is that the central government began releasing funds for semis as part of their post-Covid stimulus plan, which explains the “sudden” appearance of so many entrants. We believe the best way to think of this as there being two programs in place – the pre-Covid plan and the post-Covid plan. The earlier plan was a targeted affair with funding channeled through a few dozen private equity/venture firms largely staffed by experienced semis professionals. The more recent plan looks like it is direct funding to companies without that professional intermediary screening layer. It looks like pretty much anyone with any claim on ‘semiconductors’ can tap into these latter funds.

In our experience in 30+ years doing business in China, these broad-based national programs often end up channeling most of their funds into unrelated projects, largely real estate deals. We are being diplomatic in our wording on this. Much of this investment does not generate productive returns, at least not around the targeted industry. (Anyone care to discuss China’s Big Data program?) We do not think the world, let alone China’s nascent industry, can support 1,300 chip companies let alone 10,000.

So does that mean we can just ignore all of these companies?
The short answer is no. As we said in our prior piece, most of these companies will fade away in a few years, a few will carve out some niche (autos, IoT and industrial systems seem likely candidates), but a small number (10?, 5? 2?) will emerge as global players. The problem is that with 10,000 companies competing for attention, our ability to discern the winners and losers becomes much more difficult.

Photo by Berto Macario on Unsplash

4 responses to “Why settle for 1,000 chip companies when you can have 10,000?

    • First, find a team. You don’t need a lot of people to get started. In addition to the normal skillsets, it will help to have a couple senior people who have worked in the US for well known chip companies.
      Then approach the funds – there are a few dozen PE funds in China doing chip work.
      Alternatively, track down some of the recent crop of “10,000” chip companies and pitch your services to them. More than a few of them probably like a great team and they can fund you.

  1. Pingback: Cognitive Dissonance and Semis Funding | DIGITS to DOLLARS·

  2. Pingback: Tsinghua Unigroup Defaults Widen | DIGITS to DOLLARS·

Leave a Reply