Building blocks for the future
And also: the shape of our future digital-first lives (and jobs); the new techno-political order; the Second Cold War; hot topics for investors
Building blocks for the future: semiconductors and AI
The (new) golden age of semiconductors: Everyone was saying that data was the new oil. But now we see that the role will probably be taken by semiconductors, which are emerging as a strategic supply, without which our new ways of living and working won’t be possible
Chip manufacturing is key, and TSMC has a bright future. This week TSMC, the global leader in chip manufacturing, and absolutely dominant in the most advanced technologies, announced its 1Q21 results. And they were fully consistent with the central role that semiconductors have started to play: +17% revenue growth, to reach $12.7bn in the quarter, beating analysts estimates. Also, huge expectations for the next few years… (Bloomberg)
But they are going to need massive investments. To deliver on these high expectations, TSMC is planning an enormous investment ($100bn in 10 years), to increase production capacity, particularly for the most advanced designs (which also have higher margins). To put this into context, $10bn per year would be equivalent, with the current revenues, to approx. 25% CapEx to sales (WSJ)
A key problem is the enormous fragility of the chip supply chain. This is also reminiscent of oil in the 70s, and it is already disrupting the global economy, even starting to disrupt people’s ability to access the internet and communicate, a basic necessity in these darkish times. There are also uncertainties about when the industry will be able to add enough capacity to build some much-needed “basic” chips, because they have lower margins and are less attractive for companies like TSMC (Bloomberg)(Bloomberg2)
Using AI as a basic infrastructure: AI is becoming a central component in many collective decisions. And we expect that it will eventually be at the basis of all business and government processes
As a consequence, concerns are growing about risks of algorithmic bias or unfairness, specially dangerous because they could hide behind the aura of “data-centricity” or “100% rational decision-making” that people tend to attribute to algorithms. A discussion at the FT this week proposes some potential solutions to “counter” these bias problems: (1) Increasing diversity in the algorithm-writing teams; (2) Deploying AI only in contexts where they have demonstrable benefits; (3) Embedding “ethical thinking” into the entire design process for AI systems; (4) Submitting data sets and algorithms to independent scrutiny; (5) Creating a “regulatory approval” process for new algorithms, similar to the one regulating new pharmaceutical drugs (FT)
Snapshots of a digital-first life
With the pandemic, people’s life and work experiences have become digital-first. It is not that physical interactions have disappeared (they won’t) but they have become less frequent, at least for some segments. And the trend could remain with us. Many things in our lives are already being done within applications, and it may happen that in the future, if we want a more “physical” (vs. digital) experience, we’ll have to pay a premium
The pandemic has driven record sales of digital devices and electric cars: This week we’ve learned that digital devices (Samsung) and electric cars (Tesla) are announcing record sales, in a sign of how consumer trends have changed with the pandemic, and with lives becoming “digital-first” and environmentally conscious (FT)(Bloomberg)(Bloomberg2)
Many people now “go to the gym” through an application, and many people will keep doing this: The sustainability of home fitness and the companies enabling it, like Peloton, has recently been questioned. But it looks like it’s here to stay. The return to normality will be slow, many gyms that have opened are still subject to restrictions. And the price of an average digital membership is around -90% cheaper than a traditional one (WSJ)
Clubhouse seems to be consolidating as a mass market app, driven by “lockdown boredom”. Membership doubled during 2020, and this has apparently generated the interest of Twitter, which is supposed to have offered $4bn to acquire the company. After having rejected the offer, Clubhouse is now looking for funding at that same valuation. Meanwhile, the FT argues that growth is creating new challenges, with the app becoming “less clubby”, the meeting rooms becoming less intimate, and the overall experience becoming closer to “listening to a conference” (Bloomberg)(Bloomberg2)(FT)
Central banks all over the world are exploring digital currencies, which would be backed by national reserves, rather than by each customer’s bank account (as the current, “private” versions of digital money). Behind all this, a sort of prisoner’s dilemma, because if one country launches there is the risk that citizens from other countries will adopt the foreign digital currency in detriment of their national print money. Also, sovereign digital currencies could threaten current digital payments companies (like PayPal) (Bloomberg)
Blockchain could revolutionize how equities trading is done, shortening the transaction settlement times and reducing the costs and the risks involved in the existing process, like the use of collateral that remains in a “limbo”, with uncertain ownership, while each transaction is being settled (FT)
A new techno-political order
Privacy
This month Apple is expected to release its (long-expected) new privacy policies. So stress is growing in the digital advertising industry, with both large and small advertisers now complaining or actively working to reduce the impact:
Procter & Gamble has been testing the Chinese “anti-Apple” technology. We’ve already discussed CAID, the technology to track individual users that has been developed in China, and that is able to bypass Apple’s privacy controls. It seems that P&G has been testing CAID, as a way to keep targeting programmatic ads (WSJ)
On the other side of the spectrum, some small businesses are complaining about the potential implications of the new policies. The basic claim is that there is a trade-off between privacy and personalized experiences, and that maximizing privacy is not always the best option for end-users. But frankly, if there is any appetite for personalization (a big if), it is not at all clear that anyone is actually providing a satisfactory experience for that (WSJ)
Work
A new debate on workers’ rights. The pandemic and some recent legal decisions about workers’ rights have created pressure on “shared economy” companies. The problem now seems to be expanding to other areas of the digital industry (e.g. Amazon) and may have long term implications
Amazon is facing challenges to its workforce policies: Business schools like to discuss the “Day 1” culture that Jeff Bezos has been able to keep at Amazon, with its almost 1m US workers being managed as if they were in a startup. But (inevitable) signs of maturity are starting to emerge. And now, amid an expanding pro-worker sentiment in the US, there are claims that the “Day 1” approach, while beneficial for productivity and efficiency, does represent a risk to employees health. This has been a discussion topic in the recent vote about unionization in an Amazon fulfillment center (NYTimes)
The problem could affect Amazon’s margins and create incentives for robotization. Maybe the impact of new labor conditions wouldn’t be massive, but given Amazon’s low margins, it wouldn’t take much to hurt Amazon in a way that impacts the valuation (WSJ)
Antitrust and other control issues
China continues moving to fully control its (massive) local digital economy:
The government has imposed a big fine on Alibaba, for antitrust issues: the antitrust investigation on Alibaba has led to a record $2.8bn fine, also setting what analysts see as a significant precedent (FT)
The Central Bank of China is on the way to launch its own digital currency, a move that could also have implications on giants like Tencent and Alibaba (now selling massively successful financial services) (WSJ)
Peter Thiel believes the Chinese could turn this into a weapon. From a “Second Cold War” perspective, the famous Peter Thiel (a founder of PayPal, a tech celebrity, and a fan of Donald Trump) has said that this, and even Bitcoin, look like Chinese weapons to him (Bloomberg)
The Second Cold War and global tech fragmentation
Fragmentation is happening at all levels of the tech-stack, and local supply chains are starting to emerge:
Countries using bans as an excuse to develop local app ecosystems (that are easier to control). This is similar to what China already did in the past. Recent examples: India (local apps like ShareChat are making progress, after the TikTok ban) and Russia (the government has threatened to ban Twitter, and could make it more difficult to access and use, to create advantages for local competitors like Yandex or Mail.ru) (FT)(FT2)
Internet infrastructure also starting to fragment. An explicit example are the new submarine cables that Facebook and Google are building in South East Asia, steering clear of China (Bloomberg)
Device supply chains being reorganized. Samsung (and many others) are trying to segregate their device manufacturing and chip supply chains, reducing dependence on China (FT)
Batteries (an emerging strategic supply), with local supply chains being built both in India and in Europe (Bloomberg)
Supercomputer chips: the US just added the Chinese supercomputer companies to the official export “blacklist”, effectively banning the sales of American technologies to these firms, under concerns that the Chinese could use them to power their weapons (FT)
Hot topics for investors
Change brings opportunities, and these often lead to financial bubbles: The pandemic, together with the emergence of SPACs as a new investment vehicle for the masses to get easily exposed to startups, has led to a boom in innovative companies going public. Many of these ventures are directly linked to the acceleration of “digital-first” ways of life and work, and to the infrastructure enablers that these new models need
Signs of concern emerge around SPACs: This week there were comments about a significant deceleration of SPAC activity. This happens as some ambitious projects, e.g. on electric cars, or on flying taxis, are showing… well, that startups, after all, imply higher risks… Bill Gates had already warned us (FT)(Bloomberg)
Energy storage is emerging as a hot topic: Green energy is produced at specific locations (solar or wind parks) and specific times (e.g. daylight hours), but it is often needed at a different place, or in mobility (e.g. Electric Vehicles), and later in time. So to make it work, efficient storage mechanisms are required. This week, specific value creation opportunities were discussed for energy storage:
Next-generation, solid-state batteries: They are safer and use fewer raw materials, solving some key environmental concerns for current lithium-ion technologies. They also promise to enable shorter charging times and lower weights, which would mean significant improvements for the customer experience with electric cars. So there is a lot of interest in these new models, and both startups (like QuantumScape or Solid Power) and traditional companies (like Toyota or Samsung) are working to make them happen (FT)
Strategic battery-building programs in Europe, driven by a combination of local startups and traditional companies, and fueled by state support of at least €6.1bn. Startups like Northvolt (Sweden), Britishvolt (UK) and Automotive Cells (France) aim to become regional “champions”, in direct competition with established firms like Tesla or Volkswagen (Bloomberg)
Processes using hydrogen as a storage mechanism: Hydrogen is an effective and clean vehicle to store energy produced in solar or wind plants, for later use. The challenge is to make the process efficient enough, as with today’s technology you still need to spend three times more than with natural gas, even if production costs have declined more than two thirds in the last few years. Projects are being launched globally, to improve these numbers, led by companies like Mitsubishi (FT)
Self-driving vehicles could turn into a (long-term) opportunity, starting by trucks: Self-driving vehicles are not yet mature, but some startups are surfing the current investment wave and going public through SPAC deals”
Initial focus is on trucks, where the business case looks better: We already discussed this last week. The long distance transport market has many inefficiencies that could be addressed with self-driving vehicles, and also these would be easier to operate, as the itineraries are less variable and more predictable. Startups like TuSimple are now focused on this specific segment (WSJ)
But the case for robotaxis is also improving, as we have already said here. The emerging regulatory pressures on gig-economy companies to provide labor benefits to their workers, that could increase costs, could work as an incentive for them to adopt autonomous vehicle technology (FT)
Other opportunities discussed this week:
Enabling the coexistence of AI and humans continues to be a need. Sarcos Robotics, a startup focused on creating “intelligent” tools to enhance human capabilities in factories, or in the battlefront, has recently received a unicorn valuation (Bloomberg)
Air-taxis: an exciting (but risky) bet. These are even less mature than self-driving cars, but some startups, like Lilium ($3.3bn equity valuation!), Archer or Joby are also doing massive SPAC deals, increasingly difficult to justify by their business plans (FT)