Derisking the public funding of innovation

 The Government of India have recently concluded the first round of applications for accessing innovation funding through the $10.5 bn Research Development and Innovation Fund (RDIF) scheme. It has chosen the wholesale funding approach of financing alternative investment funds, development finance institutions, non-banking finance companies, and focused research organisations (FROs), which in turn give loans to or take equity in private firms and startups engaged in innovation. 

What model of innovation funding generates the greatest bang for the buck in terms of achieving the primary objective of catalysing innovation? 

The additionality from public funding of innovation is that it is risk-tolerant, concessional, and patient compared to commercial capital. It can be blended with commercial capital to derisk and thereby expand the envelope of investible startups. This can be done through two approaches. 

One approach would be to fund startups by investing in a fund of funds (i.e., to be an LP in a startup fund). This approach would require only diligencing the fund and leaving it to diligence the individual investments. It would leave the portfolio building responsibility to the fund, albeit within the broad sectoral scope defined by the public fund. A flip side is that the fund could play it safe and avoid the riskier bets, and use the public funding to merely juice up its returns. 

The other approach would be to co-invest with commercial investors. This would require significant in-house expertise in the due diligence of startups. In theory, it would ensure tight alignment with the specific policy objectives of the government and target investments to the riskier startups. But in practice, public decision-making processes are likely to detract from the objectives. 

There are two possible objectives associated with public funding of innovation startups. One, to de-risk and expand the envelope of investible startups. Two, fund specific products or solutions that are considered strategically important. 

In terms of these two objectives, the policy question is which one of the two approaches generates the greatest bang for the buck. An analysis of the experience of deployment of public funds through the two approaches reveals that the fund of funds approach is generally superior, except where it is required to target strategically important technologies or firms. 

Here’s the list of a few such public funds from Europe and elsewhere, assessed in terms of their mobilisation ratios (or private capital raised per unit of public fund). 

Here is a comparison of the two approaches in terms of various considerations. 

Clearly, if the priority is to maximise the volume of private capital mobilisation and to catalyse the VC market, the fund-of-funds approach is superior. However, if the priority is to catalyse a few strategically important firms or technologies where private capital is reluctant, then co-investment is useful.

In emerging markets like India, where the domestic VC funding for genuine technology innovation (as opposed to imitating technologies already mainstreamed elsewhere) is negligible, there’s a risk that a fund-of-funds approach, at least for some years, could end up catalysing and unlocking the supply of venture capital finance, rather than the supply of innovative technologies. This risk will persist even with tightly prescribing the segments where the funds can invest, given the practical difficulties in monitoring, let alone policing, them. And commercial funds will have the incentive to prioritise the use of public funding to maximise the derisking of their capital instead of expanding the envelope of investible startups. This would be the long-route to catalysing innovation (or catalysing innovation primarily by catalysing risk capital funds). 

In the circumstances, it is important to consider the option of government co-investing directly in startups with private investors to provide sufficient de-risking of their investments. Such derisking can be done by public funds taking the first-loss equity buffer, offering higher liquidation preferences for private capital, capping public returns, anti-dilution warrants, offering buyout options at modest returns, equity conversion as subordinate capital or at a discount, etc. The graphic below captures the comparative de-risking effect of public capital. 

The challenge with co-investment, of course, is the administration of such funds by a public entity and getting its governance right. While there are good examples of co-investment like the Israeli Yozma fund, European Innovation Council, and the British Enterprise Capital Fund, they have been exceptions to the generally adopted fund-of-funds approach. 

In the circumstances, DFIs are a good platform. While DFIs have traditionally focused on infrastructure financing and industrial growth, they are well placed to directly support co-investments by governments by deploying the full suite of instruments mentioned above, based on the specific nature of the investment proposal. 

I am not sure how many of the above really invest in risky startup-driven innovations as opposed to investing in promising established SMEs in emerging industries. 

As a note of caution, and this applies to the RDIF’s DFI window too, the experience of India’s infrastructure DFIs in actually derisking and catalysing infrastructure market segments has not been encouraging. Instead, they have most often been found chasing the market. See this and this.

Weekend reading links

1. On the origins of infrastructure financing in the UK.

In the summer of 1858, Britain’s parliamentarians soaked the curtains in the Palace of Westminster with lime chloride in an attempt to counter the “Great Stink” emanating from the river Thames. It failed. The prime minister demanded the Metropolitan Board of Works construct a sewerage system, legislating to allow the board to raise £3mn. This was to be repaid by a three-penny levy on all London households for 40 years. By 1900, the municipal bond market in England was about 50 per cent of the market for UK government debt. Today it is just 4.7 per cent. Back then, most infrastructure projects were not only financed by the private sector but also backed by hypothecated cash flows from money raised locally. For example, the debt for the 1892 Elan Valley Aqueduct to pipe water into Birmingham was funded by an increase in water rates on businesses and households there. So the residents of, say, Manchester were not funding infrastructure elsewhere.

2. India is marching ahead on adoption of electrotech.  

3. Germany is the poster child for central bank independence.

4. Spurred by surging data centre loads, electricity demand is rising sharply in the US.

Data centre power demand will surge from 34.7 gigawatts in 2024 to 106GW by 2035, according to BloombergNEF, a research group, equivalent to more than 80mn homes. Overall US electricity demand is forecast to jump by a quarter by 2030 and by 78 per cent by 2050, compared with 2023, according to consulting firm ICF... Between January and November, the latest month for which data is available, the cost of electricity for residential customers in the US increased by 11.5 per cent.

5. Most Indian state boards focus on memory and conceptual understanding, as opposed to analysis and application. 

6. At 16%, US after tax corporate profit margins are at historic highs.

7. Nick Bloom, Paul Mizen, Gregory Thwaites et al have an estimate of the costs of Brexit (HT: Adam Tooze)

These estimates suggest that by 2025, Brexit had reduced UK GDP by 6% to 8%, with the impact accumulating gradually over time. We estimate that investment was reduced by between 12% and 18%, employment by 3% to 4% and productivity by 3% to 4%. These large negative impacts reflect a combination of elevated uncertainty, reduced demand, diverted management time, and increased misallocation of resources from a protracted Brexit process.

8. Some data on students and migrants from Asia in the US. Even as enrolment of Chinese students has declined, that of Indian students has taken off since 2014. While Chinese students were mainly for the UG programs, Indian students have been mostly for PG studies. (HT: Adam Tooze)

Indian students dominate the workforce of computer science workers. 

Similarly, Indians dominate physicians and Filipinos dominate nursing. 

9. India has one of the lowest free floats in its equity markets.

10. Japan's public debt at 237% of GDP is way off the charts, and considerably constrains Sanae Takaichi's room to manoeuvre. 

11. Tej Parikh writes that UK's biggest economic challenge is its political instability.

The UK doesn’t have a productivity puzzle. The causes of Britain’s weak underlying growth are well known and discussed ad nauseam. Why policymakers aren’t delivering is the bigger conundrum. Instability undermines business investment, hiring and planning decisions, and absorbs the political bandwidth that could be used to address lacklustre growth... Uncertainty and slow growth weaken the appeal of UK assets too. Pound sterling and British stocks have underperformed relative to peers over the past decade. Long-term UK government borrowing costs have remained elevated compared with other G7 nations, having come under frequent selling pressure thanks to fiscal mishaps... For all the turbulence, Britain remains attractive. London remains the leading European destination for foreign direct investment. The country’s strengths in finance, university research, tech and life sciences are draws. Cheap assets add to the allure.

12. Fascinating account of China's largest hotel chain, H World, with 12,700 hotels, adding 1700 hotels in 2025, aiming for 20,000 hotels by 2030, and took in 8.3 m guests during last year's Lunar New Year holiday. As a reference, Marriot has 10,000 hotels worldwide. 

H World was launched in 2005 by Ji Qi, co-founder of online travel booking site Ctrip, after he took inspiration from the multi-brand French group Accor... Its Hanting Inn brand, a fixture of the streets of major Chinese cities, often charges well below Rmb300 ($43) a night, while mid-range Ji Hotels typically cost slightly more. H World has several other brands, while internationally it also operates German brand Steigenberger. Its franchise model mirrors an approach that other Chinese businesses have used to expand in consumer sectors from bubble tea to fast food. For the “vast majority” of its franchised hotels, H World retains a degree of control, sending in hotel managers on its own payroll under what it calls a “manachise” approach. In the US, “typically the franchisees decide almost everything . . . we think in China this would not work,” said He. “A lot of people would not abide by your rules.”

13. VC funding for India's tech startups has fallen 73% since 2021, which also coincided with the rise of Byjus.

14. Fascinating account of the Jeffrey Epstein world. 

15. Shruti Rajagopalan has a good essay on India's AI ambitions, and more. This on regulation.

The EU went first and went heavy, a binding cross-sector AI Act with tiered risk categories, compliance obligations, and a governance apparatus that could employ a small city. China took the authoritarian-efficiency route. Regulate fast, regulate specifically, and make sure the state retains control over what models can say and do. The US, characteristically, has been light touch at the federal level, leaving governance to a patchwork of executive orders, state laws, and vibes. India, with these guidelines, has landed somewhere interesting, closer to the US in its instinct to avoid a standalone AI law, but far more deliberate in articulating why it is choosing not to regulate horizontally yet. The framework’s core bet is that India’s existing legal infrastructure (the IT Act, the Digital Personal Data Protection Act, sectoral regulators like the RBI and SEBI) can handle most AI risks if enforced properly and updated where needed...

Do not regulate the technology itself, govern its applications through the regulators who already understand those domains. Build incident databases so you learn from failures instead of pretending to prevent them through preemptive compliance theater. Use “techno-legal” mechanisms (standards, system-architecture-level controls, provenance tools) so that compliance scales without armies of auditors. Create sandboxes so regulators can see what actually goes wrong before writing rules. The explicit preference for “innovation over restraint,” listed as a core principle, rejecting the EU’s precautionary posture. Both committees looked at Brussels and decided that regulating AI the way you regulate pharmaceuticals, before you know what the side effects actually are, is a bad trade for a country where AI adoption is still nascent and unevenly distributed... The liability framework it recommended is graded. The regulated entity remains liable to consumers for any losses, but first-time failures where the entity followed prescribed safeguards and reported promptly would not automatically trigger full supervisory penalties. A rigid liability regime that punishes every probabilistic error will cause institutions to constrain AI capabilities to the point of uselessness.

16. Shyam Saran writes about Marco Rubio's speech at the Munich Security Conference, describing it as an "unabashed white, racist manifesto". 

His remarks celebrated the history of conquest, exploitation, barbarity, and even ethnic cleansing, which has marked the history of Western imperialism and colonial empire-building across Asia, Africa, and Latin America. He wants this to be a source of pride and inspiration, not something to “atone for purported sins of past generations”. What is perplexing is that the history of the world after the Second World War, which is often described as an American era, is instead seen as a period of Western decline: 

“But in 1945, for the first time since the age of Columbus, it [i.e. the West] was contracting. The great western empires had entered into terminal decline accelerated by godless communist revolutions and by anti-colonial uprisings that would transform the world and drape the hammer and sickle across vast swaths of the map in the years to come.” Anti-colonial uprisings, which would include our own against British colonialism, are not celebrated as struggles for freedom and human dignity but as evidence of the abdication of the Western will to rule. Strange that this should come from a representative of a country that is celebrating 250 years of its own successful war of independence against British colonialism.

17. AI is leading to changes in software industry business models, resulting in less predictability and more uncertainty.

Software companies have, for decades, sold their wares on a “per seat” basis, where an employee gets unlimited use of a package of tools. Think of the traditional Microsoft 365 licence... In a world of AI “agents” carrying out duties autonomously, that model makes less sense. The unit of account will no longer be users but tasks completed, queries undertaken, and data “tokens” used. Sticky, predictable, year-round software-as-a-service revenue — the kind of thing that private equity firms love because it makes companies easier to load up with debt — may become an endangered species. Some are already embracing the post-seat era. Snowflake, a data management software maker, charges based on consumption, as does Databricks, an unlisted hotshot valued at $134bn, according to Crunchbase. ServiceNow is one of many working on hybrid models, where monthly fees meet pay-as-you-use add-ons... consumption-based pricing... Software companies’ predictability was an asset that contributed to high valuations.

18. Europe telecom industry facts of the day

Europe has more than 44 mobile operators that each have more than 500,000 subscribers compared with eight in the US and just four in China, according to industry group Connect Europe’s State of Digital Communications 2026 report.

19. China high-speed railway facts.

China’s railways have in recent days been ferrying about 20mn passengers a day, with half a billion train trips expected over the 40-day lunar new year period... Nearly three-quarters of passengers will travel at speeds of greater than 200kph, streaking across the country in the white and silver high-speed trains that have become a defining symbol of China’s industrial might. In December, China reached 50,000km of high-speed rail, enough track to circle the globe, compared with 8,500km in the whole of the EU as of 2023. Just over two decades after it was launched, the network now links 97 per cent of cities with populations of more than half a million... China opened its first high-speed passenger line in 2003 between Qinhuangdao and Shenyang in the north-east, with speeds of 200kph. The World Bank estimated in 2019 that China spent about $17mn to $21mn per kilometre on high-speed rail... Another 20,000km of track is planned by 2035... 

China has benefited from a combination of relatively cheap land, enormous scale, standardised designs and a permissive regulatory environment, experts say. The country’s rail project is now overseen by China State Railway Group, a huge state-owned enterprise that operates the network and helps fund new line development. The group plans to invest Rmb520bn ($75bn) this year... The group’s total liabilities have mounted to Rmb6.4tn. China State Railway reported a modest profit of Rmb11.7bn in the first three quarters of 2025, following several years of losses during the Covid-19 pandemic. Analysts said profits from the freight network helped offset losses from passenger high-speed rail. Local governments typically share the burden of building and operating the tracks. But many are struggling with their own shaky finances following the pandemic and the collapse of the property market.

20. Finally, the US Supreme Court has struck down the tariffs imposed by Donald Trump under the International Economic Emergency Powers Act. Also this. 

Demand and supply side constraints to rapid growth - the case of medical education

Business Standard points to a paradox in India’s medical education - post-graduate seats lying vacant amidst a shortage of specialist doctors. 

It points to data from the Health Dynamics of India report by the Ministry of Health and Family Welfare. 

Community health centres (CHCs) in rural India face an almost 80 per cent shortfall in specialists. As of March 2023, just 4,413 specialist doctors were available against a requirement of 21,964 across 5,491 CHCs in 757 districts, each centre serving an average population of nearly 160,000 per centre… The number of PG medical seats for MD and Diplomate of National Board (DNB) courses rose 157 per cent to 80,291 in 2025 from 31,185 in 2014, with the government planning to add another 2,000-3,000 seats by 2029… India generates over 123,000 MBBS graduates per year…

As a result, filling the seats on offer remains a challenge. After two full rounds of NEET-PG 2025, over 18,000 seats remained vacant, forcing the National Board of Examinations in Medical Sciences (NBEMS) to keep slashing qualifying cut-offs… candidates with zero percentile scores became eligible for counselling in further rounds… Zero percentile means candidates who scored the lowest in a test, or that none of the other candidates scored less. Such relaxations have become routine. Cut-offs were reduced to zero percentile in 2023, 2024 and, earlier, throughout the Covid years… Vacancies are most pronounced in private and deemed universities, which account for nearly 10,000 unfilled seats annually… 

Specialities such as cardiology, radiology, gynaecology, orthopaedics, and general surgery traditionally being the most sought after. On the other hand, seats in non-clinical subjects such as pathology, anatomy and biochemistry have vacancy rates of 50-70 per cent, as many candidates prefer to drop a year rather than opt for these disciplines. However, doctors say even traditionally sought-after clinical specialities are now seeing gaps.

The article posits several explanations for the paradox, including high fees, poor quality in private colleges, mandatory service bonds, regional imbalances in seat distribution (half the seats being in the South), declining attractiveness due to safety and other concerns, lack of commensurate increase in MBBS seats, and so on. 

Instead, I am inclined to argue that these are all symptoms of more basic problems. While there is no specific evidence, I feel that this highlights two first-order problems on both the demand and supply sides. 

On the demand side, the vacancies reflect the poor quality of education in general and basic medical education in particular. What does it say about the quality of the MBBS education when nearly 2.5 lakh students could fill only 62,000 out of the 80,000 PG seats after two rounds of counselling, even at cutoff scores of 235-276 out of 800? 

On the supply side, the sharp 2.6-fold increase in PG seats over a decade appears to have compromised quality. Does India’s medical education ecosystem have sufficient supply of good-quality teaching personnel (human capital) and well-equipped hospitals to adequately train the 2.6X increase in PG seats? I’m not sure. 

This is the point made on multiple occasions in this blog about India’s problems of poor quality of human resources and the low capital base of supply across sectors that constrain sustained high economic growth. Fundamentally, high economic growth can be sustained only if there’s a broad enough capital base (physical, human, industrial, financial, and institutional) to support that growth. 

This is also central to the larger point about high economic growth rates requiring the foundations of a broad base of human development and economic growth.

India's non-financial corporate bond market trends

Deepening of the non-financial corporate bond market has been the big endeavour of financial market reforms in India. I have blogged here and here, pointing to the limits to bond market financing of infrastructure sectors. 

In the absence of readily available longitudinal data, I ran some queries on ChatGPT to estimate the landscape of India’s non-financial bond markets, issued by public sector units and private corporates, across different bond tenors, the respective shares of infrastructure and private placements in each tenor category, and comparison across countries. The data used are from RBI/SEBI, SIFMA (the US), AFME/ECB (Europe), and BIS/Brazilian Central Bank. 

India’s bond market is dominated by private placement, skewed towards the short tenor, and issuance by the PSUs. A positive feature is the high share of infrastructure issuances.

The same result for the US showed up as below. The US corporate bond market is mostly through public placement, dominated by private corporates, and with the majority in longer tenors. The volume below is in Rs lakh crores.

The table below compares across India, the US, Europe (including the UK), and Brazil. India stands out for its dominance in private placements, a high share of public-sector issuance, and a very low share of long-tenor bonds. The private placement market is dominated by AAA and AA bond issuances, thereby squeezing out the vast majority of potential issuers. All three trends must reverse to deepen and broaden the bond markets.

How does the table look for tenor greater than 10 years?

Putting all of them together, we get the following charts. Brazil does a very good job of mobilising long-term infrastructure funds, in absolute value (see latter tables). 

As a percentage of GDP, India does well in non-financial corporate bond issuance. Its challenge, as aforementioned, is with maturity depth and structural composition. 

The total infrastructure sector bond raising is estimated in the range of Rs 2.2-2.4 lakh Cr, of which about Rs 1.3-1.4 lakh Cr is with tenor less than 5 years (55% raised by PSUs), and Rs 80,000-90,000 Cr with tenor more than 7 years (65% by PSUs). The table below has the approximate values in Rs crore.

Below is the graphical representation of the estimates.

As can be seen, the usual suspects - roads, power generation (thermal and renewables), power transmission, railways, ports, and airports - make up most of the bond issuance. 

A comparison with Brazil is instructive. Brazil has managed to develop a deeper and broader market for private corporate bond raising, including in infrastructure. In 2011, it introduced the “Debêntures Incentivadas” (Infrastructure Debentures) to finance infrastructure projects. It offers tax exemption on interest income, is mostly issued for 7-15 years, and is used in transport, energy, sanitation, and renewables. 

This is a summary of the comparison of the infrastructure bond markets in India and Brazil.

What have been the trends over the past ten years? Disturbingly, the share of private corporates in bond issuance has been falling, declining from over half to about a third. This may have been due to the focus on public capital expenditures through NHAI, Railways, PFC/REC. The share of long tenor bonds (>7 years) has risen only slightly from 8% to about 13-14%. Private placement route has remained stubbornly in the 90-92% range.

Given all the above, what are the policy takeaways?

Policy actions are required on at least three fronts. One, prioritised efforts to expand public placement, thereby ushering in greater liquidity and deeper markets. A practical intermediate pathway would be to promote the listing of existing privately placed bonds. 

Second, diversify the issuer base away from PSUs by encouraging private infrastructure developers to issue long-term bonds by offering policy support through credit enhancement schemes, partial guarantee mechanisms, takeout financing, etc. The DFIs (NIIF, IIFCL, and NaBFID) should have a specific mandate to derisk infrastructure finance instruments. 

Third, increase the tenor of bonds by expanding the use of partial credit guarantees, first-loss facilities, and blended finance models. This should also be complemented by demand-side measures to expand the pool of long-term buyers like pension funds, retirement savings accounts, and insurance funds. 

On all these, the policy actions will be in the form of numerous small steps to simplify disclosures, documentation, listing requirements, listing processes, standardisation of bond structures, improvements on rating transparency, and so on. It must be complemented with initiatives like credit enhancements and guarantees, increasing the proportion allocated to long-term bonds for pension funds and insurers, revisiting risk-weights on long-term bonds, etc.

For references, I blogged here that India could take the lead in demonstrating how DFIs can work on both the demand and supply sides to de-risk infrastructure projects and crowd in long-term capital, respectively. I blogged here on how to reduce the cost of capital for foreign investments in the infrastructure sector, here on how to revise the credit rating framework in general, and here on de-risking and lowering the cost of capital of bank loans and de-risking the use of guarantees to crowd-in bank financing. All of these ideas are consolidated in this long paper.