Fintech is dead. Again. | Fintech Inside - Edition #85 - 21st Oct, 2024

Hi Insiders, I’m Osborne Saldanha, investor in early stage startups.

Welcome to the 85th edition of Fintech Inside. Fintech Inside is the front page of Fintech in emerging markets.

Last week, RBI introduced restrictions on four non-bank lenders. Among the four non-bank lenders, Navi was also included.

This caused quite a stir among those in the ecosystem who proclaimed that India’s fintech sector is dead. Again.

In today’s post, I discuss the nuances of the lending business, the unit economics, trickle-down effects and much more.

All this, in the hope that I can help you think through the nuances as you make your judgements about the future of finance in India.

Let me know what you think.

As usual, there’s also a beautiful song recommendation at the end, if you’d like to listen to a song in the background while you read this. Do share your recommendations with me too :)

Thank you for supporting me and sticking around. Enjoy another great week in fintech!

Update: Earlier last month, I concluded my role at Emphasis Ventures (EMVC). It's been a privilege to be a part of EMVC. What's next? I will be taking time off for a couple of months. Throughout my journey, a lot of people have been kind and generous with their time, network and insights. I’m keen to pay it forward in a small way. If you are considering starting up, I’d love to help. You can reach me here.

🤔 One Big Thought

Fintech is dead. Again. Long live fintech. Again.

The Reserve Bank of India (RBI), India's central bank and regulator of the country's banking, lending and payments sectors, on Thursday (17th Oct) released a notification of its action against four non-banking financial companies (NBFC), including two mirco-finance institutions (MFI).

The four regulated, licensed entities include:

1. Asirvad Micro Finance Limited, a subsidiary of Manappuram Finance Ltd, with INR 11,880cr ($1.4Bn) of AuM as of FY24 (ended Mar, 2024).

2. Arohan Financial Services Limited, a part of Aavishkaar Group, the impact venture fund, with INR 7,112cr ($846M) of AuM as of FY24.

3. DMI Finance Private Limited, with INR 11,844cr ($1.4Bn) in AuM as of FY24.

4. Navi Finserv Limited, Flipkart co-founder Sachin Bansal's financial group company, with INR 8,527.2cr ($1Bn) of AuM as of FY24.

As you can see, none of the entities are small - all of them have millions of users, and billions of dollars in AuM (loan book outstanding). They all provide their public numbers and have pretty high credit ratings as well (e.g. Navi rating as of Aug, 24).

So what action has RBI taken on these entities? The RBI has notified these companies that they need to "cease and desist from sanction and disbursal of loans, effective from close of business of October 21, 2024". Meaning, starting 22nd Oct, these entities cannot sanction new loans or disburse loans.

Why has the RBI taken this, seemingly drastic step? The RBI, in its notification, mentioned that it observed the below "material supervisory concerns" with these four entities:

1. Usurious (high) interest rates charged to users, to the extent that the weighted average lending rate (WALR) was excessive and and not in compliance with RBI's regulations on interest spread.

2. Non-conformity with provisions of Fair Practices Code.

3. Not adequately assessing household income to fixed monthly obligations and debt obligations.

4. Deviations in Income Recognition & Asset Classification (IR&AC) resulting in evergreening of loans, conduct of gold loan portfolio, mandated disclosure requirements on interest rates and fees, outsourcing of core financial services, etc.

These observations aren't small. They're pretty serious. Obviously, RBI hasn't specified what concern it had with which entity, but even then - no entity should go scot free if they're being usurious to low-income borrowers.

It's important to note that these are only business restrictions, not a cancellation of the above entities' licenses. RBI has mentioned that the entities will have to provide remedial action and evidence of compliance to RBI's satisfaction, post which RBI could lift these restrictions.

So why the hue and cry? The main issue that everyone seems to be having with this notification, in my opinion, is that RBI's notifications seem sudden and arbitrary - partly because RBI is not very specific in its guillotine-like, ominous notifications. In my experience, RBI is not sudden and certainly not arbitrary, but I do believe it can do better with its communication.

It seems to have also become a huge issue because of Navi's inclusion in the list. Navi is the newest, most tech forward of the lot. Sachin Bansal is an important figure in tech and now in fintech. Navi's inclusion by itself is not the issue, but in my opinion, it's because it seems that even a regulated fintech startup has to live in the fear that it's license could get taken away - similar to what Paytm faced earlier this year.

The issue then becomes that RBI wants only licensed entities to operate in India's financial landscape, but then even if they're licensed, there are restrictions and fear of license being taken away. It's probably what RBI wants - all fintech startups should be God RBI-fearing. :)

There's a lot of nuance to the lending business, which is important to understand, before you make a judgement in either direction. Though everyone's already proclaimed fintech is dead, again. But I'm here to say, long live fintech, again. So, let's get into that!

• What is a high interest rate? ¯\_(ツ)_/¯

In its notification, RBI mentions usurious interest rates. It even points to its regulation from 2016 and 2023, which doesn't give any specific number that it considers to be the ceiling for interest rates of unsecured loans. So what is high? It's almost impossible to even deduce what the portfolio level interest rate of an entity is by just looking at their financial statements. You'd have to get into loan level data.

In my experience though, I've seen RBI not like anything above what credit cards charge as annual interest rates i.e. about 35-45% - but that was during Covid, when repo rates were generally low.

Even during Covid, I know of RBI conducting similar exercises of evaluating interest rates at various regulated entities. RBI would give these entities time to normalise interest rates of their overall portfolios and bring it to a manageable interest rate.

Going outside our circle of very financially educated/competent people, if you ask the regular person on the street what they consider a good interest rate, they'd not know. Nobody likes doing mental math. Plain and simple. So yes, you need regulations to keep it manageable and avoid systemic risks. But then, if you tell the person on the street you'll give them a INR 10,000, 3-month loan for productive use cases, and they'd have to pay INR 3,750 monthly (50% annual interest rate), would they be able to afford it? They'd consider their income level and whether they'd be able to repay that amount, if its manageable, they'd make their own decision. Now, instead tell them that for the same loan value of INR 10,000, 3-month loan, they'd have to pay INR 3,625 monthly (35% annual interest rate), they'd probably jump at it - any discount is good. But the difference in monthly commitment of a 50% interest rate and 35% interest rate loan is so small that they'd almost miss it. Point is, the regular borrower doesn't think of their loan in percentage terms, tell them the absolute value and you're likely to close. Almost all regulated entities sell loans this way, and its why RBI had to introduce regulations to show the user a Key Facts Statement.

(Disclaimer, no bank will give you an offer that's 15% less than their original offer. It's usually only basis points difference. This was only for illustration purposes.)

• Fair Practices Code for Lenders needs an upgrade:

It's not surprising anymore that some banks are notorious at being usurious, having despicable debt collections practices, wrongly charging random fees with no recourse, denying loan applications without explanation and many other practices. It's no surprise to anyone. In fact, thanks to all these misselling and misrepresentation experiences that our parents went through, our millennial generation has had a terrible relationship with money.

One of the main reasons I got into fintech investing (I'm a computer science engineering graduate) was because in 2012/2013, a bank manager misallocated my education loan repayments only toward interest repayments (not the principal) and extended my loan tenor without informing us, and we didn't know any better. Our mistake was that we relied on the bank manager to be truthful to us.

It is without doubt in my mind (koolaid alert!) that application of technology, led by fintech startups, has made things significantly better for the general public. It has brought transparency, access and most importantly improved competition among regulated entities. Please don't throw UPI at me as a success of banks - it was started to quash Paytm's rising dominance and secondly, there are only two bank apps among top 10 UPI apps - at #6 is Axis Bank (acquired Freecharge) and #9 is ICICI Bank.

Since 2003 (the earliest that I could find), RBI has had the Fair Practices Code for Lenders (FPCL) in place with periodic updates via circulars. Until 2011, RBI maintained a Master Circular tracking all updates to this FPCL but post that didn't bother (from what I could find, I could be wrong) maintaining the the master circular - either it got fed up with banks or there were too many updates. The latest of these FPCL update circulars is Apr, 2024 where RBI notified regulated entities about fair practices with regard to interest charges. All banks have their modified, board approved FPCL specific to their organization - see Axis Bank, Union Bank, HDFC Bank and others.

The catch though, IMO, is that these FPCL's are limited to being a guideline for Standard Operating Procedure in lending. The FPCL falls short of talking about avoiding discrimination on grounds of religion, location etc, or fair interest charges and penal fees, or fair collection practices. Sure, there are regulations around some of these aspects but the FPCL could do a lot more.

• Loan underwriting and portfolio management is more art than science:

Here's the simple thing about loan interest rates and underwriting - everybody knows this:

The more secure (less risky) you are as a borrower, the lower your interest rate. The less secure (more risky) you are as a borrower, the higher your interest rate.

Broadly, if you're employed, have monthly income above a few INR lacs and have a high credit score, you'll get an unsecured loan in the mid-teens in this environment. Interest rates are low because everyone wants to lend this segment and everyone will try to give you a better (lower) interest rate to attract you to borrow from them. This segment of borrower (demand) is less in number but supply (lender) is high, so pricing is lower.

At the other end of the spectrum, if you don't have credit history or maybe work as a freelancer or don't earn more than INR 50K or so a month, you're considered risky and not of interest to lenders. This segment of borrower (demand) is high in number but no supply, so pricing is higher.

The middle segment is where the meat is and where the money is made and where the portfolio management becomes an art.

Now, the important thing about loan portfolio management is that the central bank rates have increased over the past 2-3 years. From earlier 4% to now 6.5%. That 2.5% interest rate increase trickles down to non-bank lenders and distribution agents and snowballs to a much higher interest rate as each entity needs to make their own spreads.

For example (simplified for illustration purposes): a bank borrows at that 6.5% and further lends to its borrowers at 10% making a neat 3.5% interest spread. A small bank or larger NBFC borrows from this bank at 10% and further lends to its borrowers at 14% making a 4% interest spread. This larger NBFC borrows at 14% and further lends to its borrowers at 18-20% making a decent 4-6% spread. As you have gone down this chain, it's already gone from 6.5% to 20%+ for the end borrower. Now add on risk based pricing, default cost coverage, loan security provision, loan to security value coverage, loan use case, loan tenor and many other variables and you've gone much higher than 35% annual interest. Are you still with me?

This interest spread or yield or net interest income (interest income minus interest expense) is another cause of concern for RBI. Again, RBI has not provided a ceiling where it feels comfortable about interest spread, but it expects this spread to be typically less than 10%. In that "less than 10%" interest spread plus processing fees, penalties etc., (banks typically earn 75-85% of their income from interest - Edition #65), the lender has to manage its cost of operations, cost of defaults, cost of customer acquisition and more. In the next section, you'll understand why it becomes tough to manage all of those costs within the interest spread.

• Fintech lending math is not math'ing (buckle up):

The paradox of Indian banks is that they are largely risk-averse (the paradox being that banking is inherently about pricing and taking risk). But this risk averse nature of banking is understandable. Financial markets are built on trust and trust is this ephemeral flaky thing that could be gone in an instant. If there are signs of cracks it could have serious systemic issues. I've written about this extensively in previous posts: Edition #81, #75, and several more.

Therefore, given this nature, banks have limited exposure to high risk borrowers. What you define as high risk borrowers differs from shop to shop, but to give you a sense, it's broadly of the categories: new to credit, self-employed, low income, women and many others.

Other than the perception that these are high risk borrower categories, banks simply don't want to deal with the costs of collections, cost of defaults, cost trust erosion and more. On the other end of the spectrum, micro-finance institutions go offline and take on the risk of those mostly excluded by banking and even non-bank lending (yes yes, I know MFI's are non-bank lenders). We've not even reached addressing the informal lending economy yet.

This "massive messy middle" of borrower categories who aren't too risky nor too safe, is where margins are made and lost.

It's no wonder that fintech startups account for 77% of the market for unsecured loans of value less than INR 1lac ($1,200). It's not that fintech startups discovered a regulatory arbitrage - there is none. It's not because fintech startups were better at it than banks. It's not because fintech startups have excess capital that they want to give away - though sometimes it may feel like that. It's simply because nobody else will do it and that was the opportunity. And along with it, comes higher default rates.

Banks will only touch (lend to), let's say, less than 10% of this borrower class. Larger NBFC's will probably lend to another 10-15% of this category. Optimistically, they'll probably lend to 20%. That still leaves a massive (60-70%) untouched segment of potentially risky borrowers that none of the legacy institutions will lend to. That's financial exclusion.

Some years ago, entrepreneurs found this to be a massive opportunity coupled with the tailwinds of growing smartphone and internet penetration. Little did they know, there was a huge supply of fintech startups who wanted to target the same users, so cost of acquisition shot up, leaving the industry with first world cost of acquisition and third world income streams.

Then, the inherent math of lending to this risky category is that default rates are much higher at 5% or more. Banks are able to maintain less than 1.5% default rates because a majority of their loan portfolio is in secured housing loans, which are inherently less risky at less than 0.5% default rates. A company (fintech startup) built on providing loans to risky borrower segments, will find it very tough to maintain a sub-1% default rate.

In the end, at a less than 10% interest spread you have 3-5% default rate plus first world cost of acquisition, then add on other costs and that equation doesn't make sense beyond a point. LHS is not equal to RHS.

So sometimes fintech lenders needed to increase the interest rates to be able to cover the cost, to make the unit economics of the loan make sense - not at the cost of the borrower committing suicide. But making money/profit is against our Indian culture (/s).

CRED's Kunal Shah comes out as a genius (non-consensus and right) in all this for targeting only high credit score individuals. The company was often ridiculed for "excluding" mass India. But that's now reaping rewards severalfold given their revenues and growth.

The challenge with everything I just mentioned above, is that nobody wants nuance. Everyone wants the hot take, the reductive take. And so no matter what I say here, the fact is that the majority will believe that fintech is dead.

Has that fazed my conviction in the potential of technology in financial services? not one bit. There's a LOT more work to be done.

Has it made my work as an investor that much harder? yes - it has been getting a lot harder to find exceptional entrepreneurs build in this sector. As I've written in Edition #83, talent moving out of the financial services sector is the worst side effect of the "Fintech is dead" and "regulation is killing innovation" narrative. The sector needs more intelligent, risk-taking and execution-focused people. If the exodus of people continues, it may not be the best thing for the future of finance in India.

The need for ecosystem education and collaboration is now more important than ever and I’m here for it.

1-min Anonymous Feedback: Your feedback helps me improve this newsletter. Click UPVOTE 👍🏽 or DOWNVOTE 👎🏽

🎵 Song on loop

Fintech updates can get boring, so here's an earworm: Been obsessing over Survivin’ by Bastille, lately (Youtube/Spotify). Song of the times, I guess! :)

👋🏾 That's all Folks

If you’ve made it this far - thanks! As always, you can always reach me at [email protected]. I’d genuinely appreciate any and all feedback. If you liked what you read, please consider sharing or subscribing.

1-min Anonymous Feedback: Your feedback helps me improve this newsletter. Click UPVOTE 👍🏽 or DOWNVOTE 👎🏽

See you in the next edition.