Independent Research and Policy Advocacy

Response to Niti Aayog’s Discussion Paper on “Digital Banks – A Proposal for Licensing and Regulatory Regime in India”

Save Post


Identifying the Market Failure and Mis-identifying the Intervention

The discussion paper highlights a ‘market failure’ in the debt market for MSMEs wherein incumbent banks are unwilling to lend to MSMEs, who are otherwise creditworthy, due to lack of documentation and thereby leading to suboptimal outcomes for both banks and MSMEs[3]. Adding to these frictions is the high operational cost structure of banks which makes it unviable for them to underwrite small ticket MSME loans[4]. The paper then states that NBFCs, while being able to solve the underwriting problem that banks have, are unable to raise funds as cheaply as banks, due to the absence of a ‘bank license premium’, which, the paper claims, is a premium offered by the credit market to borrowing entities.   

As a solution to these problems, the paper suggests the creation of DBs, which are likely to have a much lower cost structure. However, the discussion paper also suggests that the onerous compliance requirements on MSMEs, for availing loan from banks, flows from the status of banks as regulated entities and fiduciaries[5]. Thus, we conclude that DBs, which would also be regulated entities and fiduciaries akin to banks, will be subject to the same compliance requirements as banks and thus the existing frictions that the paper points to, would continue to persist for the DB construct too.

Economics of the DB Construct

The discussion paper does little to explain the economics of the DB construct and how DBs would achieve profitability. It contrasts the low per-account operating cost of Webank in China with the cost-to income ratios of banks in India to suggest DBs in India would be profitable. However, the Customer Acquisition Cost (CAC) is expected to be relatively high and is expected to rise with greater competition[6]. Given that small ticket, low tenure MSME and retail loans would form a significant market for DBs, the CACs would not be a one-time cost but likely to be a recurring cost[7]. Even assuming a low operating cost structure, the historically high NPA levels in this segment implies that DBs would still face high credit costs[8]. The high credit costs, reflected by the provisioning for expected losses on loans, and relative concentration of their portfolio in the urban areas, would adversely affect their profitability. The discussion paper does not provide any arguments or illustrations of how expected losses can be brought down through the DB construct and hence, by not doing so, it does not provide a credible argument for how the DB construct will solve the key cost challenges of serving low-income and under-served segments. Also, the other argument for the profitability of the DB construct hinges on access to low-cost retail demand deposits. The reality of the artificially low-cost nature of retail deposits reflects the woefully low competition in interest rate offerings on retail demand deposits among banks in India. Unlike most banks worldwide, Indian banks enjoy virtually guaranteed profits from the ‘depositor penalty’ that they collectively and unfairly impose on their depositors. In the U.S., for instance, the ‘depositor penalty’ has been close to 0.10% on an average in the past six years, as compared to 4% in India over the 10 years since deposit rates have been deregulated[9]

Needless to say, with the majority of India’s population in rural areas and experiencing low rural broadband penetration, it is unclear how the discussion paper concludes that DBs would not be handicapped as compared to traditional banks when it comes to furthering financial inclusion[10][11]. Indeed, the example the paper describes as the target customer is that of a gourmet café/ bakery incorporated as a private limited company in an urban center.

RBI’s Mandate – Protecting Retail Depositors or Venture Capital investors?

Banking is necessarily a low-risk business. Banks work with two classes of investors, those who seek low risk in exchange for low returns and supply large amounts of capital, and a smaller number that seek much higher returns and are willing to take on much higher levels of risk. In a bank, the first are the depositors and the second are those that supply risk-capital. Well-managed banks then combine these two forms of capital and ensure the aggregate level of risk that the bank takes on matches the expectations of its combined investor group[12].

Against this, the DB license seeks to use demand deposits (presumably retail, with permissions from the RBI) to experiment untested solutions and solve for the MSME credit gap and financial inclusion gap in the country. It therefore ignores the need for banks to be low-risk in character and provides a direct channel for shifting losses from high-risk-high return-seeking equity and Venture Capital (VC) investors to retail depositors, without having demonstrated any evidence of low-risk (low-NPA) origination. This would be unacceptable and goes against the basic tenets of banking.

Is the Digital Bank Solving Problems Unique to Neo Banks?

Neo-banks are consumer facing entities which partner with incumbent banks to offer banking services, like deposits and credit, to consumers. In providing these banking services, Neo-banks rely on the balance sheet of the partner bank[13]. Through partnership with banks, these Neo-banks perform the function of a DB. However, the discussion paper identifies three major challenges to growth that Neo-banks currently face and posits that allowing them access to a DB-like construct could solve for these problems[14]. The challenges identified are[15]

  1. Limited Revenue Potential – Neo-banks currently have only 2 sources of revenue – 1) Channel partner fees and 2) Interchange on payments. The discussion paper contends that the limited sources of revenue threatens their viability and thus could hinder financial inclusion.
  2. Potential Obsolescence of the Partner Bank Core Banking System (CBS) – Neo-bank offerings are constrained by the product suite offered by their partner banks. This limits their ability to customise products for their customers.
  3. High cost of capital and no entry barriers – Relying only on equity funding, Neo-banks face a high cost capital for funding their operations. Additionally, the lack of entry barriers allows entry of all types of actors and creates consumer protection risk.

It should be noted here that none of the challenges outlined by the discussion paper stem from regulation or market failure. These are business constraints that are present due to the interplay of market forces. Taken together with the fact that these Neo-banks operate mainly for a niche consumer segment in urban areas[16], it is a stretch to contend that the demise of these Neo-banks would significantly hinder financial inclusion, thus requiring regulatory intervention by RBI.

A close examination of the challenges also reveals that none of them are unique to Neo-banks and some of them are quite misleading. For instance, the challenges of limited revenue opportunities and high cost of funding is not true as there is nothing stopping Neo-banks from applying for an NBFC or bank license, which could allow them to get credit, deposit and payments permissions. The NBFC license would allow them to access capital that costs much lower than equity capital while also providing an additional revenue stream through provision of credit. The contention in the discussion paper that only access to low-cost funding can ensure their sustainability is ill-founded at best given that NBFCs have been more profitable, in recent times, than even banks[17]. The constraint imposed by the CBS of the partner bank is an operational issue that will get resolved as the business case for including more product codes become stronger. There is no role for RBI here. Finally, the concerns on consumer protection are largely mitigated as 1) Neo-banks do not handle customer monies and hence there is low risk of fraud and 2) Neo-banks will be subject to due diligence and ongoing monitoring by their partner banks, as is already the case for all outsourced activities of banks. Issues that have arisen on account of customer protection can further be mitigated if the RBI implements the recommendations of the RBI Committee on Digital Lending that seeks to bring under the regulatory radar all credit-related activities.

The discussion paper further provides two use cases of DBs, neither of which are unique to only DBs. The first use case is that of Banking as a Service (BaaS). In this, DBs lease out their technology stack to smaller banks, like cooperative banks, for whom buying the full suite technology stack would be unprofitable. Another example of BaaS, that the paper illustrates, is the leasing out of the DB’s card issuance infrastructure to NBFCs that want to offer credit cards. Incumbent banks can also provide these services and currently there are instances of NBFCs partnering with banks to provide credit card services. Additionally, Information Technology (IT) companies also provide the option to rent their CBS on a subscription basis[18]. Similarly, the use case presented on Augmented Credit Under-writing is just a description of the working of the Account Aggregator (AA) ecosystem and adds nothing new in terms of specialised business model or additional revenues for the DB.

The Licensing Regime and Regulatory Architecture

The discussion paper proposes a three-step licensing process[19]. In the first two steps, RBI vets all applicants and the selected ones get a restricted DB license to operate in a regulatory sandbox. The relaxations applicable in the regulatory sandbox is to be decided by RBI, in line with RBI’s guidelines on regulatory sandbox. The RBI will also decide the success metrics to evaluate the performance of the licensees in the sandbox. Licensees who perform satisfactorily are awarded a full stack DB license while those licensees not achieving the pre-defined metrics would have to exit by selling their assets and unwinding their liabilities as per the process laid down by RBI.

The regulatory architecture[20] proposed by the discussion paper brings parity between DBs and existing banks on prudential and liquidity risk regulation. It also proposes that DBs have the same level of access to infrastructure enablers, like Aadhar e-KYC, NEFT/RTGS, ATM network, DICGC and the Account Aggregator ecosystem, as incumbent banks. Since the operation of DBs will be almost entirely dependent on technology, the regulatory requirements include additional regulations on DBs. For one, they would need to have one or more of the controlling persons in the entity to have an established track record in one or more fields like e-commerce, payments, cloud computing and the like. Secondly, there would be conditions requiring ex ante technological preparedness and ex post business continuity planning.

The major relaxations, in comparison to incumbent banks, for DBs, are in minimum capital requirements and branch mandate policy. The minimum capital requirements for a DB are proposed as Rs. 200 Cr, equivalent to that of Small Finance Banks (SFBs), presumably because DBs are not expected to have any physical branches, though no explicit rationale is given.

The discussion paper suggests that the branch mandates be interpreted in a progressive manner to allow for banks to decide on the channel of delivery of services (without presumably diluting the objective of delivering banking services to defined unbanked areas). However, the latest Branch Authorisation Policy defines a Banking Outlet in terms of the banking services that it should provide and does not insist on a brick-and-mortar branch[21] and to that extent, it already reflects the suggestion made in the discussion paper. Thus, it is not clear if the paper suggests this as a relaxation specific to DBs or if it is to apply to all banks.

Do DBs have any Utility?

It is clear from the discussion above that DBs do not do anything to push the envelope on financial inclusion or systemic stability. The question then arises whether the DB construct has any utility at all. An answer emerges if we look at the competition aspect of licensing DBs. Even though incumbent banks have access to all the technological solutions available to DBs, they have not adequately employed these solutions to better customise their product offerings to customers.   Given that DBs can provide all products and services being provided by incumbent banks, they could potentially increase competition in the digital financial services space and force banks to accelerate the adoption of technological solutions[22][23].  This could not only provide more choice for well-banked consumers but also create an ecosystem where banks compete to offer products that are best suited to the customer instead of offering standardised products as they are doing currently.  


Unlike differentiated banks, which have limited banking functions and thus lower risk[24], DBs differ from incumbent banks only in the channel of delivery of services and with low broadband penetration in our country, it is unclear how DBs would substantially improve financial inclusion. The entire proposal for the creation of DBs is built around developing the business prospects of Neo-banks, and thus the exit valuations for VC investors, rather than around a solid case for increasing financial inclusion or reducing systemic risk in the banking sector. In particular, the primary motive of the proposal appears to be to allow Neo-banks, and other entities, to access low-cost retail deposits and provide credit without having to go through the existing licensing process. However, there is some merit in licensing DBs to accelerate competition in the digital financial service space, which could prove beneficial to consumers. Such a construct will need to be developed without resorting to the need to solve a financial inclusion problem[25].  

[1] Dvara Research has made several contributions to the Indian financial system and participated in engagements with many key regulators and the Government of India. We have had the honour of being the technical secretariat to the Reserve Bank of India (RBI) Committee on Comprehensive Financial Services for Small Businesses and Low-Income Households (Chair: Dr. Nachiket Mor), 2013, as well as peer reviewing the chapter on consumer protection (Chapter 5) of The Financial Sector Legislative Reforms Commission (FSLRC) Report, 2011. We have also worked with the Government of India’s Committee to Review Implementation of Informal Sector Pension, the RBI Committee on Household Finance (Chair: Dr. Tarun Ramadorai), the Indian Government’s Committee of Experts on a Data Protection Framework for India (Chair: Justice Srikrishna, 2018), the SEBI constituted Working Group on Social Stock Exchange, RBI’s Committee on Deepening of Digital Payments and the RBI’s Expert Committee on MSMEs, 2019, in our capacity as domain experts.

[2] Section 4, Digital Banks – A Proposal for Licensing & Regulatory Regime for India, Discussion Paper, Niti Aayog, November 2021

[3] Section 3, Ibid

[4] Section 4, Ibid

[5] Section 3, Ibid

[6] Lending companies stare at 30% jump in user acquisition costs, Pratik Bhakta, Economic Times, May 1, 2019

[7] The small tenure of the loans implies quicker churning of the loans thus necessitating acquisition of new customers or repeatedly reaching out to existing ones, Ibid

[8] See Exhibit 11, MSME Pulse, April 2020 – Transunion CIBIL and SIDBI

[9] Fixing India’s Banks: Making Banking Boring Again, Nachiket Mor & Deepti George, BloombergQuint, January 7, 2021

[10] Rural broadband penetration in India is only 29% compared the national figure of 51%, see Broadband for Inclusive Development – social, economic and business – Deloitte and CII report, Nov 2020. Additionally, usage of internet among MSMEs was abysmally low at only 4%, see Digital Infrastructure and Policy Initiatives in the MSME sector, Biswaraj Paul Choudhury, Dvara Blog, April 15, 2019

[11] “Digital Banks: How can they be regulated to deepen financial inclusion?” Slide Deck. Washington, D.C.: CGAP, Mehmet Kerse & Stefan Staschen, December 2021

[12] Why We Need Banks… And Getting Indian Banking Right, Nachiket Mor & Madhu Srinivas, BloombergQuint, December 29, 2020

[13] Section 4, Digital Banks – A Proposal for Licensing & Regulatory Regime for India, Discussion Paper, Niti Aayog, November 2021

[14] Section 5, Ibid

[15] Ibid

[16] Rising challenges for Indian neo-banks, Tarika Sethia, ET, August 4, 2021

[17] The ROA of NBFCs has been higher than banks for the past 5 years – source: The reports of the Trend and Progress of Banking in India of the respective years and STRBI Table No. 10 Bank Group Wise Selected Ratios of Scheduled Commercial Banks

[18] From conversations with market participants

[19] Section 7, Digital Banks – A Proposal for Licensing & Regulatory Regime for India, Discussion Paper, Niti Aayog, November 2021

[20] Ibid

[21] Rationalisation of Branch Authorisation Policy – Revision of Guidelines, RBI May 2017

[22] View: India’s banking revolution has started without the banks, Andy Mukherjee, Economic Times, December 13, 2021

[23] Digital Banks: How can they be regulated to deepen financial inclusion? Slide Deck. Washington, D.C.: CGAP, Mehmet Kerse & Stefan Staschen, December 2021

[24] Chapter 2.4, Report of the Committee on Comprehensive Financial Services for Small Businesses and Low Income Households, RBI, January 2014

[25] A similar conclusion was arrived at by CGAP’s researchers, where they conclude that “a broader question is whether it is useful to link financial inclusion and digital banks in the way it is done in some of the countries. Promoting financial inclusion and serving the excluded and underserved can be done through any existing FSP license and digitization could be a strategy for a bank regardless of whether it has a financial inclusion focus or not”. Slide 25, Digital Banks: How can they be regulated to deepen financial inclusion? Slide Deck. Washington, D.C.: CGAP, Mehmet Kerse & Stefan Staschen, December 2021

Authors :

Tags :

Share via :

Leave a Reply

Your email address will not be published. Required fields are marked *

Related Posts :