Substack

Monday, December 29, 2025

China leaps into frontier innovation and research

The latest annual report of the Australian Strategic Policy Institute (ASPI) covering high-impact global research in 74 technology areas found that China leads in 66 of them, with the US in the rest. The ASPI research, covering 7.7 million unique publications, shows how the US ceded its massive lead in research output in the first decade of the century. 

It is encouraging that India has been gradually increasing its share of high-impact research. It ranks in the top five in research effort in 50 technologies, up from 43 last year. This is next only to the US and China and is higher than the UK in 48 technologies, South Korea in 32, Germany in 30, Italy in 14, Iran in 8, Saudi Arabia in 5, and Japan and France in just 4 each. 

The report also has a Talent Tracker that tracks where authors (for research published between 2020 and 2024) of the top 1 per cent of most highly cited publications are currently working. This shows that the US employs the largest share of top-tier tech talent across both cohorts, with China and the EU following. 

While Indians are only behind China and the US as the undergraduate education source for these authors, it falls far behind in being the postgraduate source and employment destination. In these, it competes with the likes of Singapore. 

John Thornhill points to how China has come to dominate not just published research but also its application. 

In 2005, China accounted for just 6 per cent of the world’s most highly cited research papers but that share had risen to 48 per cent this year. The comparable proportion of US publications fell from 43 per cent to 9 per cent… ASPI’s findings accord with Nature’s latest ranking of research institutions, tracking articles across 145 science journals. In terms of research output, nine of the world’s top 10 research institutions are Chinese with only Harvard University in the top tier. China is now mass manufacturing research; it truly has become a scientific superpower… A separate report from the Special Competitive Studies Project in the US earlier this year also highlighted the striking progress that China has made in adopting many frontier technologies. According to the SCSP’s staff assessment, the US is still leading in semiconductors, synthetic biology and quantum computing while China dominates in advanced batteries, 5G and commercial drones. But the most contested, and arguably most consequential, area is artificial intelligence.

FT long read points to how China has come to eclipse the US even in R&D and innovation spending, and is rapidly pulling ahead. 

For decades, China has been the world’s factory and companies have tapped into a low-cost labour force with few protections and cheap, dirty energy. The country’s scale — as a manufacturing base and as a consumer market — lured almost all the world’s biggest multinationals. But the underlying technology was retained by companies from the US and Europe. Now China’s research and development prowess is allowing it to compete, and potentially beat, the west. Whereas the biggest focus of US innovation has become potential moonshot technologies such as artificial general intelligence, for Beijing R&D largely focuses on addressing shortcomings in the real economy — part of Xi Jinping’s pursuit of technological self-sufficiency. After years of state, corporate and academic efforts to alleviate basic vulnerabilities, China’s advances are now setting up the country to dominate future global supply chains for energy and transport… 

China’s R&D surge has been concentrated in applied areas tied to industrial transformation. This has included advanced materials, 5G, batteries, power equipment and other so-called “enabling” technologies that serve strategic goals… Many areas of Chinese R&D are at the sharp end of technological competition with the US, including artificial intelligence, robotics and quantum computing, bioscience and pharmaceuticals, aerospace and nuclear weapons. However, the OECD data reveals that a central focus for China over the past 15 years has been basic engineering and materials. China’s research advances across areas such as batteries, renewables and alternative fuels are edging the country closer to Xi’s aims of self-sufficiency by cutting a reliance on imported fossil fuels and tech across scores of heavy industries.

The article highlights the dramatic progress made by China in R&D efforts.

The country is close to overtaking the US in total expenditure on R&D, with China spending $781bn and the US $823bn in 2023, according to the OECD. It is a stark change from 2007 when China’s R&D spending of $136bn was less than one-third of the $462bn spent by the US… While government R&D spending in China has exceeded that of the US since 2015, Chinese companies have also rapidly increased their R&D efforts over the past decade, national statistics bureau data shows. The number of corporate R&D institutions has also nearly tripled to more than 150,000. And the number of corporate R&D personnel nearly doubled to 5mn. China is also producing around 50,000 PhD graduates in science, technology, engineering and mathematics (Stem) fields annually, compared to about 34,000 from US universities.

In fact, China is rapidly becoming the go-to place for businesses seeking to expand their R&D frontiers. 

In Shanghai, long the country’s preferred hub for foreign companies, the number of foreign-owned R&D centres has increased to 631, as of September, from 441 in 2018. French automaker Renault does not even sell cars in China, but it is among those companies to have opened an R&D centre in Shanghai this year to learn from the local market. Beijing also saw 58 new R&D centres established by foreign groups in the first 10 months of the year, expanding the total number of foreign R&D centres in the city to 279, according to local officials.

China’s advantages in cutting-edge industrial R&D in industries where big data is an important input come from both its large population and lower regulatory barriers to taking research from the lab to the real world. China has been able to leverage these advantages in areas like autonomous driving, drug development, and AI. 

As the FT long read shows, foreign car manufacturers have realised that their assisted driving functionality development efforts are far more likely to succeed in China than at home. 

The technology (assisted driving functionality to customers), a forerunner to completely driverless cars, has taken the Volkswagen about 18 months to develop, test and now commercially deploy — all in China. It is the fruit of a 700-person research and development team comprised mostly of Chinese software engineers with masters or PhDs and more than five years’ experience. Asked how long it would have taken to deliver something similar back home, Hafkemeyer, who worked with Audi, Chinese state-owned auto group BAIC and tech giant Huawei before joining VW in 2022, sighs with exasperation. Typically, he says, the technology development cycle in Germany is a slog of around four to four-and-a-half years, where ideas are bogged down in endless internal debate and commercial negotiations with suppliers… Since 2018, Mercedes-Benz, BMW, Volkswagen and Stellantis have formed technology partnerships with at least 38 Chinese companies and research institutes, covering software, hardware, batteries and connectivity, UBS data shows.

Another article discusses the rapid advances made by China’s biotech industry, which has, in a short time, vaulted to the frontier in R&D and drug discovery. 

These trends extend to the pharmaceutical industry in general, where Chinese companies have been surging ahead in the drug development trajectory, and ran a third of all clinical trials last year.

At the higher end of the value chain, Chinese firms made up nearly a third of all global licensing arrangements signed by the big pharma companies. 

Cancer remains a central focus for Chinese firms… Weight-loss drugs are a hot target…local firms are not just copying. Bloomberg Intelligence, a research firm, reckons there are 160 new obesity drugs in development around the world; about a third of them are from China.

This is a brilliant description of the evolution of China’s pharmaceutical industry in the last quarter-century. 

China has steadily climbed the rungs of the global drug supply chain. In the late 1990s and early 2000s, it nurtured a crop of generic drugmakers replicating off-patent medicines. It then moved into supplying active pharmaceutical ingredients, before becoming a preferred destination for outsourced biotech manufacturing and, later, drug development itself. At each step, Chinese clinical research organisations (CROs) and contract manufacturing organisations (CMOs), which develop and manufacture drugs on behalf of biotechs, accumulated expertise in ever more complex scientific and engineering processes. 

The progression has turned outsourcing companies such as WuXi AppTec and GenScript into linchpins for drug development not only in China, but for biotech companies worldwide. “There is a joke in the US biotech industry that if you want to run a study, you ‘just WuXi it’,” says Ashoka Rajendra, founder of Orchestra, a San Francisco-based software platform for biotech companies. “There is an entire infrastructure in China that is supporting US and European R&D,” he says. “At first, companies would go to China for very specific deliverables, like making DNA or proteins. Then the work moved up the sophistication curve. It wasn’t just tasks with a precise set of instructions, but running studies and generating data.”…

Over the past decade, this same knowhow has been redirected from servicing foreign to local clients. China’s biotech sector is experiencing the same evolution seen in automotive manufacturing, consumer electronics and even textiles: from a low-cost outsourcing hub to a source of its own intellectual property. This shift is especially pronounced in biologics and cell therapy, areas that hinge on rapid experimentation and iterative lab work — tasks that are both cheaper and faster to execute in China.

The Economist has a similar description of how Chinese drug makers vaulted so quickly to the top. 

The first answer concerns how quickly they innovated in order to produce “fast followers”, which improve on existing drugs’ safety or delivery. From there they were able to move on to “first-in-class” medicines with new targets (such as a protein or a gene linked to disease) or mechanisms (which might block or boost a target’s function). Fast-follower and first-in-class treatments now make up more than 40% of the industry’s pipeline, according to interesting research published last year in Nature Reviews Drug Discovery... Wang Xingli of Fosun Pharma, a Chinese drugmaker, explains that working on fast-follower drugs gave China’s industry the “courage to do the first-in-class”.

The second answer concerns the speed, scale and low cost of other processes. China’s firms can take a drug from discovery to the start of human trials in about half the global industry’s average time. Human trials themselves, usually the slowest stage in the development of a new drug, also move faster. A vast patient pool makes enrolment easier, and a large network of trial centres also speeds things along. This model has proved especially helpful for developing antibody-drug conjugates, a novel class of treatment that seems particularly potent at tackling cancer. One big-pharma executive says that China’s appeal lies in the sheer number of firms experimenting: “You can pick the winners and improve the odds of approval”... State insurance covers most purchases, pooling demand from hospitals and forcing firms into bidding wars. To win coverage, drugmakers must often slash prices by half or more to reach a wide patient pool, or accept a far smaller private market.

This is the journey that Indian generics manufacturers failed to travel despite having started much earlier than the Chinese. They have remained largely stillborn at the generics and Active Pharmaceutical Ingredients (API) stages.

This from the FT article describes the value proposition and advantages of Chinese biotech companies. 

A McKinsey report… estimates that Chinese drugmakers can move two to three times faster than the global average in advancing a target molecule into a drug candidate and into early clinical trials. Once a therapy passes initial animal tests, patient enrolment for clinical trials in China is two to five times faster, depending on the therapeutic area, thanks to a large patient pool with unmet medical needs… It took officials just nine days to approve BiBo Pharma’s construction plans — a process that can drag on for up to six months in Europe, where Peng is scouting for another manufacturing site… According to McKinsey, the time it takes to enrol patients in clinical trials in China is about half the global average, and the cost per patient is roughly 50 per cent lower than in the US and Europe…

Chinese companies can achieve some of this speed and cost saving compared with western rivals due to generous state support… biotech start-ups that receive subsidies, cheap rent and start-to-finish assistance… Tapping into the country’s network of supply chains accelerates every stage of the research and development cycle, from discovery to scale-up manufacturing and clinical testing. Shorter timelines improve survival odds for early-stage biotechs, allowing them to test more ideas, kill failures quickly and improve their chances of discovering a blockbuster therapy…

Egan Peltan is one such US biotech founder who relies on Chinese suppliers. “Initially, the industry used Chinese CROs because they were cheaper. That is no longer true; their prices are similar. We use them because of the efficiency and time savings,” he says. For young biotech companies, supplier speed can be existential. “A project that takes a US CRO six to eight weeks would take a Chinese company about three . . . For early-stage companies with no revenue, every day you are waiting is money you have to set on fire,” Peltan says.

This speed, and associated lower costs, creates its set of advantages for Chinese companies (and their foreign partners). 

“Because the cost of developing a drug is so much lower in China, companies can afford a broader pipeline,” says founder and chief executive Stella Shi. “Where a US company might have to focus on a single asset, Chinese biotechs listing in Hong Kong typically have at least 10 candidates.” The data reflects that shift. China’s share of global innovative drug candidates in clinical trials has risen from 8 per cent in 2018 to 30 per cent this year, according to McKinsey. Over the same period, the US share has fallen from 47 per cent to 36 per cent.

China’s strengths in biotech are concentrated in fields that demand intensive engineering and high-volume laboratory work to iterate and refine new therapies. This is particularly evident in antibody-drug conjugates (ADCs), so-called “biological missiles” that send chemicals directly to tumours, as well as multispecific antibodies, immune proteins that bind to the surface of cancer cells. McKinsey estimates that Chinese companies account for 54 per cent of innovative ADC assets in phase 1 and 2 clinical trials, and 48 per cent of multispecific antibodies… China’s growing capabilities have drawn pharmaceutical multinationals to the country to either buy or co-develop innovative drugs. China’s share of out-licensing deals to the US and Europe has risen from 2 per cent in 2018 to 20 per cent this year… China’s share of new drugs approved by the US Food and Drug Administration — defined as medicines containing active ingredients not previously approved — has risen from 1 per cent in 2018 to 6 per cent so far in 2025, according to McKinsey.

Antibody-drug conjugates (ADCs) are a good illustrative example of the strengths and value proposition offered by Chinese companies. 

China’s strengths in biotech are concentrated in fields that demand intensive engineering and high-volume laboratory work to iterate and refine new therapies. This is particularly evident in antibody-drug conjugates (ADCs), so-called “biological missiles” that send chemicals directly to tumours, as well as multispecific antibodies, immune proteins that bind to the surface of cancer cells. McKinsey estimates that Chinese companies account for 54 per cent of innovative ADC assets in phase 1 and 2 clinical trials, and 48 per cent of multispecific antibodies. Sichuan Biokin, a listed company developing targeted ADC therapies for cancer, has 17 treatments in its pipeline. 

ADCs drew significant investor enthusiasm in western biotech in the early 2010s, thanks to their potency in killing cancer cells. But many programmes stalled because of the difficulty in controlling toxic side effects. “ADCs have stronger tumour-killing power. The problem in the 2010s was that no one could solve the toxicity issue,” says founder and chief executive Zhu Yi. He describes the challenge as fundamentally an engineering puzzle: identifying the “right specific technical path” that allows a highly toxic payload to attack cancer cells without damaging healthy tissue. On this front, he argues, “Chinese companies have a natural advantage” due to the abundant supply of skilled technicians to run rapid-fire experiments and the relative ease of conducting clinical trials.

As with all other sectors where China today dominates, active industrial policy has played its facilitating role.

Zhang Fangning, McKinsey’s partner of life sciences in Greater China, says, “But the origins of this growth lie in factors that have worked in concert for more than a decade: a steady growth of talent and capital; supportive policies; a robust local supply chain; and the sheer intensity of engineers and scientists who are driving the execution.” Beijing accelerated this transition by prioritising biotech as a strategic sector both for economic growth and as a means to develop drugs tailored to the ethnic Chinese population. It introduced reforms in the mid-2010s that made it easier for biotech companies to raise capital and pursue innovation. Hong Kong then loosened stock market listing rules in 2018, enabling pre-revenue companies to list and the resulting wave of biotech IPOs ushered in venture capital investment. At around the same time, drug authorities in China relaxed clinical trial requirements for certain categories of innovative therapies, making it easier to progress drug candidates… Chinese regulators drive hard bargains on drug pricing, meaning companies must expand abroad to achieve the profitability needed for long-term growth.

Also this

Approval processes have been streamlined, priority reviews conducted for drugs tackling critical conditions, and regulations brought closer to international standards. The workforce at China’s drug regulator quadrupled between 2015 and 2018, and a backlog of 20,000 new drug applications was cleared in just two years. The time taken to secure approval for human trials shrank from 501 days to 87. And the output of new medicines soared. In 2015 China approved only 11 treatments, mostly Western imports. By 2024 the figure had risen to 93, with 42% developed domestically. These reforms have been matched by efforts to lure back students and professionals who had studied or worked abroad. Many of China’s “sea turtles”, as such returnees are jokingly known, came back with experience of building biotech firms and dealing with investors and regulators. Their entrepreneurial zeal was bolstered by rules making it easier to raise funds and to list on the Hong Kong stock exchange.

Echoing much the same, The Economist has another article that describes how China has come to dominate the automotive industry and is now threatening to do the same in the pharmaceutical industry.

A deep pool of talent, a broad manufacturing base and huge scale combine to propel it rapidly up the value chain. The production of robotaxis has piggybacked on mass EV manufacturing and a dominance in the supply of lidars and the other sensors needed for self-driving; scale has also helped bring down costs. Armies of patients enlisted in clinical trials and profits from generic drugmaking have speeded up pharma innovation. 

A more surprising ingredient of China’s success is its nimble and permissive regulators. As in other industries, local governments have offered firms cheap credit and other help. But it is agile rulemaking that has really turbo-charged progress. Soon after political leaders set out their ambition for China to become a “biotechnology superpower” in 2016, the country implemented a number of reforms… Likewise, China was early to experiment with robotaxis. Local officials, keen to attract talent and investment, approved pilots at a rapid clip and installed sensors and other digital infrastructure to help guide self-driving vehicles; trials have run in over 50 cities. Many have experimented, too, with laws on liabilities and guidelines for testing. Though accidents have sometimes caused a hiatus, pilot schemes have helped engineers and policymakers understand the new technology.

Cut-throat competition at home imposes harsh conditions on individual companies, but the survivors are conditioned into becoming hypercompetitive export champions. China’s robotaxi operators compete with each other and with cheap human-driven taxis in an economy gripped by deflation. New technologies receive subsidies that ultimately come out of the pockets of its underpaid people. Many lossmaking enterprises will not survive the resulting price wars. But those that do will look overseas to make money.

The FT article also points out that the Chinese companies’ next challenge is to break into the global markets. It is here that they find value in foreign multinational drug companies. 

China’s strengths in developing novel medicines lie in fields where the foundational research has already been established, but laboratory work is needed to figure out the right formulations… “China’s early-stage R&D is now globally competitive. But for late-stage development — overseas trials, regulatory filings and commercialisation — Chinese firms still have significant room to grow,” says UBS pharma analyst Chen Chen. “That won’t always be the case,” she adds. By partnering with multinational drugmakers to take products abroad, Chinese biotechs can “secure funding for continued innovation while learning from partners’ strengths in clinical development and commercialisation”. “Over time, Chinese companies will build their own global capabilities,” she says…

Josh Smiley, chief operating officer at Zai Lab, says it will take time for companies in China to develop “systems that contemplate global development. We’re going to continue to see really good innovation come out of China in the next few years. But in many cases, those innovators are going to need a partner to help them with development,” he says. They must also create international sales forces, navigate foreign regulatory systems and cultivate supply chains outside China, especially as western governments push to create localised manufacturing for critical industries such as pharmaceuticals. For many Chinese biotechs — often unprofitable and with volatile revenue — these investments remain out of reach…

“Partnering with a multinational lets us learn first-hand how global operations work. It’s like having a personal tutor showing you, day by day, how to build a global business,” says Sichuan Biokin’s founder Zhu Yi. Working with larger foreign companies, he adds, has helped Sichuan Biokin understand the organisational structures and capabilities needed to scale. “We mirrored that and learnt by doing. But we have also seen the weaknesses of Big Pharma — the bureaucracy and inefficiency — which has helped us avoid detours as we grow.”

In other words, Chinese pharma companies are now relying on foreign firms to do exactly what their counterparts in different sectors have done to build their capabilities, and then overtake and marginalise their foreign partners. The outsourcing of even research and clinical trials, and not just manufacturing, to Chinese firms also parallels the strategy followed by multinational corporations in other sectors on their path to marginalisation. Notwithstanding these sobering tales, attracted by the immediate commercial benefits of cheaper and quicker contract research, manufacturing, clinical trials and drug development processes, the US and European pharma companies are likely to overlook the long-term consequences. In the process, they are likely to build Chinese capabilities and fall by the wayside. 

Given the demonstrated willingness to weaponise its manufacturing dominance, any Chinese dominance in an industry like pharmaceuticals will be even more dangerous. Further, unlike with other products, the issues of health safety and side effects are critical considerations with drugs, and the concerns with the quality of clinical trials and other tests done by Chinese pharma companies will remain a matter of concern. It is, therefore, important that US and European policymakers step in immediately to restrict their pharma companies from making such investments in China and forging partnerships with Chinese pharma firms. These kinds of decisions are easier to make when the market trends of outsourcing, licensing, and partnerships are in their early stages. 

In this context, China’s pharmaceutical companies with their cheaper drugs will be attractive to consumers in developing countries. Imagine an Ozempic at a fraction of the price. For these countries that are already dependent on the US and European drug companies, the prospect of replacing them with Chinese companies is unlikely to be a problem. We should not be surprised if Chinese drugs come to dominate the markets across the developing world in a short time. 

Here, India is in a unique position. It is unlike the other developing countries in that it has serious national security issues with China, and it has its own large pharmaceutical industry with aspirations to move up the value chain. Given the inevitability of Chinese drugs becoming available in the near future, and that too at lower prices compared to the Western alternatives, it may be prudent for India to tightly restrict their access to the Indian market.

Saturday, December 27, 2025

Weekend reading links

1. The rise of zero-sum politics in the West.

In the US, UK, France and Germany zero-sum beliefs on the left (eg people only get rich by making others poor) and the right (eg immigrants succeed at the expense of the native-born) are related expressions of the same underlying worldview. Namely that there is only so much to go around and we must therefore use restrictions, exactions and preferential treatment to redress the balance between winners and losers.

3. Tim Wu contrasts America's all-in bet on proprietary AI-led innovation with China's diversified bet on renewables and green technologies coupled with applications of AI through open-source models. The article has this graphic which shows how Chinese exports to developing countries has taken off exponentially.
Last year, 70 per cent of the world’s EVs were manufactured in China. China also accounts for roughly 80-85 per cent of global solar photovoltaic manufacturing, and more than 75 per cent of all global battery production.
4. Data centre construction in the US is now on par with office construction!
And this on their power demand.
Across America, data centres represent a combined capacity of about 51GW. Running at their maximum, this equates to 5 per cent of the country’s peak demand. By 2028, an estimated 44GW of additional capacity will be required by new data centres, according to S&P Global Energy. Given constraints to grid infrastructure, power capacity coming online in the next three years will only be able to provide about 25GW for these data centres. That leaves a gap of 19GW — just over 40 per cent of the power needed... After more than two decades of flat or anaemic growth, US power demand is now surging. Electricity usage is projected to rise by an average of 5.7 per cent a year to 2030, based on forecasts from utility companies... more than half of the expected increase stems from the rapid build-out of AI data centres, according to consultancy Grid Strategies.

And the constraints to power capacity expansion.

Boosting the US power grid is an enormous and time-consuming task due to a complex web of regulatory, financial and supply chain challenges. Interconnection queues — backlogs of projects waiting to plug into the network — have become a major chokepoint, slowing the rollout of new power capacity and leaving data centres facing lengthy delays... The average time from filing an interconnection request to achieving commercial operation now exceeds eight years, according to energy think-tank RMI... On average, federal permitting for a new US transmission line takes about four years, according to the Department of Energy. State processes add further delays to grid build-out. Last year, almost 900 miles of new high-voltage transmission lines were completed, according to lobby group Americans for a Clean Energy Grid. This is the most since 2020, but still far short of the 5,000 miles a year the group estimates is needed to support grid reliability and growth.
5. Like in China, India's solar manufacturing capacity is entering a glut and is also due for consolidation. 
Edinburgh-based energy consultant Wood Mackenzie estimates the country’s solar module manufacturing capacity will exceed 125GW by 2025, which is more than three times its domestic demand of around 40GW. Nomura projects capacity additions of 100-110GW in the next three years, further raising the risk of oversupply and painful consolidation.
This is an interesting comparison with China, pointing to future pain.
In China, half of the six major solar IPOs of the past two decades now trade below their issue price. In the US, SunPower has filed for bankruptcy. Even today, China’s JA Solar, with nearly 85GW each of ingot-wafer and cell capacity, is valued on par with India’s Waaree Energies, which has only a fraction of that scale... Some distress from oversupply is already visible in the supply chain. While the large, cash-rich players, including Adani, Waaree and Premier Energies, are tightening their hold on the sector by backward-integrating their supply chain and protecting their future margins, the smaller players are struggling to keep their plants running.

Worsening matters further is the lack of demand for solar power among discoms. 

India’s solar energy capacity currently totals 130GW, according to the ministry of new and renewable energy... As of September, around 44GW of tendered clean energy capacity, out of 93GW since fiscal year 2024 (FY24), remains without buyers... The ministry said in early November that it may look to cancel these projects on a case-by-case basis. One of the primary reasons for unsigned agreements is that state utilities expect solar prices to fall further. 
This capacity explosion has been facilitated by industrial policy.
In 2022, India imposed 40% tariffs on solar modules and 25% tariffs on solar cells to discourage imports from China. Last year, the country dictated that Indian solar power producers must purchase from an approved list of domestic solar-module makers—there are 93 companies in the approved list so far. Similar rules for solar cells will come into effect next year. Further restrictions on imports of ingots and wafers that are building blocks of modules and cells are expected in the coming years.

But industrial policy does not appear to be able to bridge competitiveness. 

Under new domestic content requirements, an entirely ‘Made in India’ module would cost more than double Chinese-manufactured modules, making it uncompetitive without substantial government policy support.

While India's solar manufacturing capacity is now largely at the level of modules, the government is pushing hard to integrate backwards into cells, ingots, wafers, and polycrystalline. But that's challenging. 

Owing to the complexities involved, the cell is the most capex-intensive segment. According to Nomura, capital expenditure per GW of cells can total ₹6,500 crore versus ₹2,000 crore for modules. Ingots and wafers need up to ₹4,500 to build 1GW capacity. As the supply glut in modules deepens, large integrated players such as Reliance Industries, Adani Enterprises, Waaree, Premier and Tata Power, which are investing across the value chain, are better positioned to survive... All large and listed players have announced plans to backward integrate, expecting that impending import restrictions further down the value chain will keep profits coming. According to Nomura, 70-80GW of cell capacity additions will come online over the next three years. It is 18.5GW right now.

Solar energy has the lowest tariff rate.

While the ratio of debt to GDP in the world’s biggest economy shrank from 106 per cent in 1946 to 21.6 per cent in 1990-91, it has since lurched back up to almost 100 per cent thanks to, among other things, the financial crisis and Covid-19... The postwar experience of the UK provides a case study of how these factors interact. The country’s debt-to-GDP ratio went from more than 250 per cent in 1946 to just 42 per cent three decades later. In a seminal piece of research, Barry Eichengreen and Rui Esteves show that for most of the 1946 to 1955 debt consolidation episode, the UK ran consistent, large primary budget surpluses despite the Labour government’s huge expansion of the welfare state. Yet the largest contribution to debt reduction came from inflation, which was responsible for more than 80 per cent of the debt consolidation over the period. That said, from 1955, fiscal discipline and economic growth did most of the work — surprisingly given Britain’s record at the time for economic incompetence — because the contribution of consumer price inflation, which peaked at 24 per cent in 1975, was neutralised by rocketing interest rates.

7. K-shaped income gain + K-shaped wealth gain = K-shaped economy. Rana Faroohar writes about the K-shaped economy in the US.  

Consider income growth, which was higher for low-income households right before and during the pandemic — in large part because of support from the Biden administration — but has diverged since. Wage growth for low-income workers is now lower than for middle- and high-income workers. This is partly explained by the artificial intelligence boom that is showing up in higher unemployment figures for young college graduates as more entry-level white-collar work is done by technology. Asset growth is K-shaped too, with higher-income households seeing lots of paper wealth from stocks at still near-record highs and rising home prices. According to investment group Apollo, the cash flow received in fixed income, including private credit, is nearing levels not seen in decades. That wealth effect has propelled the existing K-shaped trend in consumer spending. The percentage of overall spending done by the top 10 per cent of the socio-economic spectrum has risen from 36 per cent to nearly half since 1989, according to Moody’s analytics.

8. Daron Acemoglu points to the breakdown of the liberal democratic politics.

Liberal democracy was made by its pledges. It plunged into crisis because of their undoing. A lot of this volte-face was about the eclipse of the industrial compact and the rise of a post-industrial society, dominated by digital technologies and the college-educated professionals that these empowered. Digital technologies severed the link between economic growth and shared prosperity. With the widespread automation enabled by digital tools, companies could expand without hiring more employees and paying workers more, and the skill bias of these technologies gave a boost to the earnings of highly educated and managerial workers. The result was a staggering increase in inequality in the US, with the inflation-adjusted wages of low-education men falling most years between 1980 and 2014 — even as the aggregate economy and the urban, globalised professionals were flourishing.
That the computer age was leaving behind the working class, which used to typically support left-leaning parties, was unnoticed by the college-educated, who were becoming politically and socially ascendant in the environment digital technologies created. That they had started living separately, socialising separately, marrying separately, and holding very different views from the less educated undergirded this omission. There was also a major sin of commission on the part of left liberals. As they abandoned classic working-class or social democratic issues, they started focusing on cultural politics — in part because cultural divides had become more pronounced and in some ways more intractable in an age defined by shifting mores, globalisation and increasing immigration flows from countries with dissimilar traditions.

But the cultural divide that emerged between different education groups and ideologies did not have a simple solution. Even as norms were changing on important issues such as gay marriage, new rifts were opening related to the assimilation of new immigrants, transgender rights and cosmopolitan versus local priorities. The college-educated, fatefully, turned to social engineering efforts, trying to accelerate cultural change — in universities, schools, the entertainment industry and even workplaces. These efforts, though often well meaning, were nonetheless perceived by many working-class communities as the imposition of the priorities of college-educated values on the rest of society. The scene was set for a crisis of liberalism and of liberal democracy.

9. Importance of Samarium, a rare earth mineral, and how the US gave up its leadership.

Most rare-earth magnets are made of neodymium, which is used in everyday applications such as cellphones, auto parts and electronics. But the defense industry requires samarium-cobalt magnets, which can withstand extreme heat... Unless new sources of samarium or a substitute material can be found, American manufacturers won’t be able to build fighter jets or precision-guided missiles. They may be forced to sacrifice precision if they can’t get the right magnets... 

Although samarium-cobalt magnets were invented in an Air Force research lab in Ohio in the 1960s, the industry moved to China in the 1980s, partly because of rich rare-earth deposits there. Today, China mines, processes, sells and consumes such large volumes of rare-earth metals that it can drop the price below the cost of production when foreign competitors come online. American and European companies have struggled to stay afloat. Many either declared bankruptcy or opened factories in China.

And the difficulty of reshoring, even with good policy intent, without a crisis hitting.

In recent years, American policymakers have tried to build a domestic supply. The National Defense Authorization Act of 2023 and 2024 gradually tightened restrictions on the use of rare-earth metals from China in weapons systems, and stipulated that all such materials must be China-free by Jan. 1, 2027. But such mandates have been inconsistently enforced, partly because alternatives are not available. In 2023 and 2024, when magnets were supposed to be made of metal that was created outside China, Lockheed Martin notified the Pentagon that its F-35 Joint Strike Fighter had Chinese-made magnets. The military paused production of the jet for months but eventually issued a waiver allowing the parts.

10. The Bank of Japan raises its benchmark interest rates to a 30 year high.

Alongside, yields on 10 year government bonds have touched 2% for the first time since 1999. 

11. Assessment of the Insolvency and Bankruptcy Code implementation.

The data from the Insolvency and Bankruptcy Board of India (IBBI) shows that by September, 8,659 corporate insolvency resolution processes (CIRPs) had been admitted. Of those 1,898 cases were ongoing. More tellingly, about 1,300 CIRPs that resulted in resolution plans took an average of 603 days, while 2,896 cases that ended in liquidation took 518 days, far exceeding the statutory outer limit of 330 days prescribed under the IBC... Despite procedural delays, the IBC has had a meaningful impact on India’s banking system and credit culture. Resolved cases have delivered 32.44 per cent recovery of admitted claims, translating into more than 170 per cent of liquidation value, and have helped rescue about 1,300 firms. Equally important, the threat of losing control has altered borrower behaviour, improving repayment discipline and encouraging early settlement.

12. US corporate profits are at all time highs.

13. America's K-shaped economy.
14. Financial engineering is never far away in boom times. FT reports how Big Tech firms are shifting debts assumed to finance AI spending out of their balance sheets through project financed SPVs. 
Financial institutions including Pimco, BlackRock, Apollo, Blue Owl Capital and US banks such as JPMorgan have supplied at least $120bn in debt and equity for these tech groups’ computing infrastructure, according to a Financial Times analysis. That money is channelled through special purpose holding companies known as SPVs. The rush of financings, which do not show up on the tech companies’ balance sheets, may be obscuring the risks that these groups are running — and who will be on the hook if AI demand disappoints... 

Tapping private capital funding through off-balance sheet structures protects companies’ credit ratings and flatters their financial metrics. Meta in October completed the largest private credit data centre deal, a $30bn agreement for its proposed Hyperion facility in Louisiana that created an SPV called Beignet Investor with New York financing firm Blue Owl Capital. The SPV raised $30bn, including about $27bn of loans from Pimco, BlackRock, Apollo and others, as well as $3bn in equity from Blue Owl. The deal meant Meta could in effect borrow $30bn without any of the debt appearing on its balance sheet. This made it easier to raise a further $30bn in the corporate bond market a few weeks later.

15. Demystifying Adam Smith's invisible hand

The popular understanding of the “invisible hand” is even further off the mark. Smith borrows the phrase from Macbeth, who talks about a “bloody and invisible hand” shortly before murdering Banquo. In all his works, the economist mentions the phrase just three times, in three different contexts—and never in reference to the price mechanism... In fact, he often favoured the visible hand of government. He urged the state to provide education. He favoured legal caps on interest rates. Today, almost all free-market economists despise America’s Jones Act, which requires that shipping between American ports be conducted on vessels that are built, owned and largely crewed domestically. Smith, by contrast, favoured the Navigation Acts, a similar British law. 

Smith acknowledged the benefits of markets, but also their costs. Consider his famous pin factory. The division of labour within it allowed workers to produce thousands more pins than if they were working alone. Countries that perfected the art of dividing labour, Smith argued, would grow rich. Yet he also worried that a life spent on a few simple operations would make a labourer “as stupid and ignorant as it is possible for a human...to become”. Did Smith think the costs outweighed the benefits? It is hard to be sure.

16. VC failure rates

Sequoia’s best-ever US fund had half its investments fail.

17. Cross-border payments company Aspora is disrupting US-India remittance transfers. 

Fintechs Wise and Remitly do that by partnering with a local bank, with which they park funds worth two to three days of remittance volume as a lien. This fund lies as a security, untouched. Now, when a person in the US sends money to somebody in India, the fintech pings its Indian banking partner, which uses its own funds to make a local transfer to the recipient’s account. This reduces the speed of transfer from days to minutes. The fintech then reimburses the banking partner for all the transfers via a bulk cross-border transfer.

... 30–40% of the $135 billion remittances to India are locked in as liens to banks. Aspora does about $300 million worth of remittances a month. So, to maintain two days’ worth of remittance volumes with its local banking partners, like Yes Bank, the startup needs to park away at least $20 million of capital... In 2024, Aspora chose to route nearly a third of its cross-border remittances through stablecoins. These digital currencies operate with little legal oversight and take away the need to have so much capital as a lien... In using stablecoins, all that a fintech like Aspora had to do was partner with a cryptoexchange in the originating and receiving countries. And in the process, it can simply swap the US dollar for a stablecoin like Tether. That gets swapped out for the rupee by another exchange in India. This swapping involves a fee of 20 basis points in all, said the crypto-exchange executive, as the exchanges also take care of compliance, conversion, and the payout. That’s much less than maintaining liquidity, which can add up to 1% of the total cost...
When a fintech uses stablecoins to process remittances, the recipients are in a fix. For one, Indians have to pay a 1% tax on the money they receive. Two, the instruction that comes along with the remittance would only show that the money came from an exchange, not the sender... In fact, in some cases, when users sent money to their own accounts in India, local banks, unable to see the sender, saw it as suspicious activity and blocked their accounts.

The big risk Aspora faces is regulatory. Though not banned in India, RBI does not recognise Stablecoins or cryptocurrencies generally. Only about 1% of the $135 bn annual remittances use Stablecoins for now. 

The article also has an interesting graphic about the changes in sources of remittances into India. Declining share of Gulf remittances (except Qatar) and increasing shares from the US, UK, Australia, and Singapore. 

 18. For all talk of China's AI surge, it has few listed companies in the sector.

And the US and Europe dominate the higher end of the supply-chain. 
19. The world economy's China problem in one graphic.
20. Huawei triples local sourcing ratio in smartphones from 19% in 2020 to 57% in 2024!
Huawei increased the proportion of Chinese-made components in the Mate 70 Pro to 57%. The estimated total component cost of the Pura 80 Pro is $380, with the Chinese-made component ratio steady at 57%... For the Pura 80 Pro's system-on-a-chip, which integrates multiple semiconductors, the company used the Kirin 9020 chipset designed by subsidiary HiSilicon... For DRAM, which handles short-term memory, Huawei switched from imported products to those made by ChangXin Memory Technologies. For long-term NAND flash memory, it switched to products made by Yangtze Memory Technologies. Huawei switched to products from BOE Technology Group for the organic light-emitting diode display, which is estimated to cost over $64 per unit.
But in recent years Chinese companies have also entered the sphere with state support, led by Kaluga Queen, a farm on Lake Qingdao. And they have dived in with such stunning efficiency and focus — echoing what has happened with, say, solar panels — that Kaluga is now the biggest caviar producer in the world. Indeed, China accounts for between half and two-thirds of global production... And Chinese officials now want their entrepreneurs to expand into other gourmet foods like smoked salmon, Wagyu beef and truffles. That is creating waves: at a recent meeting of the North Atlantic Seafood Forum, a Nordic luminary flourished a 7kg Chinese-farmed salmon on stage — and declared it to be tasty, and cheap because of Beijing’s subsidies. Meanwhile, the Japanese government has restricted exports of Wagyu genetics to China to protect its beef farmers, and some Italian and French caviar houses are complaining about the pricing threat from Chinese rivals. American caviar makers are reportedly lobbying the White House for protection, too.
22. Finally, in celebration of racial integration, the Springbok rugby team.
The one thing that unites South Africans of all colours is the Springboks rugby team... South African rugby has been so “transformed” — a word the African National Congress uses to mean overcoming the grim legacy of apartheid — that affirmative action is no longer necessary. A squad, picked purely on merit, is automatically multiracial. The Springboks’ most celebrated players include Siya Kolisi, the inspirational captain, who is Black and from an impoverished township in the Eastern Cape. Sacha Feinberg-Mngomezulu, the brilliant fly-half, has a Zulu mother and a father of Jewish heritage. The 50-plus member squad named this year by Johan “Rassie” Erasmus, the Afrikaner head coach who has led the team to successive World Cup victories, contains players from South Africa’s Black, white and so-called Coloured communities. The Springboks are a case study of what successful Black empowerment looks like. Where once players were selected from among 4.5mn white people, today they are drawn from the entirety of South Africa’s 65mn population.

Friday, December 26, 2025

Indian economy's private investment problem

As the year draws to a close, the Government of India has unveiled a series of reforms - deregulation, GST rate rationalisation, Labour Code implementation, removal of limits on FDI in insurance, allowing private investment in nuclear power, etc. 

In this context, some observations on the India economy as we move into the new year. 

1. Arguably, the biggest concern with the Indian economy, and that too for a long time, has been private investment. It has been the dog that has not barked. And it has been the case since the Global Financial Crisis. 

Even as public investments have surged, especially since the pandemic, private investments have remained stuck in the 11-12% range for the last 15 years and have declined as a share of the gross fixed capital formation, which itself has declined. 

Even the corporate tax cut of September 2019 has failed to stimulate private investment. Companies appear to have used the rise in cash reserves to pay off debts. As the Indian Express report points out, the “interest coverage ratio of more than 3,000 companies — excluding those from the financial sector — has more than doubled to 5.97 in the first half of 2025-26 from 2.6 in the same period of 2020-21”. 

Further, as an NIPFP blog shows, far from raising investment, the post-pandemic period is associated with declining fixed assets as a share of total assets and rising financial assets.

The NIPFP blog also finds a distinct rise in the share of long-term and short-term financial investments of the BSE 500 companies. In 2025, nearly a quarter of the assets of these companies were financial investments, a rise from just over 15% in 2015. 

This misallocation towards financial assets is also partially driven by the attractive returns generated by the capital markets. The graphic below shows that the NIFTY 50 has outperformed corporate returns in most years since 2010-11. 

2. Ultimately, private investment is driven by expectations about aggregate demand. The tepid aggregate demand has been a major problem with India’s economic growth model. While the country has grown at a rapid clip for most of the last quarter-century, it has not translated into broad-based gains across population groups, an essential requirement for sustaining high growth rates (instead of the episodic high growth periods that have been a feature of India’s post-liberalisation growth). I have blogged herehere and here in greater detail. 

This may well be the main explanation for the stagnation in private investment. Attesting to this, as per the RBI’s Obicus Survey, capacity utilisation in manufacturing has crossed 75% of capacity (the threshold thought to be required to get firms to invest) only in 10 of the 53 quarters since the start of 2012-13. Adding to the reluctance of corporates, their profitability has been on a sharply declining trend. 

On the household side, since 2017, indebtedness, excluding mortgages, has been on a rising trend, increasing by about 40%. This is certain to have exercised a drag on consumption. 

This trend also coincides with household savings dipping to a 50-year low of 5.3% of GDP in 2023-24. 

I used PLFS data from here and ChatGPT to generate the household income index by income decile for the 2018-19 to 2023-24 period. Note the K-shaped trend post-pandemic. 

It shows that the D3 to D5 deciles grew the fastest, with the bottom three deciles being the weakest (the lowest decile declining). 

blog at Marcellus Investment Managers points to three back-to-back stimulus measures unveiled by the government to boost consumption - the income tax relief in the Union Budget of February 2025 by raising the tax slabs (Rs 1 trillion stimulus), the RBI’s five repo reductions (reduced interest outgo by Rs 2 trillion), and the GST 2.0 (Rs 2 trillion stimulus). The combined effect of the total of Rs 5 trillion amounts annually to about 1.7% of GDP, and its impact remains to be seen.

3. In any case, for all its big population size, the consumption potential of the Indian economy is woefully limited. This nicely sums up the challenge of making money in the country. 

While India’s population of 1.4bn offers enviable scale, its market has proven difficult to monetise. According to Sensor Tower, Indian internet users downloaded 24.3bn apps in 2024 Andy spent 1.13tn hours on them, but total spending was just $1bn.

This highlights the importance of well-paying jobs in creating a large enough consumption class. Unfortunately, the gig economy, the largest creator of jobs, may not suffice in this regard, though it has its value in absorbing the large pool of workforce entrants. 

4. In fact, it’s a possibility that the gig economy may actually be displacing more productive jobs. Bloomberg has an excellent article that captures a big labour market paradox, attracting workers to the factory floor amidst an acute shortage of good jobs and high youth unemployment. Surveys indicate India’s manufacturing sector faces a skills gap of 10%–20% across major functions, and around 75%–80% of employers report difficulty recruiting qualified talent. 

But when you can easily get the same Rs 15000-18000 per month at your convenience with gig work like delivery riders (the gig sector employment is growing at 13% annually), why would you want to do the regimented hard grind of factory floor jobs? As the Bloomberg article writes,

Even as the world’s fastest-growing major economy bets on manufacturing to fuel its rise, the factory floor is no longer where many young Indians see their future. India’s biggest economic challenge isn’t choosing between factory floors and delivery apps — it’s proving it can grow both. Manufacturing drives exports and global clout; gig work brings opportunity to a nation struggling with the highestyouth unemployment in Asia.

Tiruchirappalli offers a vivid case study of this dilemma… more than 15,000 industrial units line the highways around the city, from fabrication shops to distilleries. For decades, these plants relied on workers from nearby villages. But rising incomes and better education in southern India have changed aspirations: locals now favor jobs in larger cities, and manufacturing is widely seen as carrying less social standing. Labor rules are loosely enforced — nine-hour shifts often stretch to 11 or 12 hours with no overtime pay. Gig work, by contrast, rewards longer hours and offers flexibility: Drivers can work two hours one day or 12 the next. To delivery driver Mark Sebastian Raj, a motorbike, a smartphone, and the choice of where and when to work, feel more appealing than clocking in at a foundry… The challenge is whether India can scale up factories fast enough to compete with China while protecting gig workers — without regulating them so heavily that it erodes the very flexibility that makes such jobs appealing.

Reflecting the lack of good jobs amidst the surge in gig and contract jobs, a graduate in India is more likely to be unemployed than an illiterate youth. 

Tertiary education in India continues to remain excessively concentrated in liberal arts, in stark contrast to China. 

5. India’s IT industry was a great opportunity for the country to leverage its head start and chart out a broad-based services sector-led economic growth model. Unfortunately, as I have blogged earlier, it has missed a succession of technology waves in the sector. And it has already fallen way behind on the AI-wave, too. 

In fact, the IT sector may well be described as a good example of unproductive appropriation of corporate surpluses. While the IT industry generated large profits, which, instead of investing in R&D and emerging technologies, it chose to return to shareholders in the form of buybacks

IT services still dominate with exports set to reach $210bn this financial year, India Ratings and Research forecasts. It has been a powerhouse industry for India but as IT services presented so much low lying fruit, the sector sucked up tech talent and capital from elsewhere. India’s SaaS sector in particular punched below its potential as a result. Software majors treated their services businesses as cash cows, deploying a small share to intellectual property assets. The 10 largest IT services companies had consolidated profits of $114bn in the past decade; 75 per cent of this was paid out via dividends and buybacks.

Their R&D expenditures are abysmally low.

The top five Indian IT firms had free cash flows of nearly $13bn in the 2023-24 financial year, according to HFS Research. And Infosys said on September 11 it had approved a $2bn share buyback offer — a week before the Trump order. Yet the R&D to sales ratio for India’s IT industry is abysmal: 0.88 per cent on average, according to a 2024 report by India’s Ministry of Corporate Affairs.

This flush of free cash flows has also not translated into meaningful increases in employee salaries that could have contributed to the emergence of a broader consumption class.

And now the industry faces disruption and job losses from AI technologies and the emergence of vertical integration of activities within multinationals in the form of Global Capability Centres. 

6. Finally, contrary to the impression created by recent public debates, Indians may be among the hardest-working people.

Since the 1970s, Indians on average have consistently worked more than 2,000 hours per year, even as other developed and developing countries saw this figure decline as productivity rose… As per an ILO database, Indians worked the highest average hours — at 56.2 hours per week or 11.2 hours per day for a five-day week — among 56 countries for which 2024 figures were available. In 2023, too, India topped the list (56 hours per week) among the 92 countries for which data were available… figures compiled by Our World In Data (OWID), a research publication based at the University of Oxford in the UK, showed that the average Indian worked 2,383 hours a year (or 6.5 hours per day, including weekends and national holidays, or 9.5 hours per day assuming 250 working days per year) in 2023. This puts India at ninth on the global list of the most time spent working. 

However, the increasing hours of work are not translating into higher productivity. In fact, labour productivity is not only stagnating but may even be declining in certain sectors. 

As per the Reserve Bank of India’s (RBI) KLEMS database — which analyses industry-level data focusing on capital, labour, energy, materials and services to measure economic growth, productivity, and efficiency — since 2019, at least nine industries have recorded a decline in labour productivity, including key sectors such as mining. Five other industries, including construction, have only seen marginal growth in the same period among a total of 27 industries analysed. This declining productivity is despite the share of labour’s income in the industries’ gross output remaining largely unchanged since 2019, barring four industries that reported small increases in labour’s income share. The 2024-25 Economic Survey found that while corporate profits grew by 22.3% in 2024, employment only grew by 1.5%, and the expenditure on employees fell to 13% from 17% in 2023. This indicates a preference for cost-cutting over workforce expansion, coupled with declining or stagnating wages, all of which are contributing to labour productivity and output lagging, apart from growing inequality.

In India’s case, the problem appears not to be working more hours, but rather working more productively. This productivity is a measure of both the use of capital and technologies, and most importantly, the quality of education imparted. The latter is arguably the biggest development and economic growth problem that we have, and one which requires prioritised, national engagement.

In conclusion, to make a high-level sense of the direction of the Indian economy, we can use a framework involving four categories of contributors to economic growth. The first consists of inputs like physical infrastructure and financial capital. The second involves the quality of human capital formation. The third consists of enabling factors like laws and regulations. And finally, there’s the state capability to formulate laws, execute contracts, and deliver quality public services. 

While much has been and is being done on the first and third, the second and fourth categories remain inadequately acknowledged and addressed. They are critical since broad-based economic growth and productivity enhancements cannot come without these contributors.