By Jeff Ferry, Chief Economist
President Trump’s decision over the weekend to allow US component and chip companies to sell to Chinese network builder Huawei was a questionable decision.
We need independence from Huawei, not reliance on them as either a customer or network provider. Yet the president’s decision was an inevitable result of the structure of the global Internet infrastructure industry. The world does not have enough network providers today. More urgency is needed to fix this situation.
The telecom infrastructure business is a global $200 billion business and Huawei accounts for half of that market today. As we know, Huawei is a state-owned Chinese enterprise that built its business by stealing technology from Cisco, Motorola, and a host of other innovators. Even if today it is designing its own systems, evidence from Business Week and others shows old habits die hard. Huawei still enjoys stealing technology, especially from defenseless American entrepreneurs who lack the resources to fight back.
The problem is that the major customers in this market are large nationwide telecom companies—think Verizon, Deutsche Telekom, Vodafone, or NTT. They need at least three providers to create a competitive situation. They spend billions of dollars each year building their networks. Verizon, for example, has an annual network budget of around $17 billion. The way a telco typically spends that money is to award its network contract to two providers. This creates a competitive situation within the network footprint, with each provider motivated to do a good job to win a few more cities or regions in the network.
In the course of building a network, providers make dozens—even hundreds—of promises to their customers. These promises involve software fixes, hardware fixes, software and hardware upgrades, aggressive installation timetables, and other things. No provider keeps all its promises. Most providers keep most of their promises most of the time. But sometimes a provider slips up. In that case, the telco needs to be able to threaten to throw out the provider and replace him with a new one.
As an example, in April, Nokia reported that it missed its Q1 revenue target because it was unable to “recognize” 200 million euros ($230M) of revenue because it failed to deliver some 5G equipment to some customers. If that delay was all to one customer, Nokia might be looking at a “throw-out” situation today.
But with just two major western network providers, throwing one out is just not possible for telcos who want to avoid Chinese providers. The 5G wireless revolution illustrates the problem. World wireless infrastructure is a roughly $50 billion market today, and set to grow dramatically. Huawei is number one, there are two western companies, Ericsson and Nokia, in the next two slots, and then comes China’s rising provider, ZTE. A large number of America and European telcos are now using Nokia and Ericsson. They want a third provider, and for those in Europe that provider is Huawei. For many, Huawei is their main provider, with Nokia and Ericsson taking the role of secondary provider and the outsider that can be brought in if necessary.
So when Secretary of State Mike Pompeo or one of our national security officials wants to kick the Chinese providers out, the local telcos in Europe and Asia say: sorry, we need that third provider.
What’s missing is a US provider. That would solve this problem. It would also, of course, bring back American innovation and job creation at the network layer to our shores.
It’s important to understand that the lack of a US provider in this business is a fluke, an accident of history. Back in the 1990s, the US had three leading providers: Lucent, Nortel [Full Disclosure: I worked for Nortel for two years], and Motorola. All three disappeared partly because of their own mistakes and partly because Huawei, with its government-subsidized billions, began putting downward pricing pressure on the network industry, driving them out of business.
The problem is, to quote Stephane Teral, a leading wireless industry analyst, “The business case for 5G is murky at best.” The network providers are in the position of investing billions of dollars in developing systems today, and the telcos investing their own billions, in a technology that will not be profitable for five to ten years. All of my old colleagues in wireless networks are in the position of that man in the TV commercial, working on inventing the world’s first television in 1925, and his wife calls out to him from the bedroom: “Come to bed dear, there won’t be anything on for 25 years.” All these applications (remote surgery, self-driving vehicles) won’t be mainstream for at least five years, maybe more. All the hype from Verizon and AT&T about installing 5G today is all marketing. It’s not really 5G, nobody really needs it, and the telcos are spending only the minimum on it, just enough to fund the marketing.
Nokia and Ericsson are great companies. They are not small, with some $20 billion in revenue each. But a Finnish and a Swedish company do not have the resources to fund huge R&D projects as they need to be funded today. They can’t keep pace with the Chinese developers. In Shenzhen - whether it is Huawei, China Telecom, China Unicom, SMIC or Unisoc - you can hardly throw a rock without hitting the offices of a Chinese-government-supported technology company investing in 5G networks. The Chinese are investing probably $100 billion in 5G, all of it ultimately funded by the Chinese government. They intend to dominate the world in this technology, just as they dominate the world today in steel and aluminum.
Networks and the Internet are critical technologies for today and tomorrow. Everything from education to defense to transportation to health care to news to personal communications to your love life depends (or will depend) on networks. If the US views this as critical to its civilian and military future, it must invest to make sure it has the resources to build these networks without depending on an unfriendly nation like China—let alone a known IP thief, espionage enabler, and all-round, unsavory operator like Huawei.
The New York Times called recently for a “Manhattan Project” for US networks. As an old network guy, networks to me are not as glamorous as the world’s first atomic bomb or going to the moon. On the other hand, the innovation is not that colossal either. Most of the great innovations that make 5G possible were developed by US chip and component companies. We need a network provider to put them together into a world-class system, and scale up with the sort of reliability and resiliency that will make leading telcos trust them.
The federal government needs to work with existing network providers, and bring some money to the table, to encourage the creation of a US-based network provider focused on the wireless industry. Ericsson is a solid company, with a significant US R&D effort. Nokia inherited Lucent’s resources. Both of them are great companies and could be investors in this venture. Cisco is the US networking leader, one of the best-managed companies in Silicon Valley, and widely respected worldwide. When I was in the industry, Cisco sold to every single company in the Fortune 500. Cisco has avoided wireless because it reasons quite sensibly that doing business with telcos is less profitable than large corporations, and that the R&D cycles are too long and too expensive.
The US government already spends billions building networks for its own purposes, some well-known and some confidential. Now is the time for the federal government to step up and support the creation of a networking provider that will buy from US chip and component companies, build great networks, and deploy them worldwide. The presence of American partners like Cisco and others can ensure that the business is built on a profitable business plan. Previously, the federal government considered an option that would have involved working with the telcos on a government-supported 5G network. That was the wrong way to go. That would have plugged the 5G hole, but not solved the fundamental problem. The telcos are looking for cheap, reliable networks—they are not committed to building and maintaining an innovative industry.
The US needs to restore the US network layer. It needs to bring just enough funds to launch the business and keep it going for the first five years. It also needs to use its clout to make sure that there is enough profit between the demands of the extremely profitable chip companies and the extremely tightfisted telcos to make sure the network layer can make a profit.
That’s what a supply chain is. And when you have worked in a complex supply chain, as I have, you realize that the talk of “globally distributed supply chains” with products crossing borders dozens of times is nonsense if they involve countries that have their own agenda for global domination.
We need a US supply chain. We need it for 5G. We need it for the health of our own chip and component companies. We need it for American innovation. And we need it for national security.