Industrial policy to overcome the China problem may require market making

There were two important industrial policy decisions taken by the US government last week that underline its commitment to securing the supply chains for critical minerals. They also offer important lessons about the direction of industrial policy in general. 

In the first, the US plunged into the rare earths industrial policy by taking a direct stake in a private rare earths firm and setting a price floor for its output. 

The Pentagon last week said it had become the largest shareholder in MP Materials, operator of the Mountain Pass facility in California, which sent shockwaves through the industry and drove the company’s shares to record highs. The US government committed $400mn of direct investment in the Las Vegas-based company and guaranteed a decade-long price floor for its output at nearly double the current market rate. MP, led by founder James Litinsky, is also building a domestic magnet manufacturing facility… The arrangement is part of the White House’s drive to break the Chinese dominance of critical minerals and strengthen the domestic supply chain. It is rare for the government to make direct investments in businesses, and critics argue this deal went too far... Rare earths are vital to modern manufacturing and to the magnets used in high-tech industries, from electric cars to fighter jets. Low market prices have deterred western rare earth projects from coming online.

... the US government had overnight become the dominant “influence” in the market for the rare earth neodymium-praseodymium (NdPr oxide). The price floor for NdPr, which is used in magnets, has been set at $110 a kilogramme, almost double the current market price of about $60. The guarantee covers all the NdPr that MP would produce over 10 years… MP does not currently make magnets at commercial scale, but the US has also guaranteed to buy roughly 7,000 tonnes of magnets annually for a decade that the company would produce using its own NdPr at the as-yet unbuilt facility. Industry veterans say this volume far exceeded US defence demand. Despite industry unease, the rare earths agreement has drawn bipartisan support in Washington.

In the second decision, the US announced anti-dumping duties of 93.5% on the imports from China of anode active materials (graphite), a mineral vital to batteries in EVs. This increases the total US tariffs on Chinese graphite to 160%, and comes on top of the US Energy Department’s $750 million loan last year to the Australian company Novonix to construct the largest synthetic graphite factory in North America at Chattanooga. 

Anodes are widely regarded as the most difficult part of the battery for the west to reduce its dependence on China, due to low prices and the near-complete dominance of Chinese groups over global supply. Chinese companies accounted for 95 per cent of the global anode market in the first five months of 2025, with South Korean companies, led by Posco and Daejoo, accounting for 2.7 per cent, and Japanese makers for 2 per cent, according to market information provider SNE Research. Tim Bush, a Hong Kong-based battery analyst at UBS, said efforts in Asia and North America to build a non-Chinese anode supply chain had been “undermined by US automakers’ unwillingness to underwrite the costs”. That reflects, in part, scepticism among battery and electric vehicle producers about the ability of North American producers to supply the battery-grade graphite they require.

The tariffs undoubtedly come at a significant incremental cost to the US consumers. 

Given the current paucity of non-Chinese anode suppliers, the additional import costs are likely to sting Asian battery providers, including those serving American EV manufacturers such as Tesla, General Motors and Ford, with the extra costs potentially passed on to US consumers. An average EV battery contains 50 to 100 kilogrammes of graphite, meaning the new tariffs could deprive battery and EV makers of up to 20 per cent of the value of generous federal production credits that were introduced by the Biden administration and which survived the recent passage of President Donald Trump’s “One Big Beautiful Bill”.

Incidentally, the US decision comes even as China finalised new restrictions on the export of technologies essential for making cutting-edge lithium iron phosphate (LFP) batteries, which have a low-cost chemistry composition. 

What do these two examples inform us about the emerging trends in industrial policy?

The earliest industrial policy actions were regulatory, primarily in the form of infant industry protection measures, by way of outright bans and prohibitive tariffs. Industrial policy of the kind pursued by the East Asian economies has focused on capital subsidies and fiscal concessions aimed at firms. The Chinese expanded their boundaries further with economy-wide subsidies, cheap capital, regulations like hukou that assured a supply of cheap labour, economies of scale from operating in a massive market, and generally provided capital subsidies and fiscal concessions on a scale never before seen. The result is that global manufacturing sectors are either overwhelmingly dominated by their manufacturers or they far outcompete their peers from other countries. 

This China problem means that traditional industrial policy instruments deployed by the East Asian economies are hardly sufficient. No country, except perhaps the US and that too only in certain sectors, can outspend the Chinese. 

There’s a chicken-and-egg problem. There’s very little or no domestic manufacturing of good quality. Domestic manufacturers cannot match the very low prices of Chinese imports. In the absence of an assured domestic market, no investor or firm is willing to put money. This gridlock perpetuates Chinese dominance. It must be broken.

Accordingly, the US government's decisions to invest public funds and assure a floor price (for rare earths), and raise prohibitive tariffs and give state loans (for synthetic graphite) go beyond traditional industrial policy. It effectively creates a market that can support domestic manufacturing. But this comes at a higher cost (of production compared to the Chinese imports), which is passed on to the domestic consumers who must pay higher prices. 

There are lessons for India. Traditional industrial policy must be complemented with market-making policies. Policies must be tailored to leverage the attractions of India’s massive market to incentivise domestic manufacturing. Apart from prohibitive tariffs and other barriers, this can be successful only with a mandate to purchase domestic produce. 

There are several product markets where market-making levers are available to the government. They include solar power and smart meters (where state-owned discoms are the monopsony buyers), telecom equipment (which are B2B markets), surveillance cameras and drones (which, while sold to consumers, also have a dual-use/security aspect). In these products, there’s a case for public policy to either mandate sales of domestic manufactured goods, or mandate a minimum domestic content, or raise disability bridging tariffs. Given the nature of these products and the volumes required, they are an unprecedented opportunity to catalyse a domestic ecosystem that covers the entire value chain involving component manufacturing, chip design, product design, and product manufacturing. Taken together, they have the potential to catalyse a serious deepening and broadening of India’s manufacturing base. 

However, in many products, Indian manufacturers will not be able to meet the domestic demand and that too with quality. Further, the domestically made goods will come at a much higher price. This gridlock can be broken only gradually. 

A prudent strategy would be to phase in the domestic purchases and domestic content requirements gradually. For example, the domestic manufactured equipment requirement can start with at least 5-10% in the first year, going on to 50-75% over five to seven years. Similarly, domestic content requirements can be progressively increased over time. Apart from facilitating the emergence of strong domestic manufacturing bases in these products, such phasing will also enable the domestic firms buying and selling these products to at least partially mitigate the high costs of domestic manufacturers. 

This strategy creates the risk of being captured by inefficient domestic vested interests. India’s memory of autarky is not too distant. The North East Asian economies mitigated this risk by building incentives around export competition and creating mechanisms to ensure that the market itself rewards the efficient manufacturers and weeds out the inefficient ones. Export focus would anyway be required to realise the economies of scale to become competitive with the Chinese manufacturers. However, given the political economy that emerges once such policies are put in place, this will be a big challenge for the government.

Such measures will invariably generate opposition from trade partners and likely violate WTO commitments. However, it appears that there is no other alternative. 

There’s a paradox with liberalised trade regimes. It’s like the bullies who favour the status quo. The US and the West created the WTO to institutionalise the prevailing global economic balance in their favour. The Chinese overcame this imbalance and turned it to their advantage. It’s therefore natural that they are now the most vocal supporters of the WTO. 

The WTO status quo is not in the interests of a country building its manufacturing base. The Chinese largely ignored the WTO regulations while building up its manufacturing industry, while enormously benefiting from the global market access it provided. This hard reality must be recognised while pursuing industrial policy. 

Weekend reading links

1. How is global trade changing due to Trump tariffs?

US tariff revenue surged almost fourfold from a year earlier to a record $24.2bn in May, while imports from China fell 43 per cent from the same month in 2024... China's exports are up 4.8% on last year despite a sharp drop in trade with the US.

There's also emerging evidence that Chinese firms are rerouting exports to the US through South East Asia and EU countries to avoid the high tariffs on Chinese exports.

The value of Chinese exports to the US dropped 43 per cent year on year in May, according to figures published by the US census bureau — equivalent to $15bn-worth of goods. But the country’s overall exports rose 4.8 per cent in the same period, official Chinese data showed, as the shortfall in trade with the US was offset by a 15 per cent increase in shipping to the Association of Southeast Asian Nations trade bloc and a 12 per cent rise to the EU... 

Separate research by Capital Economics estimated that $3.4bn of Chinese exports were rerouted through Vietnam in May, a rise of 30 per cent compared with the same month last year. Indirect trade through Indonesia also increased markedly, with an estimated $0.8bn rerouted in May 2025, 25 per cent higher than May 2024. Exports of electronic components such as printed circuits, parts of telephone sets and flat panel display modules to Vietnam were up 54 per cent, or $2.6bn, in May 2025 compared with a year earlier, Chinese data shows... Indian exports to the US jumped 17 per cent in May compared with a year earlier, while imports from China and Hong Kong rose 22.4 per cent according to Ajay Srivastava, founder of the Global Trade Research Initiative, a research group.

This is a good graphic on what products have been squeezed following the tariffs. 

2. The US equity markets are going about their merry ways overlooking the real costs that are introduced by the Trump tariffs. 

Despite the carve-outs and climbdowns, the US’s overall average effective tariff rate now stands at 15.8 per cent, according to calculations by the Yale Budget Lab — the highest rate since 1936 and an increase of more than 13 percentage points since Trump returned to office in January.

A big cause of concern is the uncertainty associated with Trump policies that are taking its toll on investments.

The most tangible consequence of the Trump tariffs so far is not supply chain reordering, but the sudden dearth of dealmaking, according to Persson of EY. A survey of dealmakers by PwC in May found that 30 per cent were either pausing or revising deals because of the uncertainty caused by tariffs. Among those pushed back amid the uncertainty included bids for Boeing’s navigation unit and an expected £4bn sale by buyout group Apax of insurance group PIB. The sudden slowdown flew in the face of investor expectations that Trump’s return to the White House would trigger a wave of M&A activity on the back of a deregulatory splurge, according to Josh Smigel, partner in PwC’s deals practice. As a result, Smigel calculates, private equity firms are holding about $1tn worth of assets that — absent the Trump uncertainty — could have been redeployed back into the market if planned exits had not stalled.

3. Has Israel won the battle, but only to lose the war?

Mr. Netanyahu’s relentless and unapologetic military response to the Oct. 7, 2023, Hamas-led attack that killed 1,200 people and took 250 people hostage has cemented the view of Israel as a pariah, its leadership accused of genocide and war crimes, and disdained by some world leaders. In opinion polls globally, most people have a negative view of Israel. In Gaza, the war against Hamas has taken a devastating toll, killing tens of thousands of people and leaving more than a million homeless and hungry. Much of the enclave has been reduced to rubble. Poverty and hopelessness are rampant... Israel’s actions have shattered a rock-solid, bipartisan consensus in the United States for defending Israel. Now, support for the country has become a fiercely contentious issue in Congress, the subject of angry debates and protests on college campuses and fuel for a surge in antisemitic incidents in the United States and around the world... Israel has created a new wave of global opinion critical of its goals and methods. And many Israelis now feel threatened while abroad, even as they are more secure at home... 

In a Pew Research survey of 24 countries around the world published last month, negative opinions about Israel have surged. In 20 countries, more than half of the people said they had an unfavorable view of Israel. In eight countries — Australia, Greece, Indonesia, Japan, the Netherlands, Spain, Sweden and Turkey — more than 75 percent held that view... Just 46 percent of Americans in the latest Gallup survey expressed support for Israel, the lowest number since the company began asking the question a quarter-century ago. A third of the respondents in the United States said they sympathized with the plight of the Palestinians, up from just 13 percent in 2003... Inside Israel, the decision to prioritize military victories over the return of the hostages has deeply wounded many people. And the violence has strained the good will of the country’s allies and neighbors.

4. Thrive Capital, founded by Josh Kushner, the brother of Jared Kushner is charting a new model of VC investing.

The approach Kushner has developed since launching Thrive 14 years ago: get close to founders, remain loyal through crises and concentrate funds in a small number of companies. Betting a billion dollars or more on a behemoth inverts the classic venture model: firms typically write dozens of small cheques in young start-ups; most fail, but the flops are more than offset by a few spectacular successes... venture capital has mutated from a cottage industry into an institutionalised asset class... The shift has left VCs with a choice: remain faithful to early-stage investing and hope for outsize returns, or scale up funds to meet increasingly massive private companies. Thrive is attempting to manage both, writing cheques for multibillion-dollar start-ups its team believe can still multiply 10 or 100-fold in value... Most VCs split funds between dozens of start-ups, but the vast majority of a Thrive fund will go to just 10-15. The firm has put 10 per cent or more of earlier funds to work in single companies, including workplace messaging app Slack, GitHub, Instagram and Stripe. Thrive first invested in Stripe, then valued at $3bn, in 2014, and has increased its stake multiple times, including investing close to $2bn last year... the firm has quietly shown intense fealty to founders during moments of crisis, such as during the boardroom coup that briefly ousted OpenAI’s Altman last year. Kushner was instrumental in returning Altman to the company after less than a week... 

Thrive’s rivals, including more established West Coast firms, dismiss the approach as closer to asset management. “We invest in companies, they trade in stocks. It’s like an ETF [exchange traded fund] for venture,” says a partner at one Silicon Valley firm. “But private companies are not stocks. You can’t get out when they start going down.” Speaking privately to the FT, some institutional investors question whether Thrive’s massive bets can ever deliver “venture-style returns”. Others say it is too soon to judge a group whose biggest investments have not yet cashed out. Thrive’s biggest portfolio companies, including OpenAI and payments start-up Stripe, have racked up massive paper gains. But until they go public or are acquired, profits won’t be returned to institutional investors in Thrive’s funds... The payout for Thrive and its backers would be enormous should Stripe, OpenAI, or defence tech company Anduril go public... Thrive has raised a total of $12.3bn, and now has almost $25bn under management, making it one of the largest VCs in the country.

Interesting that Mukesh Ambani has a 3.3% stake in Thrive capital as part of a consortium of investors! 

5. This is a very good graphic that shows how VCs are experiencing a squeeze in their cash flows.

Much the same could be said about PE funds.

6. Private equity in data centres.

The private equity giant Blackstone spent $10 billion in 2021 to acquire QTS, and has been pouring billions more into the company to help it expand its data centers... This largely unglamorous industry is critical for A.I. leaders to get right. QTS leases its facilities to companies like Amazon and Meta and supplies the electricity and water needed to power and cool their computers... Blackstone calls data centers one of its “highest conviction investments.” Blackstone is already one of the world’s largest owners of office buildings, warehouses and science labs, but it has sunk more money into data centers and related infrastructure than into almost any other sector in the firm’s 40-year history. All told, Blackstone has put more than $100 billion into buying and lending to data centers, as well investing in construction firms, natural gas power plants and the machinery needed to build them... (it) says it still sees strong demand from tech companies, which are willing to sign what they describe as airtight leases for 15 to 20 years to rent out data center space... 

Blackstone is not alone. Data centers are drawing a crowd on Wall Street — investment giants like KKR, BlackRock and Blue Owl have collectively plowed hundreds of billions into the industry. As investment firms announce larger and larger deals, one Wall Street executive says he jokes about “Braggawatt” deals, as data centers are typically measured by the wattage they use. The spending frenzy has created concerns about whether too many data centers are being built... The complexity and cost of running A.I.-focused data centers stem from the vast amounts of power they guzzle, which can be about 10 to 20 times as much per server or rack as general cloud computing. There is also the need to keep the centers operational 99.999 percent of the day, or the “five nines” in industry parlance. That equates to about five minutes of downtime all year for maintenance or to switch out servers.

7. China's dominance of clean energy technologies

China has also begun to dominate nuclear power, a highly technical field once indisputably led by the United States. China not only has 31 reactors under construction, nearly as many as the rest of the world combined, but has announced advances in next-generation nuclear technologies and also in fusion, the long-promised source of all-but-limitless clean energy that has bedeviled science for years.

And, buoyed by President Trump's policies, America retains leadership of fossil fuels.

This reversal is striking.

Americans created the first practical silicon photovoltaic cells in the 1950s and the first rechargeable lithium-metal batteries in the 1970s. The world’s first wind farm was built in New Hampshire nearly 50 years ago. Jimmy Carter installed solar panels on the White House in 1979... In 2008 the United States produced nearly half of the world’s polysilicon, a crucial material for solar panels. Today, China produces more than 90 percent.

This is a good description of China's manufacturing prowess.

Last June, the Urumqi solar farm, the largest in the world, came online in the Xinjiang Autonomous Region in China. It is capable of generating more power than some small countries need to run their entire economies. It’s hardly an anomaly. The other 10 largest solar facilities in the world are also in China, and even bigger ones are planned. The Chinese automaker BYD is currently building not one but two electric vehicle factories that will each produce twice as many cars as the largest car factory in the world, a Volkswagen plant in Germany.

Finally, a graphic that captures China's clean energy investments globally.

Chinese firms are building wind turbines in Brazil and electric vehicles in Indonesia. In northern Kenya, Chinese developers have erected Africa’s biggest wind farm. And across the continent, in countries rich with minerals needed for clean energy technologies, such as Zambia, Chinese financing for all sorts of projects has left some governments deeply in debt to Chinese banks. Since 2023, Chinese companies have announced $168 billion in foreign investments in clean energy manufacturing, generation and transmission, according to Climate Energy Finance, a research group.

8. Tim Harford points to a new paper by David Autor and Neil Thompson who use an "expertise" framework to explain the impact of automation and AI on jobs. Autor and Thompson pose a question

Would we expect accounting clerks and inventory clerks to be similarly affected by automation? There are several well-established approaches to analysing this question, and all of them suggest that the answer is “yes”. Back in the day, both types of clerk spent a lot of time performing routine intellectual tasks such as spotting discrepancies, compiling inventories or tables of data, and doing simple arithmetic on a large scale. All of these tasks were the kind of things that computers could do, and as computers became cheap enough they took over. Given the same tasks faced the same sort of automation, it seems logical that both jobs would change in similar ways. 

But that is not what happened. In particular, say Autor and Thompson, wages for accounting clerks rose, while wages for inventory clerks fell. This is because most jobs are not random collections of unrelated tasks. They are bundles of tasks that are most efficiently done by the same person for a variety of unmysterious reasons. Remove some tasks from the bundle and the rest of the job changes. Inventory clerks lost the bit of the job requiring most education and training (the arithmetic) and became more like shelf-stackers. Accounting clerks also lost the arithmetic, but what remained required judgment, analysis and sophisticated problem solving. Although the same kind of tasks had been automated away, the effect was to make inventory clerking a job requiring less training and less expertise, while accounting clerks needed to be more expert than before. 

The natural worry for anyone hoping to have a job in five years’ time is what AI might do to that job. And while there are few certainties, Autor and Thompson’s framework does suggest a clarifying question: does AI look like it is going to do the most highly skilled part of your job or the low-skill rump that you’ve not been able to get rid of? The answer to that question may help to predict whether your job is about to get more fun or more annoying — and whether your salary is likely to rise, or fall as your expert work is devalued like the expert work of the Luddites.

9. Two graphics that capture the essence and outcome of One Big Beautiful Bill (OBBA). One, stripped off all its hype, OBBA is a giant tax cut bill.

And its biggest beneficiaries will be the richest.

Analysis by scholars at the University of Pennsylvania suggests that Americans earning under $18,000 would lose $165 in 2027, or 1.1% of their income. By 2033 their annual losses would rise to $1,300 on average—about 7.4% for the group. The richest 0.1%, earning over $4.45m, would gain more than $300,000 in 2027, a 2.3% increase. Much of this comes indirectly, via changes to corporate taxes, which are usually assumed to benefit wealthier households who own stocks... Analysis of the House version by scholars at the University of Pennsylvania suggests that Americans earning less than $16,999 would lose about $820 a year—a 5.7% reduction in median income for that group. The richest 0.1%, earning more than $4.3m, would gain $390,000, a 2.8% increase.

10. Driverless EV trucks are already operational in a coal mine in Inner Mongolia. (HT: Adam Tooze).

Yimin is one of the five largest open-cast coal mines in China. During peak season, it used to require about 300 trucks, operated by around 1,200 drivers working shifts around the clock, to transport coal to processing sites, and soil, sand and rocks to dumping grounds. But managers said the mine faced a shortage of drivers. Dangerous driving conditions led to high attrition rates, compounded by declining interest among younger generations in pursuing this profession. “Truck drivers face exhausting workloads that often lead to health issues,” said Yimin mine director Shu Yinqiu. The solution came earlier this year with a fleet of 100 photovoltaic-battery-powered, self-driving trucks. They represent the world’s largest deployment of autonomous electric mining trucks, highlighting China’s resolve to upgrade its traditional industries with advanced technologies, as the nation grapples with a shrinking labour force and an ageing population...

Key partners in the project include Huawei Technologies, Xuzhou Construction Machinery Group, State Grid and the Beijing University of Science and Technology. Now, instead of a thousand-man crew, just 24 people, divided into four teams, are needed to operate the 100 new trucks. Staff monitor and control the vehicles from the comfort of a remote control room, where live-feed videos and real-time traffic information are displayed on multiple screens... As of September, the China National Coal Association (CNCA) estimated there were over 1,500 automated mining trucks in China. It predicted that number would triple to 5,000 by the end of this year and exceed 10,000 by 2026... A fleet of 100 unmanned trucks could save coal mine operators 40 million yuan (US$5.6 million) in driver salaries annually, according to CNCA estimates.

11. Major announcement for the establishment of a PCB and Copper Clad Laminate (CCL) manufacturing facility by Syrma SGS Technology at Naidupeta in Andhra Pradesh with an investment of about Rs 1800 Cr and in partnership with South Korean company Shinhyup Electronics Ltd. The project is expected to be commissioned by 2026-27 and can avail incentives under the GoI's Electronics Component Manufacturing Scheme (ECMS). In 2024, the GoI had imposed a 30% anti-dumping duty (ADD) on bare PCBs to boost domestic production. The Indian PCB market was valued at $6.2 bn in 2024 and is estimated to grow by a CAGR of 16.4% from 2025-33. 

12. Spain wants to avoid the costs of being part of NATO, while wanting to access its benefits. It was the only standout against accepting the goal of 5% of GDP defence spending target by NATO members at the recent NATO summit. At the same time, as FT reports, one of its defence firms, Indra, which is 28% owned by the Spanish Government, is benefiting from NATO defence spending. 

In April, the group was given a role in 12 European Defence Fund research and development projects and made the leader of one involving radars. Its executives were in Ukraine last month pitching their wares... In the air, Indra is Spain’s lead participant in Europe’s flagship fighter jet project, the Future Combat Air System, a sometimes prickly partnership with Airbus, which represents Germany, and France’s Dassault Aviation.

13. India's derivatives market, and how Jane Street abused it before SEBI cracked down.

In December 2020 — when Jane Street first set up its Mumbai arm — the monthly turnover of futures and options markets on the National Stock Exchange had reached nearly $300bn, from just $134.7bn four years earlier, and by December 2024 stood at $512.7bn. This became a fertile terrain for Jane Street. Between January 2023 and March 2025 the firm netted an overall profit in India of about $4.3bn, Sebi said in its order on Thursday.

14. Using dupes of expensive brands appears to be a trend in the US, as seen from the ongoing fight between Lululemon which has sued Costco of copying at least six patented clothing designs, including its popular Scuba hoodie and Define jacket.

Once seen as embarrassing parsimony, buying knock-offs has become a fashion statement of its own. Egged on by hashtags, TikTok videos and media articles, customers are leaning into the fun of finding cheaper but still good alternatives, turning the search for dupes into a public treasure hunt. Nearly half of US consumers surveyed by analytics firm First Insight said they had tried a product specifically because it was a “dupe”, and 70 per cent of shoppers who make more than $150,000 said they were more likely to try a dupe than other private label goods...

The warehouse store’s $20 sweatshirt mimics the ornamental stitching and pouch pockets of Lululemon’s Scuba offering, which sells for six times the price. And Costco’s dupe of the Design jacket mimics an unusual line of curved stitching across the back. Lululemon contends in its lawsuit that those specific details violate the “trade dress” patents that it has registered over the past two years, as well as a trademark on the colour description “tidewater teal” that it applied for one day before filing its claim that Costco had “unlawfully traded upon Plaintiffs’ reputation, goodwill and sweat equity”.

Interestingly, US laws allow considerable flexibility in the interpretation of design patents.

US rules protect makers from infringement claims if the similarities are based on function rather than distinctive design. The warehouse group could also try to turn the dupe craze to its advantage by arguing that consumers are unlikely to be misled into believing that they are buying a Lululemon original. Costco’s products are clearly marked with either the Kirkland brand or the manufacturer’s name. Despite the publicity, most patent attorneys expect the dispute to settle, as Deckers’ first Uggs lawsuit did last year. Each side has too much to lose from a trial. Costco could be on the hook for gigantic monetary damages, while “if Lululemon were to lose, it would be open season” for other duplicates, says Josh Gerben, a DC trademark attorney.

15. Good story on how Tamil Nadu's industrial development strategy has brought about broad-based regional development across the state.

16. AI in primary health care.

Shishu Mapan, an artificial intelligence (AI) tool trained on over 30,000 infants, built by scientists at the Wadhwani Institute for AI, a non-profit that develops AI-based solutions for social impact. Using a short, arc-shaped video while the newborn is undressed and laid on a cloth sheet, the app estimates the infant’s weight and growth metrics, which eliminates the need for scales or guesswork... AI-powered tools like Wadhwani AI’s app could become frontline essentials, capable of transforming child health outcomes where the system often falls short. It also eases the burden on frontline health workers, who often struggle to keep up with high demand in rural areas... AI-powered tools like Wadhwani AI’s app could become frontline essentials, capable of transforming child health outcomes where the system often falls short. It also eases the burden on frontline health workers, who often struggle to keep up with high demand in rural areas... AI-powered tools like Wadhwani AI’s app could become frontline essentials, capable of transforming child health outcomes where the system often falls short. It also eases the burden on frontline health workers, who often struggle to keep up with high demand in rural areas.

17. Interesting that even as the overwhelming majority of the world has no confidence in Donald Trump, India stands alongside Israel in having the highest confidence!

Wonder what actions of Trump warrant such confidence?

18. Patent cliffs facing pharma companies.

Keytruda... cancer medicine is one of the world’s best sellers, earning Merck $29.5bn in sales last year... In 2028 Keytruda’s patent ends... Drugs worth about $180bn of revenue a year are going off patent in 2027 and 2028, according to research firm Evaluate Pharma, representing almost 12 per cent of the global market. Bristol Myers Squibb and Pfizer are also facing 2028 patent expirations for top-selling drugs. 

Interesting aspects of the Pharma industry.

While all innovations can be patented, the pharma industry suffers from patent cliffs in ways that others such as the tech industry do not. This is mainly because the key active ingredient in a drug is covered by one main patent, which is hard to invent around, and chemical formulas are relatively easy to copy. Sampat of Johns Hopkins says the median number of patents per drug is around three to five, not the hundreds or thousands that cover, for instance, an iPhone. “So any given patent expiring doesn’t matter all that much for something like the iPhone, as it would for a drug,” he says. Also unlike the iPhone, few patients are loyal to their brands and healthcare systems are eager to cut costs by moving to generic versions quickly after they are released. Many countries have laws allowing pharmacists to automatically swap out branded prescriptions with generics.

19. The problem with the rail ticket subsidy of Indian Railways

This monopoly network transports 13 million people every day and its non-premium services are heavily subsidised. According to the railway minister, the cost of travel per km by train is ₹1.38 but passengers pay only 73 paise, a subsidy of 47 per cent. Though the government dishes out large sums for passenger subsidies, part of the gap is supposed to be covered by freight services and premium air conditioned passenger services. The problem with this cross-subsidy policy is that railway freight services have been steadily losing share to road transport over the decades and its profits are not enough to cover the losses from passenger services. As for AC services, some of which make money in some years, they account for a minuscule 5 per cent of overall passengers. The proliferation of low-cost airlines and growing air connectivity — ironically, this, too, is government policy — is likely to diminish demand for this segment, despite the investment in semi high-speed premium Vande Bharat service.

20. The NPAs on bank loans to MSMEs are at historic lows.

Gross NPAs in the system have touched a new low of 2.3 per cent of loans, with a sharp drop in NPAs in MSMEs. Gross NPAs in MSMEs declined from 6.8 per cent in 2022-23 to 4.5 per cent in 2023-24 and further to 3.6 per cent in 2024-25. NPAs in the MSME sector have historically been of the order of 9 per cent or more... bankers have found innovative ways, such as the Trade Receivables Discounting System (TReDS), to finance MSMEs... The TReDS book was about ₹2.7 trillion, or 10 per cent of the MSME book, in 2023-24. It cannot explain the current NPA level of 3.6 per cent on the entire MSME exposure. The NPA level in the Emergency Credit Line Guarantee Scheme (ECLGS) is 5.6 per cent. Recall that the ECLGS was introduced during the pandemic in May 2020 in order to facilitate additional lending to MSMEs and prevent a secular collapse in the sector on account of a crisis of liquidity. The eligibility conditions were pretty stringent. Only MSMEs that were solvent prior to the onset of pandemic were meant to qualify. The loans granted under ECLGS in the period 2021-23 amounted to ₹3.68 trillion or 12 per cent of loans outstanding to MSMEs in 2024-25. If gross NPAs on the ECLGS loans were 5.6 per cent and NPAs on total MSME loans are 3.6 per cent, that makes the performance on the remaining 88 per cent of MSME loans truly impressive.

21. Finally, a graphic below on the spectacular reduction in the price of green energy sources since 2010.

The role of domestic corporate groups in manufacturing value addition

There are two distinct stages to manufacturing: product assembly and localisation of components. Industrial policy measures tend to focus on product manufacturing. But countries generally start at the lower part of the manufacturing Smile Curve, with low-value-added activities like mounting components on the PCB (SMT), product assembly, testing, marking, and packaging (ATMP). Domestic value addition comes only with component manufacturing. Then countries move slowly up the value chain with research and development, product design, and branding. 

Having boarded the manufacturing train through the production-linked incentives (PLI) scheme and contract manufacturers, the next step is to localise component manufacturing. But given the vice-like grip of the Chinese component manufacturers (arising from their massive volumes, low prices, and deep integration with the global value chains), even countries like Vietnam have struggled to make much headway in component manufacturing. 

In this context, The Ken has an article on how Tata Motors is tackling the electric vehicle (EV) manufacturing challenge. Central to its EV ambitions is Tata Autocomp, an associate company of Tata Motors (its shareholding is 26%), established 30 years ago to meet the component needs of Tata’s commercial and passenger vehicles. Accordingly, Autocomp delivers EV components to Tata Motors. It had revenues of Rs 13,600 Cr in FY24. Autocomp does this through a “constellation of joint ventures” and subsidiaries.

With over 60 manufacturing facilities globally and a slew of strategic joint ventures, Pune-based Tata Autocomp claims to be one of India’s largest end-to-end EV-component suppliers. Be it battery packs, motors, seats, or even the body of Tata’s EVs, a significant share of it all flows through Tata Autocomp. 

And this reflects in Autocomp’s revenue, about half of which comes from partnerships with overseas OEMs. 

Tata Motors claims a very high degree of localisation in its Tata Harrier EV model’s drive trains, battery packs, motors, etc., though it’s more likely that many of the components themselves are subassemblies of imported subcomponents. It’s important to make the distinction between components made by Autocomp’s partners globally and in India. It must be critically scrutinised as to how much of the claim of domestic value addition is actual manufacturing and not an assembly subterfuge. 

In this context another Ken article has the graphic below which shows how India’s top EV makers mostly rely on outsiders for critical motors (and most certainly battery).

The first article describes the strategy behind Autocomp’s business model.

Building every capability in-house for an EV delays the time to market, says Danish Faruqui, chief executive of Fab Economic, a US-based automotive and gigafactory consultancy, making it difficult to capitalise on evolving demand. “As a result, many auto companies across the globe, including Autocomp, are looking to get external capabilities and turning the EV business into something like an iPhone supply-chain delivery, where the best components are sourced from specific suppliers and the final product is delivered to the customer, bypassing the R&D cost.”

Further, joint ventures give more agility for companies in the short-term. “With technology changing so fast, and cell chemistries and the EV market still evolving, it wouldn’t have been right to invest and develop these in-house while starting from scratch,” Faruqui said. And the other best part of a partnership? The freedom to walk away. Tata Autocomp has already shut down multiple joint ventures, such as Taco Sasken Automotive Electronics (its partnership with Bengaluru-based product engineering firm Sasken Technologies to design and develop automotive electronic products) and Seco Powertrain (a collaboration with the Korean clutch maker Seco Seojin).

The Tata Group is uniquely placed to leverage synergies across the Group companies.

The Tata group’s many companies, spread across 10 sectors, give it an unmatched ability to cross-sell and co-develop. For instance, Tata Motors taps TCS for software; Tata Chemicals for cell chemistry; Tata Power for charging infrastructure; Tata Agratas for cell production; Tata Technologies and Tata Elxsi for design, development, and simulation; Tata Autocomp for hardware; and Tata Capital for loans.

But there are challenges arising from the corporate structure of the Tata Group that are limitations on realising synergies and efficiencies. 

Tata Autocomp helped Tata Motors crack the EV code. But here’s the thing: the latter still can’t call the shots on when and how to scale. It’s forced to run at the pace of its other in-house entities… The group is trying to align its moving parts through the “One Tata” approach, a push for internal synergy across entities… But that’s easier realised on Powerpoint slides than on shop floors. “In-house agreements and tech-building often don’t turn out as desirable as designers and engineers want them to be for production,” he added. Essentially, there’s no option to penalise an in-house supplier for any delays… If four to five in-house entities are not performing at the same pace, the whole value chain gets hit.”

This challenge of internal misalignment isn’t unique to Tata Motors. The group’s own super-app experiment, Tata Neu, is a cautionary tale. Despite access to multiple retail brands, the app has failed to deliver a seamless consumer experience, largely due to poor cross-entity integration. Even within Tata Motors, the complexity of managing over 100 entities breeds both independence and inefficiency. 

Tata Motors, through Tata Autocomp, may be doing to the creation of EV component manufacturing in India what Apple did with iPhone for mobile phone manufacturing in China, albeit in a completely different manner and at a much smaller degree and scale. 

Apple pursued a model of an OEM actively engaging (including embedding its engineers and purchasing some of the equipment) with its contract manufacturers and component makers by building capabilities and handholding them to ensure very high quality. Tata Autocomp is following a vertical integration model of building in-house capabilities through joint ventures and buying up component makers, and then gradually moving their manufacturing (and hopefully design) into India. 

This kind of engagement may be essential to develop a component manufacturing ecosystem. It may be necessary that either a large OEM or a vertically integrated corporation provide the de-risking required to attract component manufacturers. By financing and ensuring captive demand, it derisks and encourages component makers to relocate from China. A PLI top-up will be a bonus. 

This highlights the importance of large companies in establishing a manufacturing base that is at scale, globally competitive, and strategically significant in key sectors. Apart from EVs, Tatas are now also spearheading India’s mobile phone manufacturing pursuit. Tata Technologies accounted for 26% of iPhones made in India in 2024 and is expanding rapidly. 

There’s a strong case that India’s large corporate groups, Ambani, Adani, Mahindra, Birla, etc., could emulate the Tata Group’s EV strategy. The Ambani Group may be well placed to play an important role in petrochemicals, telecommunication equipment, and green energy technologies. Similarly, with the Adani Group in solar cell manufacturing, other green technologies, and smart meters; the Mahindra Group in automobile/EV manufacturing, and so on. 

Only these groups have the heft to overcome the coordination problems, financing deep pockets, and risk-aversion to be able to manufacture at a scale that’s attractive enough to break away from the vice-like grip of Chinese manufacturers, and shift component manufacturing in India. Even with the most generous industrial policy and the favourable geopolitical tailwinds of diversifying out of China, there are hard limits to how much industrial policy and market dynamics on their own can go in relocating component makers from China. At the least, the engagement of deep-pocketed corporate groups like Tatas can expedite domestic-scale manufacturing in India. 

For countries like India, without any large domestic OEM and without the unique circumstances behind Apple’s engagement with China, local corporate groups may be valuable assets to succeed with the objective of deepening and broadening their manufacturing base.

In this context, there’s always the risk of crony capitalism. Already, there are strong trends of business concentration across sectors in the Indian economy, driven by the large corporate Groups. But this risk cannot overlook the large body of evidence from across the world about the value of large domestic corporate groups in national economic growth. I have blogged on the emerging domestic monopolies in infrastructure and digital markets (also this), the role of family-owned businesses in driving commercial scale across East Asian economies, and the inevitability and need for monopolistic firms in certain sectors. 

In this context, a new MGI report highlights the role of a few large firms in driving national economic productivity growth. It finds that a small number of firms contribute the lion’s share of productivity growth and that the most productive firms find new ways to create and scale new value. I’ll blog separately on this report.

Making best use of the smart meter opportunity

I have blogged here about the installation of smart meters by electricity utilities across India. Now that smart meters are a reality, this post offers a few suggestions to make the best use of them.

Before that some updates on the status of smart meter installations. Till March 2025, 222.3 million smart meters have been sanctioned, of which 138 million have been tendered and awarded, and nearly 22.9 million have been installed. 

The state-wise sanctions and installations show that the uptake has been surprisingly weakest among those states with the better discoms. 

The target is to install 250 million smart meters by 2027.

India has adopted the TOTEX (total cost of ownership) model in smart meters, whereby the private firm invests upfront in the entire smart meter infrastructure (meters, communication and telemetry system, software etc.) and the utility makes monthly payments, including for billing. In this model, infrastructure contractors like Adani, GMR, GPIL etc., contract with smart meter manufacturers like Schneider, Secure, HPL etc., and deliver metering as a service for the contract period (60-96 months). 

Now that smart meters are on us, some observations on ensuring their effective installation and utilisation.

1. The value of smart meters comes from comparing the energy input and billing for each feeder before and after their installation. It allows discoms to identify energy losses, feeder-wise, and take remedial action to lower them. Accordingly, smart meters can help reduce aggregate technical and commercial (AT&C) losses in distribution. 

In order to achieve this objective, it’s important to ensure that smart meters are installed feeder-wise in a manner that covers all services (including the distribution transformers) under an 11 kV feeder so that it becomes possible to conduct reliable end-to-end energy audits. Like with unmetered agricultural services, scattered smart metering of services within a feeder makes accurate energy audit difficult, if not impossible. 

Another level of prioritisation would be to cover the highest energy-consuming feeders. This will ensure that in the initial stages, the implementation bandwidth is focused on a few high-value supply centres instead of being spread out wide and thin. Besides, it will also ensure that the discoms start to realise meaningful financial benefits from the implementation.

2. I blogged about the assumptions behind the installation of smart meters to reduce energy losses and improve the quality of supply. 

One, the smart meter's telemetry is always synchronised to the grid and it reliably delivers information. Two, the smart meter itself is operational and does not suffer from technical problems and failures. Three, the smart meter is not tampered with or even bypassed at the consumer level. Four, even if the data is acquired and processed to make available actionable decision support, the discoms have the capabilities to undertake rigorous energy audits and monitor and enforce them. Five, the discoms are able to disconnect errant connections.

There is daunting state capability and political economy challenges to be overcome to meet these assumptions. Therefore, these assumptions should be salient and prioritised considerations during the installation of smart meters. They should be monitored closely and rigorously for compliance. 

If these assumptions are not complied with, there’s a strong risk that the smart meter will end up like using an expensive computer to do the functions of a typewriter. 

3. The nature of electricity distribution and energy audit makes it ideally suited for a technology intervention like smart metering. If, even after its implementation, the discoms are unable to substantially reduce losses, it would point to failures in either enforcement (not able to disconnect services despite knowing the sources of leakages or pilferages) or implementation (bad installation, poor telemetry, tampering with the smart meter etc.) failures. 

There’s a strong risk of such failures. The speed and scale of this roll-out must be calibrated with the capabilities of the systems on both the supply and discom sides. Careful thought and effort must go into testing and iterating the robustness of the smart meters themselves, their installation requirements (safeguards like earthing and other protections, network connections etc.), stabilising the telemetry, and billing and accounting systems of the smart meter network. 

The worst outcome would be for the vendors to erect the meters without the requisite safeguards and auxiliaries and claim their first bills. The poor quality of installation will invariably result in various kinds of meter failures. Given the difficult and widely dispersed contexts coupled with the speed of the roll-out, the vendors will have enough excuses to blame the government and get away. It’s a recipe for contractual failures. 

For these reasons, it may be useful for each discom to test and refine the installation in a few feeders for a few months before scaling them up. Such a stabilisation pilot approach should be made a mandatory requirement before full-scale roll-out. 

4. I’m not sure about how effective smart meters will be in serving as a service control device, i.e., in disconnecting services. Their primary value will be in communicating information about energy flows and consumption. This means making available the right kind of actionable information for different levels within the discom and government. 

Unfortunately, there are too many examples of technology experiments in development that have struggled at the last mile by failing to provide actionable information. Given this and the common institutional contexts and problems, it may be useful to standardise the reporting formats for energy audits across 3-4 levels to ensure that the system provides the basic first-order actionable information for each level. 

5. The 250 million smart meters offer a great opportunity to create an invaluable end-to-end domestic manufacturing ecosystem. Apart from the product itself, there’s a big opportunity to promote the emergence of an Indian System on Chip (SoC) that drives the smart meter. Given the biggest constraint to breaking into the tightly controlled chip design market is access to deployment use cases in large volumes, smart meters are an unmatched market-making opportunity. 

This requires very high-level coordination involving different Ministries of the Government of India, to integrate this strategy with public procurements, mandate product standards, expedite regulatory enablers and permissions, etc. For example, in the case of smart meters, the meter supply and O&M contractors (6-9 year TOTEX contracts) could be mandated to supply domestically designed and manufactured meters. This mandate could be phased in over time, with the initial supply being restricted to 10% of all installations and going up each year. The support provided by the Ministry of Power for smart meter installations under the Atmanirbhar Bharat Scheme could be leveraged to catalyse domestic chip and meter design. The government coordination could bring together the OEM/contract manufacturer, and chip and product design firms. 

At the least, this public policy mandate on smart meters must create a couple of globally competitive and competing domestic smart meter manufacturers. 

6. The smart meters also offer a great financial market opportunity. It presents a low-risk, long-term, assured revenue-generating regulated asset. Accordingly, in Europe, with its deregulated supply market, smart meters have been attractive investment destinations for private equity firms. In the existing TOTEX contracts, the financial risk is borne on the balance sheet of the discom in so far as it directly commits to repaying the meter costs (even if recovered through the monthly bills). In forthcoming tenders, as the market gets derisked, contracts could be structured to distinguish between the capex (meter) and opex (billing) risks, with only the latter being fully borne by the discom.