Supervising what you cannot inspect

by Maninder Singh Juneja and Renuka Sane.

In traditional financial regulation, supervisors are able to inspect the thing being regulated. For example, a scorecard by a bank was usually a short list of factors that one could comprehend. These factors could be traced to key documents within the bank, banks had a stated rationale for why they were being used, and there was some common sense in making weighted averages. The scorecard itself was static - the same weighted factors were applied over a period of time. Inspecting this process of constructing the scorecard and using it for management decisions was how trust was built. The AI world is different. Models are rented rather than built, change continuously, and behave probabilistically. They cannot be fully inspected even by the institution deploying them let alone by the regulator.

How should we then think of regulation?

One approach is to intensify the traditional approach where regulators demand more explainability, more documentation, and more validation. This will drive up the costs of compliance. But more importantly, this is ill suited to the new world where the technology changes rapidly, where the bank does not control the AI it uses, where there is no clear artefact that the bank can give the supervisor such as a model or a document (Board of Governors et al 2026). When regulators push traditional approaches, banks will respond by choosing AI models which are easy to document rather than the ones best for them, or defer AI deployments altogether. All these are unhappy consequences. What we need are policy makers who understand the live systems of the new world.

In this article, we analyse these emerging problems from first principles. We start from scratch, understand the landscape of market failure in the world of AI in banking, and think about how regulators can grapple with this world.

Our key idea is that AI systems resist replicability. Regulatory strategies that demand replicability will flounder or choke technology deployment. We suggest the regulatory standard applied should be AI deployments that are "supervisable" - the outcomes should be observable, they should be attributable to causes, and reversible by humans.

How AI is actually deployed

Before we get to the puzzles faced by regulators, we need to describe what AI in banking is. This categorisation is not unique to banking. Banking is simply where their consequences are regulated. Five modes span the range.

AI as Tool: AI augments a step, a search, a calculation, a first draft that the human controls. The human is cognitively engaged.

AI as Collaborator: The human and AI co-produce iteratively, and the human participates at every stage.

AI as Recommender: AI generates scores or options, and the human makes the decision. This is the classic human-in-the-loop, but credible only if the human can meaningfully interrogate the recommendation and not degenerate into cognitive surrender.

AI as Preparer: AI does the work and the human signs off. The approval here is closer to a check by a supervisor rather than a re-derivation.

AI as Autonomous executor (agentic AI): AI executes autonomously inside guardrails and the human monitors on exception. Emerging forms of agentic AI for banking run from single-task agents (a payment released, a service query resolved) to multi-step workflows and customer-facing agents that transact.

A central issue here is the true (de facto) role of the human. A reviewer who approves a thousand recommendations a day is not overseeing a model, the model is overseeing her. A feature of any deployment is the measured divergence between the model recommendation (and estimated uncertainty) vs. the human decision. Managers of banks will need to worry about relapses of human behaviour inside the organisation, a bit like how hospital managers worry about bad behaviour by doctors within their organisation.

What AI does to market failure in banking

We now shift gears to look at the standard knowledge on market failure in banking. Regulation may be justified when (and only when) there exist market failures which cannot self-correct fast enough, and there is adequate state capability in banking regulation to be able to correctly identify them and intervene. AI's distinctive feature is that it can cure several classic failures. Better default prediction reduces credit rationing, better fraud detection cuts deadweight loss, richer risk assessment lets banks serve customers they previously could not price, AI advisors help customers avoid some malpractices by the bank. But there are also some new problems that are anticipated.

Information asymmetry: This happens in banking at two levels: borrower-to-lender (adverse selection, hence credit rationing) and firm-to-consumer (product complexity, hence mis-selling). AI narrows the first through alternative data and may widen the second. On one hand, the customer armed with AI can see through many things proposed by the bank which are not in her best interest. But the consumer cannot observe why they were shown a product, offered a price, or steered toward a specific insurance plan. The sales process itself becomes more opaque. Personalised pricing approaches first-degree price discrimination, extracting consumer surplus. A single flawed model can mis-sell to millions simultaneously, converting isolated conduct failures of the pre-AI world into a big correlated event. And redress weakens when a denial comes from a model the firm itself cannot explain.

Systemic externalities from shared infrastructure: Each bank chooses its models, data sources and vendors to optimise its own performance. When multiple entities choose the same ones, the sector's exposures become correlated, which is a cost no individual bank prices in. We list the vulnerabilities below:

Correlated model risk: banks on similar models and the same foundation providers respond identically to an event. The regulator, at the system level, has to manage what happens when institutions move together, because no single entity has experience of such behaviour or of the impact synchronisation adds.

Third party concentration: one vendor's failure propagates everywhere at once. India has already run this experiment, when a ransomware attack on one shared technology provider knocked roughly three hundred cooperative and regional rural banks off the payments network.

Correlated cyber breach: shared stacks mean one exploited vulnerability is every institution's vulnerability. AI lowers the attacker's costs (automated vulnerability discovery, deepfake social engineering) and adds new attack surfaces (data poisoning, model inversion, prompt injection against agents that can move money).

Runs on banks at level 3. The bank runs of old were a queue on the pavement. Then we got to Silicon Valley Bank where over a weekend, customers took away money from the bank. Now we can be at level 3: autonomous agents managing customer cash can turn a shared signal into a self-reinforcing run at machine speed.

The various market failures listed above behave differently across the five modes of AI use. For example, when a human constructs the offer, opaque pricing can get contained. However, when an agent personalises autonomously at scale, this may become severe. Systemic correlation is moderate when AI advises and severe when fleets of similar agents act simultaneously. Any regulation that grades by model type alone, or by use case alone, misses half the object. The next question is the mode of regulation itself.

Regulatory strategy

Regulation can work in two ways. Process-based regulation is ex ante: it prescribes how the firm must operate, defines required controls, mandates oversight, validation standards, limits on autonomy. Outcome-based regulation is ex post: it prescribes ends, fair treatment, solvency, and judges results, leaving the choice of methods to the firm.

Outcome-based regulation is the efficient default. It is technology-neutral, so it does not ossify as methods change; it lets firms find the least cost route to compliance; and it does not require the regulator to understand the firm's production function better than the firm does. But it has important preconditions: the outcome must be observable and measurable; it must be attributable to the firm and, ideally, to the cause; and the harm must be reversible or compensable. Process regulation is the right departure from the default when those preconditions fail and where outcomes are unobservable, harm is catastrophic or irreversible, or damage manifests only systemically or with a lag.

Traditional doctrine treats these preconditions as given: examine the activity, choose the mode. This does not work for AI. Whether an AI deployment's outcomes are observable, attributable, and reversible is an engineering choice, which needs to be settled at design time. We suggest that regulation should mandate observability. This makes it possible to have an "outcome-based supervision" model. There are three ways to ensure observability.

1. Telemetry implies that the institution keeps a track of every decision such that the system records which version of itself it was using, what information it was given, and any time a person stepped in to overrule it. This makes outcomes attributable. That way if a certain group starts getting more (or less) approvals than before, the organisation can evaluate what caused the shift - was it the model, or the group itself. Such a capability is being mandated elsewhere in the world for similar use-cases (European Parliament and Council of the European Union, 2024). The regulator should also consider if it wants to set a minimum common telemetry standard.

2. Boundaries and rollback include putting caps on what the system is allowed to do, rolling out new updates to just a small number of cases first (instead of everyone at once), and having a tested plan for switching back to the older version if something goes wrong. This ensures that if a bad update slips through, it only affects a small slice of decisions.

3. Probes make bias observable. One way is "paired testing": one submits two applications that are exactly the same except for details that hint at things like someone's race or gender, and see if they get treated differently. One can also compare approval rates against the company's own normal levels. Together, these checks can catch an unfair credit model in just a few weeks, instead of waiting years to see who actually pays back their loans. One can also check rejected applicants against credit-bureau data to see which ones got approved by someone else, and how they fared.

These are similar to the idea of decision receipts that record which rules were applied to which facts and in what sequence for every decision made by a government or public system (Srivastava, 2026).

Over time, the supervisor should also build a repertoire of its own test cases drawn from incidents, complaints and examinations across the system and run it against every material AI deployment, much as stress-test scenarios are run against every balance sheet today. This will ensure that what surfaces in one institution becomes a probe for all others. It thus allows the regulator to set its own observability layer.

Process regulation is then reserved for the harms that are systemic, correlated, or irreversible at machine speed. For such events there need to be protections such as circuit breakers that halt things before they cause damage, limits on how much the systems can do on their own, model diversity so they don't all fail the same way, and rehearsed back-up plans.

What follows for the supervisor, the board, and the customer

For the supervisor: The unit of examination shifts from the model to the deployment, and the examiner's question shifts from "show me the validation report" to "show me the behaviour": what boundaries were set, what exceptions were thrown, what overrides were exercised, how far the system drifted from its baseline. Supervisors will also need to find the intellectual clarity to avoid a wide variety of extraneous claims about regulation of AI, e.g. the push for economic nationalism which has nothing to do with market failure.

For the board: A board cannot certify systems it cannot inspect or understand. Its role is to govern the framework which includes the limits on what the system is allowed to do, and making the rules for when a decision must be escalated to a human. The board then needs to continuously check the exceptions and overall performance.

Exception-handling should also be written into policy. If a problem stays unresolved beyond a defined size or time limit, it automatically gets escalated to the board. Internal auditors should double-check that the numbers are real. The board sticks to this audited framework and does not inspect the system directly.

If a board is asked to approve something they can't understand, they will default to saying no. But if you let them govern the limits and the exceptions instead, they can say yes.

The customer: The customer becomes part of the supervisory architecture. The widening firm-to-consumer asymmetry has a structural corrective the pre-AI world lacked: the customer now has AI too. Mainstream assistants abroad have begun connecting directly to users' accounts. If product terms (rates, fees, eligibility criteria) are mandated to be structured and machine-readable, the customer's own AI does the comparing, the explaining, and the policing of mis-selling, continuously and at zero supervisory cost. The redress channel weakened by opacity is restored the same way: an adverse decision should carry its reason to the customer, and what would have had to be different for the answer to change. This restores the ability to contest. The same asymmetry that AI widened, AI-equipped customers can close, but only if regulation hands them the data.

An example

Consider a debt-collection example. When borrowers fall behind on payments, the bank has to decide how to chase each overdue account. Contact methods differ in cost. Automated SMS and IVR (the automated phone system - "press 1 to pay") are cheap; having an employee actually phone the customer is expensive. So the bank builds a model that allocates accounts: cheap automated nudges for most, and the costly human call reserved for the accounts where the model predicts that talking to a person will actually "cure" the account (get it back to paying). Human calling works better but it costs more. The model is rationing an expensive resource to where it thinks it'll pay off.

With AI, the collection systems would retrain continuously or run reinforcement-style optimisation against a live reward (cure rate per rupee spent). The model would be moving on its own, faster than the review cycle, toward a target that is a proxy for what the bank may actually want. Under the conventional "inspect-the-artifact" approach, the model is checked before deployment, signed off, and reviewed on a schedule (say quarterly or annually). However, if AI is an active optimiser, a quarterly review discovers deterioration only after it has touched thousands of accounts. It may drift toward a mis-specified target in a way no one can read off the model itself. Further, if the collections model is a shared vendor product, or several banks fine-tune the same foundation model on similar data, they all go bad the same way at the same time when borrower behaviour shifts. In contrast, in a "supervisability-built in by design" approach, a small slice of accounts is deliberately kept on the previous allocation method, running live alongside the new model. This allows comparison between the old and the new in real time.

Every decision should keep a record of a few things: which version of the AI was used, what action it chose, which group of customers it was dealing with, and how things turned out in the end. This is how it helps: Say the AI's overall success rate is going up, but one particular group is quietly getting fewer phone calls from real people and doing way worse because of it. With all that recorded, that gap shows up right as it's happening. You can compare it against what's going well elsewhere, trace it back to a specific version of the AI, and undo it that same day, because the older setup is still up and running as a backup.

The company should never have to dig into the AI's inner "thought process" to realize its behaviour has taken a bad turn. None of this replaces the normal testing that the firm would do anyway. The difference is that here, the ability to observe what's happening, pin down what caused it, and shut it down are all built right into the system from the start, instead of being things you have to go do to the AI afterward.

Conclusion

AI deployments sit uneasily in conventional mores of banking regulation. The intelligence is rented, the behaviour is probabilistic, and harms can move at machine speed. We suggest that the response should not be more inspection of what cannot be inspected, nor blind faith in outcomes that arrive too late. Efficient regulation will emerge from observability, attributability, and reversibility. Regulators should require banks to build these.

References

Board of Governors of the Federal Reserve System, Federal Deposit Insurance Corporation, and Office of the Comptroller of the Currency. "Supervisory Guidance on Model Risk Management." SR Letter 26-2. April 17, 2026. https://www.federalreserve.gov/supervisionreg/srletters/SR2602.pdf.

European Parliament and Council of the European Union. Regulation (EU) 2024/1689 (Artificial Intelligence Act), arts. 12, 26(6), and Annex III(5)(b). https://ai-act-service-desk.ec.europa.eu/en/ai-act/article-12.

Srivastava, Manish. "Digital Governance Needs Decision Receipts." Episode 73 of Big Ideas. XKDR Forum, June 1, 2026. Podcast, video, 13:06. https://youtu.be/WFX4ITb9yok

Trust in the Era of the AI-Informed Customer

by Maninder Singh Juneja.

A patient grappling with a longstanding eye problem was diagnosed with MGD (Meibomian Gland Dysfunction) by an eye specialist. To make sense of the medical jargon, she photographed her meibography report, the gland scan, and uploaded it to an AI, which confirmed the diagnosis and the line of treatment. A few weeks later, wanting to avoid buying shades of cosmetics she already owned, she uploaded a photograph of her stock and of the items she was about to buy to the same AI. Instead of commenting on the shades, the AI told her the waterproof eyeliner she had used for years was blocking the pores of the gland, like wax in a drain, and making her dry eyes worse. She stopped, and this one change brought immediate relief. The AI connected across domains, unasked. Neither the physician nor the cosmetics counter had connected the two. The ophthalmologist saw the eye. The counter saw the product. AI saw the person. Neither was wrong. Each was trained to look at one domain. Every institution serves its own; AI serves the customer. This is not a failure of individuals but a structural shift with strategic implications. In this article we examine these implications.

Verification in markets

For as long as markets have existed, buyers have had to trust sellers, because there was no way to verify their word, or the effort and cost were too high. The economist George Akerlof spent many days puzzling over the nature of the vegetable retail market in Delhi. In 1970, he offered new insights into the market failure that arises when the buyer cannot verify what the seller knows. He later won the Nobel prize in economics for this work which helps us understand asymmetric information.

The market economy combats asymmetric information using substitutes, brands, professional licensing, statutory audits, regulators, and guarantees. For example, the customer trusts the bank's brand name instead of reading the fine print, or the doctor's medical degree instead of evaluating the diagnosis. Half a century of institutional architecture in financial services is built on this logic.

The use of AI has brought about a foundational shift in trust. AI has made verification in many situations quite feasible. Customers who once accepted substitutes for trust in institutions can now check a claim, compare alternatives, or challenge a recommendation on an inkling of doubt. The barriers to specialist access, cost, time, language, jargon, have all collapsed at once. The result is not an end of trust, but the pillars on which trust rests have changed.

How it impacts the brand

The brand is an informational shortcut, one of Akerlof's substitutes for verification. It stood in for the customer's inability to verify. The bank's or insurer's reputation stood in place of the customer's understanding of provider soundness, and the terms and conditions of the contract. AI changes that, but the picture is complicated, because the brand connotes two things at once: A promise of what we will do for you, and an aspiration of who you become or which tribe you join by choosing us.

The promise of what the brand will do is now more verifiable. The customer's AI checks every promise before purchase, searching the wider internet, cross-referencing user reviews and triangulating from multiple public sources. The bank that claims to be customer-first has its complaint-resolution record extracted from annual reports or instantly summarised from X. The insurer that promises easy claims has its claim-rejection ratio surfaced against peers. Promises that survive verification strengthen the brand. Promises that do not are revealed in seconds.

The aspiration is not in the product, it lives in the customer. People buy Apple products partly because Apple-ness signals something about themselves. Customers bank with a private bank not for any major service benefit but for who else banks there. People ride Royal Enfield partly because owning one says something no specification sheet captures. AI audits the promise. It cannot interpret the tribe.

The split deepens in the AI-to-AI world, where the customer's AI transacts with the institution's AI. The customer only experiences the outcomes, settlement speed, dispute rate, complaint-resolution time and median application-to-approval. It does not watch advertisements. In this perspective, brands need to invest in advertising that bolsters tribal loyalty, but advertising that is supposed to bolster the promise is now less important.

How it impacts labour

Like the brand, the professional is going to be hit by the AI wave. The professional of the old world was, by training, organised around the domain. The cardiologist gave the right answer to the heart in front of her, not to the medicine cabinet or the financial situation at home. The mutual fund agent recommended within his manufacturer's product set, not against the seven existing funds in the customer's portfolio. The AI can see the landscape comprehensively and the professional has to now compete with it.

In most domains the gap between the bottom and the top decile of professionals has been wide. AI compresses the gap from below. The advisor who tests his recommendation with AI before delivering it catches the portability clause he had not considered. The relationship manager who has the model argue the customer's case against his own pitch will close more often. Every word the professional says can now be cross verified; the smartest professionals will go up against 'unsophisticated consumers' with more respect.

In recent research, Brynjolfsson, Li and Raymond (2023) studied 5,179 customer-support agents at a Fortune 500 software firm and found that average productivity rose 14% with access to a generative AI assistant. Within this overall average, novice and low-skilled workers improved by 34%, while experienced and highly skilled workers showed minimal gain. By this reasoning, AI-powered unskilled labour will be tough competition against skilled practitioners.

The pattern is visible in India also. At one NBFC, the productivity of fresh-college LAP underwriters rose 40% with AI assistance, while experienced underwriters showed minimal gain. The bottom of the labour quality distribution comes closer to the top.

As the floor rises, the implications are twofold. First, the customer's worst experience disappears, and with it the customer's reason to switch providers. Second, the professional's competence becomes the table stakes. They have to now provide what AI cannot supply: Trust earned over time, judgement under ambiguity, the willingness to take a customer-friendly call when things are not going well with the business. These are the qualities that live on the aspiration side of the brand, and they are the real moats of professional competence.

When the asymmetry reverses

So far the shift has run one way, AI in the customer's hand against the institution. The same architecture runs the other way too. Institutions have always known what customers did. But AI interactions reveal something deeper, what the customer considered doing. The questions asked, the scenarios tested, the decisions abandoned. These are cognitive traces, and they sit closer to intent than anything an institution has had access to before. If applied to underwriting, pricing or customer acquisition, they create a new informational advantage that did not exist a year ago.

The trust consequence runs deeper than the privacy one. Behaviour is what the customer did once it was done. Intent is what she rehearsed before she was ready to be seen. A customer can absorb the knowledge that her behaviour was logged. When she learns the institution priced her, or declined her, on the strength of her question, the breach is of a different order. The first asymmetry was about information she did not have. This one is about information she did not know she was giving. This may lead the customer to stop being candid with the one tool that was working for her, because she now suspects it is also working for the firms. The independence this piece began with is the first thing she loses when the architecture turns around. We will have to face a new world of consumer protection complexity, going beyond the simpler questions of data privacy.

What then survives in trust? Not the part that rested on the customer's inability to check. What survives is what AI cannot manufacture. Judgement under ambiguity, the call no model will take responsibility for. A relationship proven over time, the banker who backed the customer through a bad cycle and was proved right. The human presence in a hard moment, the advisor who delivers difficult news with care. None of this can be read off a document, so none of it can be verified, and so none of it can be commoditised. The trust that survives is the trust that was never about information in the first place.

The institutions that Akerlof described were built on substitutes for verification. Those substitutes served a purpose. They filled a gap the customer could not fill herself. That gap is closing. What remains when the substitute is no longer needed is the thing the substitute was always standing in for. Genuine expertise, honestly applied, in the customer's interest. The institutions that had that all along have nothing to fear from the informed customer. The institutions that were selling the substitute will find, quietly and permanently, that the customer has stopped calling back.

The scarce asset is the question

If one risk is that the institution reads the customer's question, the other is that she asks a wrong question. While AI can reduce the information asymmetry, friction has not entirely disappeared. It rests on arriving at the right question to ask. Speed and convenience applied to the wrong question produce a confident wrong answer faster. The customer who compares home loans on interest rate alone misses the prepayment clause. The customer who has already decided to switch insurers asks questions that confirm the decision. The informed customer is powerful. The misinformed customer with AI is powerfully wrong.

The question itself is the unclaimed opportunity. No bank will build a question set that surfaces its own prepayment clause weaknesses. No insurer will build a question set that exposes its own claim-rejection ratio. This needs independent actors: non-profits, researchers, consumer bodies, who can make a GitHub for the questions consumers should be asking financial institutions.

What the Boardroom should debate

Every earlier wave of technology was institution first. Computing, internet, mobile was adopted and absorbed by the organisation and then passed onto the customer on their terms. AI exploded in consumers' hand, 100 million users in two months. The Board is now governing businesses where customers will know as much if not more than the organisation.

The first institutional response to AI has been operational, adding chatbots, analytical tools, dashboards, voice bots, service automation, that are good cost-saving initiatives. The strategic question is the viability of the business model itself.

What happens when the customer arrives informed? Which elements of the value proposition survive verification? Which revenue streams depend on customer ignorance or high search costs? Which promises would survive an AI audit? Which parts of the sales process assume an information asymmetry that no longer exists? These are not technology questions for the CTO. They are business-model questions for the Board.

The institutions that emerge stronger will not be those that adopted AI fastest. They will be those whose value remains after verification becomes cheap.

References

Akerlof, George A. (1970). "The Market for 'Lemons': Quality Uncertainty and the Market Mechanism". Quarterly Journal of Economics, 84(3), 488-500. https://www.jstor.org/stable/1879431

Brynjolfsson, Erik, Danielle Li and Lindsey R. Raymond (2023). "Generative AI at Work". NBER Working Paper 31161. https://www.nber.org/papers/w31161

Reuters / Similarweb (2023). ChatGPT user-growth figures. https://www.reuters.com/technology/chatgpt-sets-record-fastest-growing-user-base-analyst-note-2023-02-01/

Reserve Bank of India. Annual Report of the Ombudsman Scheme. https://www.rbi.org.in/Scripts/AnnualReportPublications.aspx

Maninder Singh Juneja is a partner at True North. He serves on the boards of Pine Labs, Nivara Home Finance and Integrace, and has previously served on the boards of Niva Bupa Health Insurance, Federal Bank Financial Services and HomeFirst Finance. The author thanks participants at a talk at XKDR Forum for myriad good ideas, and Ajay Shah, Renuka Sane and Aditi Mascarenhas for comments on earlier drafts.

Prepaid Payment Instruments: How has regulation impacted market outcomes?

by Amol Kulkarni and Renuka Sane.

Regulatory interventions often function as de-facto industrial policy, by favouring certain business models over others, effectively picking winners and losers. In this article we present an episode in recent Indian history where regulatory decisions of the Reserve Bank of India on prepaid payment instruments (PPI) significantly influenced competition and product evolution. PPIs are instruments that facilitate the purchase of goods and services, financial services, remittance facilities, etc., against the value stored therein (RBI, 2021). They saw a phenomenal growth of more than 100 times in volumes and more than 25 times in value between September 2012 and November 2024 when volumes had reached 584.78 million and the value Rs. 192.14 billion (RBI, PSI). By this time, the Unified Payments Interface (UPI) volumes and value were more than 25 times and 100 times that of PPI volumes and value. The data suggests that something happened between October 2016 and September 2018, which put the brakes on the PPI growth story and shifted the momentum towards UPI.

UPI enables transfer of funds directly between bank accounts, and does not require parking of funds in non-interest bearing accounts like PPIs. Some may argue that UPI was a superior product that led to reduced interest in PPIs. However, for others, the need to park and transact with only a limited amount in a PPI, ring fences them from larger amounts in a bank account reducing their total risk exposure. Nonbank transaction accounts can also improve access to digital payments for unbanked households (Toh, 2023). The argument that UPI winning over PPIs because the former is a better product would have more credibility if, during the specified period (Ocotber 2016 - September 2018):

• There were no regulatory interventions that adversely affected PPI operations,
• Regulations did not favour UPI,
• Exits of private players from the PPI space were dispersed as more players realised the futility of competing with UPI.

The article argues that this was not the case, and there is reason to believe that the policy and regulatory environment during this period contributed to the slowdown in the PPI growth trajectory, particularly those of non-bank PPIs, and favoured UPI.

The build-up of the PPI market

On 8 November 2016, the Government of India withdrew the legal tender status of Rs. 500 and Rs. 1000 denomination of banknotes issued by the RBI till that date (RBI, 2016). Such large-scale demonetisation gave a fillip for the demand of digital payments. Consequently, the volume of PPI transactions shot up from 127 million in October 2016 to 261 million by December 2016 and steadily rose to around 340 million by March 2017. The value of PPI transactions also surpassed Rs. 100 billion during this period (RBI, PSI). In the months following demonetisation, around 10 entities, all non-banks, received permission from the RBI to issue and operate PPIs (RBI, 2025). The number of non-bank PPI issuers increased from 37 to 55, and surpassed bank PPI issuers for the first (and only) time in this period (RBI, AR).

The draft PPI Master Directions: March 2017

A few months after demonetisation, on 20 March 2017, the RBI issued draft Master Directions on Issuance and Operations of PPIs in India for public comments (RBI, 2017). One of the stated objectives of the draft Master Directions was to encourage innovation in the segment in a prudent manner, taking into account safety and security along with customer protection and convenience.

By this time, three types of PPIs were regulated: semi closed PPIs with minimum KYC, semi closed PPIs with full KYC, and open PPIs. The key difference between semi-closed and open PPIs was that the former could be used to purchase goods and services at specific or clearly identified merchant locations, while the latter could be used for purchase of goods and services at any card accepting merchant location. Cash withdrawal was not permitted through semi-closed PPIs but allowed through open PPIs.

The draft Master Directions proposed two important changes:

1. Increase in the capital and networth requirements: Before the draft, PPI operators were required to maintain a minimum positive net worth of Rs. 1 crore at all times, along with a minimum paid up capital of Rs. 5 crores (RBI, 2016).
2. Limits on monthly fund transfers: Earlier there were no limits on monthly fund transfers. The 2017 draft directions proposed restrictions on the minimum amount outstanding at any point of time, the maximum amount that could be loaded on to a wallet in a month, and the amount that could be transfered out in a month.

Some key proposals of the 2017 draft directions were:

All figures in Indian Rupees

Issue	Semi closed PPIs with minimum KYC	Semi closed PPIs with full KYC	Open system PPIs
Minimum positive net worth to be maintained at all times	25 crores	25 crores	25 crores
Amount outstanding at any point of time	20,000	1,00,000	1,00,000
Maximum amount that can be loaded during any month	20,000	Cash loading limit: 50,000	Cash loading limit: 50,000
Monthly fund transfer limit	10,000	For pre registered beneficiary: 1,00,000 For other cases: 10,000	For pre registered beneficiary: 1,00,000 For other cases: 10,000

There was pushback from stakeholders on the proposals of the draft Master Directions (IGIDR, 2017; CUTS, 2017; IFMR, 2017; NASSCOM-DSCI, 2017). For instance, the Finance Research Group (FRG) at IGIDR called out disproportionate proposals around capital requirements, and transaction limits on consumers, and suggested rolling them back. Specifically, with respect to transaction limits, the FRG pointed out:

The intention for imposing transaction limits and restricting consumers from transacting with their own money is unclear. Every payment system in an economy is susceptible to fraud. However, we do not impose limits on the use of payment systems to pre-empt frauds. For example, the susceptibility of credit card transactions to frauds does not lead us to impose limits on individual credit card transactions. On the contrary, imposing transaction limits on consumers is contrary to their interests.

The final PPI master directions: October 2017

In October 2017, the RBI issued Master Direction on Issuance and Operation of PPIs, after examining comments and feedback received on draft directions (RBI, 2017A). Some key provisions of the Master Directions were:

All figures in Indian Rupees

Issue	Semi-closed PPIs with minimum KYC	Semi closed PPIs with full KYC	Open system PPIs
Minimum positive net worth at the time of application	5 crores	5 crores	5 crores
Minimum positive net worth by the end of third financial year of receiving final authorisation	15 crores	15 crores	15 crores
Amount outstanding at any point of time	10,000	1,00,000	1,00,000
Maximum amount that could be loaded during any month	10,000	Cash loading limit: 50,000	Cash loading limit: 50,000
Cap on amount that could be loaded during a financial year	1,00,000
Monthly fund transfer limit	10,000	For preregistered beneficiary: 1,00,000 For other cases: 10,000	For preregistered beneficiary: 1,00,000 For other cases: 10,000

The Master Directions had reduced networth requirements from Rs 25 crore proposed in the draft to Rs 15 crore. However, additional restrictions were imposed on transaction limits. For instance, the maximum amount which could be loaded during any month in a semi closed PPI with minimum KYC was reduced from Rs. 20,000 to Rs. 10,000. In addition, a new limit on maximum amount that could be loaded in a semi-closed PPI with minimum KYC during a financial year of Rs. 1,00,000 was included in the final directions. It is important to note that this had not found mention in the original draft proposal that was circulated for comments.

This effectively meant that the average amount that could be loaded in a semi-closed PPI with minimum KYC on a monthly basis was Rs. 8,333. Alternatively, for 10 months during a financial year, Rs. 10,000 could be loaded in a semi-closed PPI with minimum KYC every month, but for the next two months, no amount could be loaded, to comply with the annual limits. Coversations with stakeholders suggest that such restrictions potentially frustrated recurring payments and auto-debit mandates which rely on minimum balance being available in a wallet. Further, it restricted the use case of PPIs to very narrow set of transactions.

The impact of regulatory changes: the size of transactions

The proposals in both the draft and final Master Directions had an immediate and severe impact on the PPI market. Figure 1 shows the impact on the PPI transaction volumes, and the turning point when UPI overtook PPIs.

Figure 1: PPI and UPI transaction volumes

Figure 2 shows the impact of Master Directions on the PPI transaction values, and the turning point when UPI overtook PPIs - immediately after these Directions.

Figure 2: PPI and UPI transaction values

However, one must also note that PPI transactions were more than UPI's, for about a year since its launch in May 2016 till about October 2017, when the final master directions on PPIs came into effect. UPI got a fillip due to demonetisation but was still used less than PPIs for several subsequent months. This suggests that PPIs were relevant in a market with UPI.

The impact of regulatory changes: the number and type of players

Figure 3 presents the changes in the number of players in the PPI market. Between March and October 2017, i.e. between the draft and final PPI directions, licenses of seven PPI operators, all of which were non-banks, were cancelled. Of these, four voluntarily surrendered their license, perhaps owing to the restrictive regulatory framework proposed in the draft Master Directions (RBI, 2025). This shows the impact that the draft directions, which also reflected regulator's thinking, had on the the PPI market. Once the directions were finalised, licenses of 20 PPI operators were cancelled by the RBI, all of which were non-banks. Overall, we see that in this period, 16 players voluntarily surrendered their licenses, four ceased operations, five converted themselves to payments banks, and one license was revoked. Further for three years from 2018 to 2020, not a single permission was granted to non-banks for the issuance and operation of PPI (RBI, 2025). Given that licenses were being voluntarily surrendered, one may assume that very few or no fresh applications were received by the RBI. This was also the period during which the government began massively incentivising use of UPI through cashbacks and other schemes (Kulkarni, 2018).

The number of licenses issued went up only in 2021, potentially as a response to the creation of a new category of semi-closed PPI with minimum KYC which could be loaded only through bank accounts.

Figure 3: Non-Bank Licenses issued and cancelled

The regulations led to a shift in the composition of players as well. Earlier, the market saw both bank and non-bank entities offer PPI products. The commercial consequences of the regulatory changes were much larger on non-bank PPIs relative to bank PPIs. Consequently, the number of non-bank PPI operators reduced from 55 in 2016-17 to 36 in 2020-21, while the number of bank PPI operators increased from 54 to 56. In fact, it went upto 62 in 2019-20, possibily indicating the interests that banks retained in offering PPIs, which ideally should have not been the case, if UPI was a superior product. Of the non-bank PPI operators which continue to operate in spite of the regulatory changes in 2017, many are legacy operators enjoying a loyal user base, some are part of larger groups having deep pockets and providing financial or digital services, others have expanded into offerings like digital lending, some offer niche services like foreign exchange, money transfer, and transit payments, while few had to undergo change in management and control.

Figure 4: Bank and non-bank players

Further, non-bank issuers have faced more stringent compliance burdens, particularly around KYC norms, fund loading restrictions, and interoperability requirements. For instance, while non-bank PPIs must maintain an escrow account with a partner bank, bank operated PPIs can leverage their own deposit accounts, reducing costs and operational friction. Additionally, regulatory decisions such as restricting credit lines on PPI wallets have disproportionately impacted non-bank issuers, limiting their ability to innovate and compete with banks that can seamlessly integrate PPI-like functionalities within their broader suite of financial services.

RBI's review of its stringent conditions (late 2019-early 2020)

More than two years after the October 2017 Master Directions, the RBI decided to review some of the stringent conditions imposed on the PPI market, particularly those related to loading of PPIs. In December 2019, it decided to create a new category of semi-closed PPI with minimum KYC which could be loaded only through bank accounts. For such PPIs, the amount which could have been loaded through bank accounts during a financial year was increased to Rs. 1,20,000 i.e. Rs. 10,000 per month could be loaded in such PPIs. The RBI recognised that such leeway was necessary to ensure regular bill and merchant payments (RBI, 2019A). For other semi-closed PPIs with minimum KYC, the annual limit of Rs. 1,00,000 for loading was retained. However, this move appeared to be too little too late and perhaps failed to uplift the momentum in PPIs. Consequently, in January 2020, in addition to bank accounts, credit cards were permitted as a mechanism of loading semi-closed PPIs with minimum KYC having an annual loading limit of Rs. 1,20,000. This move, so far, has also failed to push PPI towards previously experienced growth rates.

Conclusion

The analysis of RBI regulation of PPIs over the past decade, changes in PPI volumes, value, and operators during this period, suggests that regulation has indeed impacted market outcomes for PPIs. While the RBI provided some relaxations, these were not enough to push PPIs back into high growth trajectory.

While it is difficult to pinpoint the exact regulatory requirement which might have impacted the PPI market most, it is clear that the restrictions introduced in October 2017 collectively contributed to a significant decline amongst industry's interest in PPIs as payment instruments. This was validated through interactions with key stakeholders as well.

One reckons that the RBI would have introduced such restrictions in the interests of safety and security of the payments market, prevent fraudulent actors from misusing the instrument, and obstruct unreliable entities from issuing and operating PPI instruments. It might not have predicted that the restrictions could have had such adverse consequences of reducing the attractiveness of PPIs as an instrument, and the users and market players, moving away from the PPI market. It is here where robust public consultations, and ex-ante cost benefits analyses can prove useful to estimate the potential impact of regulatory instruments, avoid disproportionate regulation, and ensure balanced market outcomes. While the RBI did invite suggestions on draft directions in March 2017, several new requirements, such as the annual cap on loading semi-closed PPIs with minimum KYC, were directly incorporated in the final directions of October 2017. This prevented stakeholders from providing their inputs on such new requirements.

The RBI should use tools like public consultations more often and thorougly. In any case, frauds and transaction failures are equally possible with UPI, as recent events have shown. In fact, with UPI, the entire bank balance of the customer is at risk, making them more susceptible than PPIs could ever have been.

Some may argue that with the advent of UPI, PPI was bound to lose market share. However, as this article shows, PPI was restricted by regulations, and did not get an opportunity to effectively compete with UPI, which on the other hand was booming on the back of regulatory relaxations, incentives, and zero fee mandates. One must also not forget that it was the non-banks which propelled UPI to unimaginable heights, and non-banks were also behind PPIs initial growth before the regulations hit them badly.

References

CUTS, 2017: Comments on draft Master Directions on issuance and operations of PPIs in India, 2017, https://cuts-ccier.org/pdf/Advocacy-Comments_on_RBI_Master_Direction_on_Issuance_and_Operation_of_PPIs.pdf

IFMR, 2017: Comments on draft Master Directions on issuance and operations of PPIs in India, 2017, https://dvararesearch.com/wp-content/uploads/2024/01/IFMR-Finance-Foundation-Comments-on-the-RBI-Draft-Master-Directions-on-Issuance-and-Operation-of-Prepaid-Payment-Instruments-in-India.pdf

IGIDR, 2017: Finance Research Group, Inputs on draft Master Directions on issuance and operations of PPIs in India, IGIDR, 16 April 2017, https://ifrogs.org/PDF/201703note_inputsToCpOnPpis.pdf

Kulkarni, 2018: Amol Kulkarni, (Not) the way to promote digital payments, CUTS Discussion Paper, February 2018, https://cuts-ccier.org/pdf/DP_the_way_to_promote_digital_payments.pdf

Mukherjee, 2025: Mukherjee, India: UPI enabled for prepaid payment via third-party apps, Coingeek, 7 January 2025, at https://coingeek.com/india-upi-enabled-for-prepaid-payment-via-third-party-apps/

NASSCOM-DSCI, 2017: Inputs on draft Master Directions on issuance and operations of PPIs in India, 2017, https://www.dsci.in/resource/content/nasscom-dsci-submission-rbi-master-directions-ppis

RBI, 2008: RBI Press Release regarding Inviting Comments on Approach Paper on Guidelines for PPIs dated 7 November 2008, at https://rbi.org.in/scripts/BS_PressReleaseDisplay.aspx?prid=19419

RBI, 2016: RBI Master Circular dated 1 July 2016 regarding Policy Guidelines on Issuance and Operation of PPIs in India, at https://rbi.org.in/scripts/NotificationUser.aspx?Mode=0&Id=10510

RBI, 2016A: RBI Press Release dated 8 November 2016 regarding Withdrawal of Legal Tender Status of Rs 500 and Rs 1000 Notes, at https://rbi.org.in/Scripts/BS_PressReleaseDisplay.aspx?prid=38520

RBI, 2017: RBI Draft Master Directions dated 20 March 2017 regarding Issuance and Operations of PPIs in India, at https://www.rbi.org.in/Scripts/bs_viewcontent.aspx?Id=3325

RBI, 2017A: RBI Master Direction dated 11 October 2017 (updated as of 17 November 2020) regarding Issuance and Operation of PPIs at https://rbi.org.in/scripts/NotificationUser.aspx?Mode=0&Id=11142

RBI, 2019: RBI Notification regarding Introduction of a New Type of semi-closed PPIs dated 24 December 2019, at https://www.rbi.org.in/Scripts/NotificationUser.aspx?Id=11766&Mode=0

RBI, 2019A: RBI Press Release regarding Statement on Development and Regulatory Policies dated 6 December 2019, at https://www.rbi.org.in/Scripts/BS_PressReleaseDisplay.aspx?prid=48803

RBI, 2021: RBI Master Directions on PPIs dated 27 August 2021 (updated as on 27 December 2024), at https://www.rbi.org.in/Scripts/BS_ViewMasDirections.aspx?id=12156

RBI, 2024: RBI Notification dated 27 December 2024 regarding UPI access for PPIs through third-party applications, at https://www.rbi.org.in/Scripts/NotificationUser.aspx?Id=12756&Mode=0

RBI, 2024A: RBI Press Release dated 5 April 2024 regarding Statement on Developmental and Regulatory Policies, at https://www.rbi.org.in/Scripts/BS_PressReleaseDisplay.aspx?prid=57639

RBI, 2025: RBI, Approvals/ Certificates of Authorisation issued by the Reserve Bank of India under the Payment and Settlement Systems Act, 2007 for Setting up and Operating Payment System in India, 9 January 2025, at https://rbi.org.in/Scripts/PublicationsView.aspx?id=12043

RBI, AR: RBI Annual Reports at https://rbi.org.in/Scripts/AnnualReportMainDisplay.aspx. There was a change in financial year from June to March from 2019-20. Also, data of bank PPI issuers for 2015-16 is not available.

RBI, PSI: RBI Payment System Indicators (monthly) at https://www.rbi.org.in/Scripts/PSIUserView.aspx

Toh, 2023: Ying Lei Toh, How Much Do Nonbank Transaction Accounts Improve Access to Digital Payments for Unbanked Households?, Payment System Research Briefing, 29 November 2023, Federal Reserve Bank of Kansas City, at https://www.kansascityfed.org/Root/documents/9919/PaymentsSystemResearchBriefing23Toh1129.pdf

The authors are researchers at the TrustBridge Rule of Law Foundation.

Digital transformation and the paradox of financial inclusion in India

by Suyash Rai.

India has made great strides in digital technology, becoming a leading exporter of digitally delivered services to the global economy. These capabilities with computer technology fuelled hopes that digital transformation could yield gains for the Indian state that are comparable to those seen in the private sector. The `Digital Public Infrastructure (DPI)' approach, with India's Aadhaar digital ID system as a prime example, is presented as a path to higher GDP growth for developing countries. There is an emerging debate on the role of the state in shaping the development and deployment of DPIs.

Two key pillars of the Indian story with DPIs are identity services ("Aadhaar") and their impact on financial inclusion. In a new working paper, Economic development and digital transformation: Learning from the experience of Aadhaar and financial inclusion in India, I critically examine the Indian progress on financial inclusion between 2011 and 2021, revealing a paradox: while account ownership surged, account usage remained low.

The facts

The paper analyses India's performance compared to other lower middle-income and middle-income countries. The evidence shows:

• Impressive account opening: India witnessed remarkable progress in account penetration, surpassing the average improvement in middle-income countries.
• High inactivity: A significant percentage of accounts in India were inactive, far exceeding the average for middle-income countries.
• Low account usage: India lagged behind in account usage for both consumption smoothing (regular deposits and withdrawals) and digital payments, indicating a gap between account ownership and actual financial inclusion.

The role of government mandates and Aadhaar

We argue that the rapid scale of account opening was caused by a series of government and Reserve Bank of India (RBI)mandates, particularly the Pradhan Mantri Jan Dhan Yojana (PMJDY). While Aadhaar played a role, it was primarily used as a physical ID for KYC, rather than as a digital ID through e-KYC. The gains in account opening may have a lot to do with state coercion and less to do with DPI.

The primary objective driving these initiatives was to facilitate direct benefit transfers (DBT) for welfare schemes. The government's focus on DBT aimed to reduce leakages and improve attribution for its welfare programs in the eyes of voters.

Why did this approach yield disappointing results?

The paper explores several reasons for the limited account usage despite the increase in account ownership:

• The lack of a viable business model: No-frills accounts, with zero minimum balance and free transactions, are commercially unattractive for banks.
• Mismatch between the solution and the problem: The focus on account opening for DBT didn't necessarily translate into accounts that address the richness and complexity of finance for the poor, of meeting the diverse needs of users for consumption smoothing and payments.

Lessons

The top-down approach, with a readiness to utilise the coercive power of the state, has limitations. While the government achieved its objective of scaling up DBT, this came at the cost of genuine financial inclusion and limited the potential uses of Aadhaar as a DPI.

We highlight the need for a more balanced approach, considering market forces and user needs, so as to obtain better outcomes with DPIs. We stress the importance of political creativity, institutional reforms, and a broader understanding of public value, beyond narrow fiscal objectives, when designing and implementing DPIs.

We offers insights into the complexities of digital transformation and financial inclusion, challenging the simplistic narrative of Aadhaar's success. These experiences invite us to rethink the role of the state in shaping DPIs and consider alternative approaches that can truly leverage technology for inclusive and sustainable development.

Suyash Rai is a Fellow at Carnegie India and a Visiting Research Fellow at the xKDR Forum

.

Concerns about recent developments on SEBI's regulation-making on market rumours and insider trading

by Bhavin Patel and Renuka Sane.

Insider trading is one of the areas of financial regulation where the order of complexity required of state capability is relatively high, and the gains to society from successful implementation are relatively low. The Indian environment on SEBI enforcement against insider trading has accumulated many difficulties. In recent months, we have seen the next step forward in SEBI's journey, with a novel legal idea around rumours swirling in the market. In this article, we summarise the recent developments, and show two substantive problems with the direction taken by SEBI. These substantive problems are ultimately grounded in failures of process. There are signs that the process adopted in recent months for these developments on rumours, has more deficiencies than usual.

Recent developments on how SEBI thinks about rumours

Over the past year, SEBI has been concerned about the impact of market rumours on security prices. It proposed that certain listed entities be required to verify "market rumours" related to their firm. Amendments to this effect (the June 2023 Amendments) were made in the SEBI (Listing Obligations and Disclosure Requirements) Regulations, 2015 (the LODR Regulations), though the amendments are yet to be enforced. On December 28, 2023, it published a Consultation Paper on Amendments to SEBI Regulations with Respect to Verification of Market Rumour (the Current Consultation Paper) suggesting that listed entities should verify market rumours only if they are material. 'Materiality' meant that such rumours should lead to price movement in the security. It also proposed that if a listed entity classifies certain information as Unpublished Price Sensitive Information (UPSI) but does not verify a market rumour related to such information, it should continue to be treated as UPSI under the SEBI (Prohibition of Insider Trading) Regulations, 2015, (the PIT Regulations). In its Board Meeting held on March 15, 2024, SEBI approved this proposal, confirming that if a rumour related to UPSI is not verified, it will continue to be treated as UPSI. In other words, if a speculative story appears in the public domain, pertaining to UPSI about a firm, it will still be treated as UPSI until the listed entity verifies the story.

These steps contain difficulties that are generic to the working of SEBI. Firms are already mandated to release certain information (through listing obligations), and not release some (through prohibition on insider trading). SEBI's proposals on verification of market rumours have not yet explained the market failure proposed to be solved and how the selected intervention is the least costly way of doing so. In this particular case, the legal effects of the new law can be particularly damaging. The decision in the March 15 Board meeting, to continue to treat Generally Available Information (verified or not) as Unpublished Price Sensitive Information (till the company verifies it), goes against the grain of how price discovery works in the public market. A greater process discipline will help improve the thinking, and the democratic legitimacy, of insider trading regulation.

Substantive problem #1: The impact on the OTC market

Consider the following illustration:

• Assume a listed entity classifies "defaulting on a large supplier contract" as UPSI.

• Since the information is classified as UPSI, insiders would be prohibited from trading based on such information under the PIT Regulations. Note that the definition of insider under R. 2(1)(g) of the PIT Regulations includes persons with access to UPSI. This implies that if a retail investor, with no connection to the firm, has access to this information, the retail investor is also classified as an insider. This speaks to the over-inclusiveness of the term 'insider' under prevailing securities laws.

• Assume that a newspaper carries a story about a "rumour about the default". Since this is now reported in the media, it will be accessible to the general public. The listed entity chooses not to verify the information since it does not trigger the materiality threshold under R. 30 of the LODR Regulations (for example, it might not cause a price movement or may only cause a price movement lower than the limits specified in the threshold).

• Under the current proposal, an unverified event or information reported in the media would not be considered 'generally available information' under the PIT Regulations. It would continue to be UPSI. The definition of insider regarding UPSI would extend to the general public. As a result, the PIT Regulations would generalise the prohibition on trading based on such information, even though it has been reported in the media, because the listed entity chose not to verify it. This may effectively hinder the price discovery process, a core function of securities markets.

Further, the explanation to R. 4(1) of the PIT Regulations states that if a person possesses UPSI, their trades would be presumed to have been motivated "by the knowledge and awareness of such information". The proviso to this explanation states that an insider can prove their 'innocence' if: (i) the transaction is an off-market inter-se transfer between insiders who were in possession of the same unpublished price sensitive information without being in breach of R. 3 and both parties had made a conscious and informed trade decision, (ii)the transaction was carried out through the block deal window mechanism between persons who were in possession of the unpublished price sensitive information without being in breach of R. 3 and both parties had made a conscious and informed trade decision.

As a result, under proviso (i) two people who read the same unverified story in the media are consequently insiders and who have made a 'conscious and informed trade decision' can trade in the listed entity's securities if such trade is off-market; and under proviso (ii) a block deal trade is also allowed under similar circumstances.

The effect would be that regular trades on the floor of the exchange based on unverified media reports would be prohibited. However, off-market transactions or block deals conducted based on such information would not count as 'insider trading', creating a problem of arbitrary discrimination. This also drastically reduces the pool of investors able to trade, which may possibly lead to a liquidity collapse.

Substantive problem #2: Contradictions with existing regulations

The proposals contradict a literal interpretation of UPSI as set out in the PIT Regulations since they suggest that information which is neither unpublished (since it has been published in the media) nor price-sensitive (since it does not affect the price of securities and therefore falls outside the scope of the materiality threshold verification requirements under R. 30 of the LODR Regulations) would still be considered UPSI. This would suggest a need to amend the definition of UPSI under the PIT Regulations and, potentially, the provisos to the explanation to R. 4(1) and the definition of insider under R. 2(1)(g). The proposals in the Current Consultation Paper also necessitate changes to determining materiality under R. 30(4) of the LODR Regulations. This is because the proposed definition of materiality is not a part of the June 2023 Amendments. The materiality thresholds proposed impact which market rumours listed entities need to verify.

Difficulties with the process

Assuming that SEBI has considered the potential impact on liquidity and does not regard it as problematic, we are still faced with the problem of how these proposals may be implemented. The current proposal may require changes in other substantive regulations. If they are indeed to be implemented, they should be done through amendments to the PIT and LODR Regulations. Board minutes, Circulars, or Guidelines should not suffice to effectuate substantive changes. Under Section 30 of the Securities and Exchange Board of India Act, 1992 (the SEBI Act), changes to regulations must be laid "as soon as may be after it is made, before each House of Parliament, while it is in session, for a total period of thirty days." This allows for Parliamentary oversight of the regulator's exercise of powers of subordinate legislation and prevents the use of such powers beyond permitted limits. The June 2023 Amendments are scheduled to take effect from June 1, 2024, there is little time left for listed entities to prepare, and such amendments should be published soon. The process under S. 30 of the SEBI Act was followed in the case of the June 2023 Amendments, and the Amendments were laid before Parliament on August 11, 2023. There is no reason this process should be side-stepped now. SEBI should come out with a precise legal instrument, and the amendments to the PIT and LODR Regulations, to implement its proposal.

Ideally, the regulator should, in addition to inviting and analysing public comments, identify the problem or market failure these seek to address, the principles governing the proposals, the outcome the regulator aims to achieve through such changes, and an analysis of their costs and benefits as recommended by the Financial Sector Legislative Reforms Commission (FSLRC) in its Handbook on adoption of governance enhancing and non-legislative elements of the draft Indian Financial Code of December 26, 2013. The Financial Stability and Development Council had also approved the implementation of the FSLRC's recommendations in its meeting in October 2013. If this process had been followed, the failures in substantive thinking described earlier would have been avoided.

The authors are researchers at TrustBridge. We thank Amol Kulkarni and Madhav Goel for useful comments.

Does financial and macro policy explain household investment in gold?

by Renuka Sane and Manish Kumar Singh.

Gold plays a significant role in the portfolio of Indian households. Several explanations have been offered: gold is an important source of credit, it matters for socio-cultural and political reasons, provides women with agency as women are likely to have more control over the gold they own relative to financial assets. Research has, however, not paid adequate attention to the performance of gold as an asset class. Investment in gold is often brushed aside as irrational based on the evidence that gold has delivered near zero real returns (in USD) over a 100 year period (Siegel, 2014). In a new working paper, Sane and Singh (2022): "Does financial and macro policy explain household investment in gold?" We argue that investments in gold have to be seen in the context of Indian financial markets. Gold is a a far more sensible investment that international research would suggest.

Indian financial markets

Household saving is a function of the financial environment within which the household operates. The following characteristics of Indian markets are worth noting:

1. High inflation: India adopted a formal inflation target of 4 per cent within a band of +/- 2 per cent in August 2016. Before this, high levels and volatility of inflation had been a persistent problem in India. The average inflation in the four years prior to inflation targeting was around 7.26% - this dropped to 4.19% after the adoption of the framework (Patnaik & Pandey, 2020). When there is such high and persistent inflation, households will naturally look for instruments which are able to, at the very least, beat inflation, even if not provide a complete hedge.

2. Interest rate management: India has consistently followed a policy of managing long-term interest rates on government borrowings. This has led to an environment of low interest rates for government borrowing, and has prevented long-term yields from rising. Interest rates on fixed deposits which are benchmarked to long-term government yield is also relatively low and have been consistently falling over the 1999-2021 period from an interest of 9-10% to about 4%.

3. Volatility in equity markets: Emerging markets are generally more volatile than markets in OECD countries. The annualised 10 year standard deviation on the MSCI Emerging Markets Index was around 17%, while that of the MSCI World Index (based on large and mid cap representation across 23 Developed Markets (DM)) was around 13%. An asset that serves as a hedge assumes greater importance in emerging markets relative to developed markets.

4. Capital controls: One way to hedge a portfolio is to diversify across different markets. However, this has been difficult in India, owing to a complex framework of restrictions on the current and capital account till the year 2000. In 2000, the current account was made fully convertible, and a modified framework for capital controls was put in place (Patnaik & Shah, 2012). There continue to be restrictions, on both the current and capital account, which differ depending on the type of investor, and the assets in question.

5. Currency interventions: Patnaik & Sengupta (2021) study RBI interventions and find that when there has been pressure on the rupee to appreciate, the RBI has responded by intervening in the forex market and buying dollars. When, in the aftermath of the 2008 global financial crisis, there was pressure on the rupee to depreciate, the RBI allowed the rupee to fluctuate in this period. Indian investors, therefore, benefit from a larger depreciation of the rupee for those assets where the price is determined in international markets.

High inflation levels and volatility, low interest rates on account of financial repression, inability to invest in international markets until very recently, depreciation of the rupee have a bearing on the choices that are available to households. Financial repression changes the risk-return trade-off between fixed deposits and gold. Similarly in an environment where individuals are restricted from investing in overseas markets, gold offers a way for doing international diversification. These become important considerations as we evaluate the performance of gold vis-a-vis the Indian equity market.

How does gold fare?

We use data from June 1999-March 2021 and find that:

1. In the last 20 years, real returns on gold have always been positive.

2. Apart from a few years around 2018, gold has consistently beaten returns on fixed deposits.

3. RBI interventions in the currency market changes the dynamics of gold return for Indian households. Our regression estimates suggest that if the Indian rupee depreciated against the U.S. dollar by 10% in a month, then the gold price in Indian rupees increased on average about 3.63%. While the exchange rate pass-through is far from complete, it implies that currency interventions by the Reserve Bank of India have implications for the gold price that is seen by Indian investors.

4. Gold and NIFTY seem to have moved together till about 2008, after which NIFTY saw a sharp fall, while gold continued with its upward trajectory. The two asset classes moved together again till about 2014, and then from 2015 till early 2020. There seems to be a divergence in the series around 2014, when NIFTY was rising steadily while gold prices fell before rising again. Gold is a strong hedge against the NIFTY when measured in daily frequency. In the last 10 years, this relationship had become stronger.

5. The global minimum variance portfolio which only includes gold and NIFTY suggests a 63% weight to gold for target annual return of about 13%. As the target return increases, we see that the weight allocated to gold drops to about 3%. When one does a similar optimisation exercise including the S&P 500 returns, the global minimum variance portfolio suggests a weight of 46.5% for gold, 31.3% for NIFTY and 22.2% for SPX. Once international diversification is possible, the weight of gold has fallen by almost 16 percentage points. The confidence intervals, however, on these estimates are wide given the paucity of longer time-series data on returns.

Conclusion

Gold has provided the means to Indian households to overcome the difficulties associated with high inflation in a financially repressed macroeconomic environment with capital controls. Given the performance of gold, fixed deposits and NIFTY, and the difficulties of international diversification households have not been entirely unreasonable to hold gold in their portfolios. If policy has to channel household savings to more productive uses, it has to confront the underlying issues in the macroeconomic environment which make gold a preferred investment choice.

References

Patnaik, I. & Pandey, R. (2020). Four years of the inflation targeting framework. NIPFP Working Paper Series, No 325.

Patnaik, I. & Sengupta, R. (2021). Analysing India's exchange rate regime. India Policy Forum (forthcoming).

Patnaik, I. & Shah, A. (2012). Did the Indian capital controls work as a tool of macroeconomic policy? IMF Economic Review, 60, 439-464.

Sane, R. & Singh, M. (2022). Does financial and macro policy explain household investment in gold?, Dvara Research Working Paper Series No. WP-2022-01.

Siegel, J. J. (2014). Stocks for the long run: The definitive guide to financial market returns and long-term investment strategies. McGraw Hill.

Renuka Sane is a researcher at NIPFP, New Delhi. Manish Kumar Singh is a researcher at IIT Roorkee.

A cooperative liquidity window for mutual funds: A debate

by Harsh Vardhan vs. Josh Felman and Ajay Shah.

Problem statement

There is a mismatch between the growth of the mutual fund industry versus the maturation of the financial markets (Shah, 2018). This generated trouble after the IL&FS default of August 2018, and will likely make trouble in the future also. Mutual funds are in an awkward place, promising liquidity to their customers but lacking a liquid bond market. Some years ago, the exchanges were getting better, and there was a path to building the Bond-Currency-Derivatives Nexus, so we could hope that progress on both paths would come along and solve the problem of the mutual funds. Now, both elements (exchanges and bond market reform) have a weak outlook. Is there a way out of this conundrum? Can a liquidity window for mutual funds be created, through which the problem of the mutual funds can be solved?

Why we need this and how it can work, by Harsh Vardhan

Indian debt mutual funds have grown rapidly over the past few years. Debt funds got a strong push after demonetisation. Currently the total assets under management (AuM) of debt funds are ~ Rs 15 Trn. There are individual debt fund schemes with AuMs of over Rs 1 Trn.

Debt funds invest their corpus in debt securities. In India there are two main classes of debt securities – those issued by the government including central and state governments and those issued by companies. Both have very poor liquidity. In the case of government bonds, while there is a somewhat liquid interbank market, a large part of the liquidity is in a single ‘benchmark’ paper which is typically a 10 year bond. When a new 10 year bond is issued, the old one ceases to be the benchmark and its liquidity drops sharply. The lack of liquidity is even worse with corporate bonds.

Most debt mutual funds promise high liquidity to their investors. For liquid and short duration funds, redemption proceeds are credited to the investor on T+1 while for most other debt funds it is T+2. MFs suffer the agony of liquid liabilities and illiquid assets. They manage this challenge through two pathways: (a) holding cash (typically less than 5% of AuM) and (b) having credit lines from banks.

There is considerable systemic risk in the Indian financial system, and situations where these two pathways prove to be inadequate. As an example, Franklin Templeton shut down six debt schemes when redemptions were unusually large and the bond market was unusually illiquid. The redemption pressure that they faced had nothing to do with their money management; it was induced by an episode of systemic risk.

In the anatomy of these recurrent debt market crises, one interesting feature is market failure in the form of a negative externality. Purely at random, when large redemptions show up at any one door, the selling that this induces drives down prices (as the overall market is illiquid and impact cost is high), which adversely impacts the NAV of all other funds. For any rational economic agent that sees the first inkling of higher outflows (either by watching flows or by looking at NAV changes), it is rational to yank all debt investments. This creates a channel through which selling by one fund induces redemptions for others.

Another way to locate these problems in the framework of market failure is to see that market liquidity is a public good. As an example, the liquidity of Nifty futures is non-rival (your consumption of liquidity does not adversely impinge on my access to liquidity) and non-excludable (everyone can access the Nifty futures market). When we build liquid markets, we are creating a public good.

All market failure is ultimately a problem of coordination between economic agents. We should look for collective action through which some of the problems of debt mutual funds can be addressed.

There are two solutions going around, for this problem of bond market illiquidity, which just don’t make sense. One strategy is for regulators to demand that mutual funds hold more capital. Mutual funds are not balance-sheet based entities and the journey of trying to amplify their equity capital requirements is conceptually wrong. Another strategy is for the central bank or the government through any other agency, to run a liquidity window for mutual funds. When the full consequences of this play out for mutual funds, it is likely to leave them worse off.

Is there a way out of this jam? I believe we can establish a Cooperative Liquidity Window (CLW), built by mutual funds for mutual funds -- with a small involvement of the state -- which can help solve this problem. For the people who are too used to state leadership in such things, we should point out that the Bank of England played this kind of function -- liquidity support for distressed banks -- for centuries as a purely private organisation; it was only nationalised in 1946. During the great depression in the US in the 1930s, J P Morgan, founder owner of the eponymous bank, orchestrated a bail-out of the American banking system through co-operative efforts of larger, stronger banks. These experiences are food for thought, and the design proposed here draws on this history.

For such an emergency liquidity support mechanism, we should establish five conceptual objectives:

• It should use no public money.
• There should be an extremely low amount of state coercion involved, in getting some MFs to participate in the CLW, and no role for the state in terms of regulation, management, appointments, or rule-making of the CLW.
• The governance of the mechanism should be within the AMCs that participate in it; it should operate as a self regulatory organisation.
• The capital to set up and operate the mechanism should be provided by the participants; it should operate as a mutual co-operative; rules of access to the mechanism should be defined by the participants.
• It should be only an emergency liquidity support system. The criteria for defining an emergency, and the extent of support that can be provided to individual entities, should be defined by members as the by-laws of the mechanism.

How would the proposed CLW work?

1. The participating AMCs would create a vehicle by contributing to the equity of the vehicle. The vehicle could be set up as a trust or any other legal form that minimizes transaction costs.
2. Some members would be coerced by SEBI (the largest firms adding up to perhaps 75% of the category AUM) and others would be voluntary participants (those who would like to benefit from its services even if not forced by SEBI). Apart from this, there would be no role for the state power in the CLW, in any fashion.
3. The equity contribution of each MF should be determined by its debt fund corpus. For example, all MFs with debt fund AuM of over Rs 1 Trn might contribute Rs.5 Billion, those with an AuM of Rs 0.5 Trn to 1 Trn might contribute Rs. 3 Billion, and so on. The CLW governance must write the specific rules of equity contributions.
4. The CLW would leverage up and create a corpus that supports a securities repurchase (repo) operation in the event of stress.
5. When a member AMC faces severe redemption pressure (way beyond what is deemed normal by the members collectively as defined by the governance rule of the CLW) it would pledge its eligible debt securities to raise short term liquidity. This would be akin to a bank accessing the repo window in the event of a run.
6. This window would also accept liquidity from members like a normal repo window.
7. The rules regarding the extent of liquidity support provided, the tenure, the bid-ask spread, acceptable securities as collateral and hair cuts, etc. would all be defined by the members collectively.
8. The CLW would operate as a not for profit entity or provide a modest return on equity to the member shareholders.

Currently there are ~45 AMCs in India. If we assume that 40 of them participate, each contributing an average of Rs 1 billion of equity capital, we would have Rs.40 billion of equity capital in hand. Assuming 4x leverage, the resources of the organisation would be Rs.160 billion. It is easy to go to much higher values.

The CLW should support participating MFs only in dealing with liquidity issues and not credit risk issues. This should be enshrined in the governance and operating rules of the CLW. Considering that the CLW will be managed by the AMCs themselves, who are all deeply informed players, it is reasonable to assume that they will be able to differentiate between liquidity and credit issues, Further, at a security level, the CLW will determine eligibility of securities and haircuts applicable. This will ensure that even in providing liquidity support, credit issues are not ignored. The rules of operation of the CLW should be well known, ex ante, so all the participating MFs face a predictable environment.

Let us simulate how the Franklin Templeton crisis might have played out, if this CLW was in place. The issues faced by Franklin Templeton’s shuttered debt funds schemes were purely liquidity issues: Over the last 18 months or so, they have returned upwards of 90% of the AUM at the time of shutting the schemes. Further, the return on these funds during the time was comparable with other funds in the same asset class. As the Franklin Templeton crisis was a liquidity crisis and not a credit crisis, the CLW would have been in play to support the liquidity crisis at Franklin Templeton. With illustrative assets of Rs.160 billion, it would have had the financial depth to deal with this situation, where all six affected funds put together had a total AUM of about Rs. 250 billion.

This design is not a substitute for a deep and developed bond market. A liquid market for securities is always the best solution to deal with any liquidity issues. But we face a problem today: We have a situation where the debt mutual funds corpus has grown very significantly and yet the bond market, especially the corporate bond market, remains very illiquid. The CLW is a mechanism where enlightened self interest can create a cooperative which helps the sector deal with a dangerous liquidity challenge.

In my proposal, there is only one use of state power: I feel SEBI should force large debt funds adding up to (say) 75% of the industry AUM to be members of the CLW, and force non-members to communicate this lack of membership in their customer-facing communications. The justification for this use of state power lies in the extent to which this would help reduce systemic risk (innocent bystanders being adversely affected in the next mutual fund crisis). This coercion addresses the free rider problem, where any one MF may derive benefits from the more stable mutual fund / bond market system, but try to be stingy in not paying for this stabilisation. Apart from this, I propose there should be no state involvement / control / regulation of the analysis, design, staffing, rule-making or operation of the CLW.

All members would have the self interest of making the facility work well -- as they are both owners and customers -- and they would thus exert governance. This is a problem where a cooperative solution works well. There is no market failure in the working of the CLW, and thus no role for regulation or any other involvement of the state.

There is one limitation in this design. The CLW will not be adequate if there is a full fledged financial crisis, such as what was experienced in 2008. In that case, the CLW would become one more element of the financial system that would have to be analysed in the crisis management at MOF.

There is no solution which can cover up for the lack of a bond market, by Josh Felman and Ajay Shah

Bond mutual funds are facing a serious dilemma. On the one hand, they promise investors liquidity, the ability to withdraw money at short notice. But on the other hand, they hold assets that are largely illiquid and difficult to sell. As a result, they face a mismatch between what they promise and what they can actually deliver.

Investors typically pay little attention to this mismatch, because most of the time it isn’t apparent. That’s because on most normal days, the investors who want to withdraw their money are more than counterbalanced by the many investors who are putting their money into the funds. It is only when this balance is disrupted, when a large proportion of investors “run” to take their money out, that mutual funds must sell their assets and the liquidity mismatch is revealed (Sane, Shah, Zaveri 2018).

Of course, banks face a similar mismatch problem. They, too, promise that depositors can withdraw funds easily, even as they hold assets (loans) that are even more illiquid than bonds. But in the case of banks there is a firewall against runs, namely the deposit insurance provided by the Deposit Insurance and Credit Guarantee Corporation. With this insurance, depositors know that their deposits are always safe. Accordingly, they have no incentive to rush to banks to withdraw their money, even if they find out that their bank’s loans have turned bad.

Could a Cooperative Liquidity Window (CLW) provide a similar firewall for debt mutual funds? At first blush, it seems like it would. After all, if the problem is that bonds are illiquid, then it seems logical to create a window that would allow funds to exchange bonds for cash. Moreover, the CLW proposal has some particularly attractive features. It would be a private initiative, involving no public money; and it would be employed only in emergencies, reducing the risk that it would distort financial markets. It avoids state failure by having no state involvement, apart from coercing large mutual funds (MFs) to become members.

But we see difficulties in translating this concept into a working liquidity facility. Consider the following problems with the proposal:

• The illustrative corpus – Rs 160 billion – is relatively small, about the size of a single mutual fund group (such as Franklin Templeton). So, if several groups get into trouble at once, there won’t be enough liquidity to go around. In our thinking about the CLW proposal, we should think of something more like Rs.0.5 trillion of dry powder.
• The proposal envisages that lenders will be willing to purchase Rs 120 billion of CLW debt. Would they really be willing to lend so much money to an unknown institution engaged in the risky activity of buying illiquid debt? And even if they did, what interest rate would they charge?
• Assuming that lenders charge a relatively high rate of interest, how will the economics of every day operation of the CLW work out? In most years, its assets will simply be sitting in safe but low-yielding government securities, so it will suffer from a negative cost of carry. That means it will need to make compensating large profits on its occasional liquidity activities, by buying debt at very low prices and selling at high prices.

Let’s assume optimistically that these problems can somehow be overcome. We think the proposal still won’t work, because it has an important flaw: it is based on the premise that mutual funds facing runs are merely suffering from liquidity problems. But things are usually not this simple. Most runs involve credit risk issues, which means that there is a danger of defaults, which could saddle the CLW with large losses. And this makes all the difference. To be concrete: we don’t agree with Harsh’s relatively sanguine assessment of the Franklin Templeton story.

Runs on mutual funds follow a standard sequence. Initially, investors find out that a large bond-issuing firm is in serious trouble. In response, they start examining the portfolios of their mutual funds. And when they find the funds that are heavily exposed to the teetering firm, they run. This is precisely what happened in the case of Franklin Templeton. This firm invested aggressively in risky assets: even its “safe” Ultra Short mutual fund invested more than one quarter of its portfolio in assets rated A or below, rather than the AAA assets that such funds would normally hold. In addition, Templeton invested heavily in zero coupon bonds issued by Yes Bank. So when financial markets turned risk averse and Yes Bank ran into trouble, investors fled the Templeton funds.

In restrospect, it turns out these investors were correct: there was indeed credit risk. It is now almost two years since Templeton shut six of its funds, and the 300,000 investors in these funds still haven’t received all of their money back. Even if investors are reimbursed eventually for their full nominal amounts, they have suffered an opportunity cost. Inflation will have eaten away at the real value of their money, and they will have lost the opportunity to use the funds to meet last year’s expenses (such as Covid hospital bills) or make other investments. In particular, they were unable to place this money in the stock market, which has nearly doubled since withdrawals were frozen in April 2020.

The complexity of correlations and asymmetric information about credit and liquidity risk means that the proposed CLW will run into three problems:

1. It could distort the incentives of mutual funds. Right now, mutual funds face market discipline. They know that if they invest in risky, illiquid bonds, they will get into trouble if investors panic and demand their money back. So most mutual funds – unlike Franklin Templeton – try to confine their purchases to safe, relatively liquid bonds. Precisely for this reason, most funds were able to survive the runs on Templeton largely unscathed.

This discipline could disappear if a liquidity window is established. In this case, mutual funds will feel more free to buy risky, illiquid bonds. In fact, they might try to buy as many such bonds as possible. After all, risky bonds carry higher interest rates, so mutual funds that buy them will be able to advertise higher returns. And if things go wrong, these funds will always be able to pass the problem onto the CLW.

Of course, they will not be able to transfer all their risk, since they have contributed to the equity capital of the CLW. For example, if they own 10 percent of the CLW, they would have to bear 10 percent of any losses faced by the CLW. Still, they might be able to pass on 90 percent of any potential losses. And this is enough to distort incentives.

So, the CLW will try to stop such behavior, by limiting the types of debt they will buy. But this will not be easy.

2. The CLW will find it difficult to use rules or discretion to determine what types of debt are eligible for the facility. If the CLW tries to use rules, that is to define the types of debt that they will buy, firms will employ ‘financial engineering’ to create debt that nominally conforms to the rules but in fact remains highly risky. This was how the US wound up in a financial crisis in the mid-2000s: because firms created synthetic bonds that were rated AAA but were actually highly risky. Closer to home, there are also examples of bonds that were deemed safe – like the AAA-rated bonds issued by ILFS – that nonetheless ended up defaulting.

If the CLW consequently eschews rules and says instead that it will handle episodes using case-by-case discretion, users will fear that they cannot rely on the CLW, since such an approach would mean that other members could veto their attempt to unload their bonds to the facility.

We request the reader to not envision peaceful times, when some trades are taking place and spreads are fine, but instead to think of times when spreads are high, recent trades have stale prices, and a pall of fear hangs over the market. Consider a situation like late 2008, when bond prices were plummeting. At that time, buying bonds was considered a foolhardy act, comparable to ‘catching a falling knife’. Would a consortium of mutual funds really have the courage to intervene in this situation?

It is important to recall that the shareholders of the CLW are, themselves, bond market traders. They are the ones refusing to buy the bonds at any price on their own books – that is why the bonds are illiquid! So why would they allow their agent (the CLW) to do this? Consider the calculation of the other firms. If the CLW purchases the bonds, and the bonds default, the cost will have to be borne by the members of the cooperative. In contrast, if the CLW doesn’t purchase the bonds and the mutual fund is forced to shut down, the other firms might even benefit. Recall how the rest of the financial system `ganged up’ against LTCM in 1998, as they stood to gain from declining prices of LTCM’s positions.

3. Even when the CLW is willing to purchase bonds, it will not be easy to agree on a price. When bonds are illiquid, their price is not known to anyone. The distressed mutual fund will plead for a high price – and it will have a say in the running of the CLW. But other shareholders would object, as they would not want to suffer losses. So the Board of the CLW will work themselves into a tizzy trying to agree on a sale price.

Let’s assume the majority on the Board gets to decide the price.They will face an inherently difficult problem. Because the bonds are illiquid, the Board will need to guess the true value of the bonds on offer. And because the CLW would be running with an elevated leverage ratio, the consequences of guessing too high would be disastrous. At a 3:1 debt-equity ratio, a 30 percent fall in the price of the CLW’s assets would wipe out the entire equity capital. So, the CLW will need to offer a low price.

These three problems would haunt the CLW. It might freeze up with decision-making paralysis precisely at the times when decisive action is most required. Alternatively, it might proceed, but with excessive caution. It might purchase only select assets, meaning that many mutual funds facing runs would find the liquidity window closed. And even where the CLW was willing to purchase their assets, it is likely to offer a low price, which would prove ruinous to already-stressed MFs. These features interfere with the stated function of the CLW.

Many people remember stories from the Panic of 1907, where one person -- J.P. Morgan -- was the buyer of the last resort. This mechanism worked because Morgan was a self-interested profit-maximising individual who made a decision to use dry powder. He drove a hard bargain and purchased assets very cheaply, and turned a tremendous profit. He took enormous risk in the process, for he could have gone bankrupt himself. And, it could easily have been that the demand for liquidity insurance was bigger than his balance sheet, in which case his intervention would have gone badly wrong. For each J.P. Morgan who is celebrated for a 1907 event, there are many others who failed at various moments in history. We dream that a CLW will be able to think and act like J.P. Morgan, but its shareholders + board + management would find it impossible to have the entrepreneurial and risk-taking acumen of an individual. This is perhaps why we don’t see such a co-operative liquidity window in the world today.

A final point. We have stayed within the construct of no state intervention other than forcing MFs adding up to 75 percent of category AUM to become members. We fear, however, that when faced with the difficulties described above, the Indian state will not hold back even though there is no market failure. Once this happens, the familiar litany of state failure would commence.

We see this debate as a special case of a general principle. An economic policy strategy that addresses the surface symptoms is unlikely to work; the scope for financial engineering in public policy is very small. For a policy to succeed, it needs to engage in a root cause analysis, to address the underlying economic problem. If the problem is that investors are running from bond funds because they are inherently illiquid, then the only way to solve this problem is by reducing the mismatch between what these funds promise and what they can actually deliver. And that requires some fundamental financial reforms.

Hence, we would argue that the future of Indian finance remains along the strategy of the Financial Sector Legislative Reforms Commission (FSLRC). Once this is done, the need for a liquidity window will gradually fade away.

Bibliography

Mutual funds with feet of clay. Ajay Shah, Business Standard, 22 January 2018.

Runs on mutual funds. Renuka Sane, Ajay Shah, Bhargavi Zaveri. The Leap Blog, 12 October 2018.