Hello everyone, 

Welcome to the 208th edition of The Pattern. 

Earlier this month, I was in Nairobi for the Africa Fintech Forum. A few days, a handful of conversations with lenders across markets that look very different on paper: different products, different regulations, and entirely different customer segments. 

But somewhere around the third conversation, the differences stopped mattering. 

What I was noticing wasn't what people were talking about. It was what had changed about how they were talking about it. 

The vocabulary was familiar: approval rates, portfolio quality, and collections efficiency. These are the same words you hear in every lending room, in every market. What was different was the order they came in. 

A few years ago, you'd walk into a room and growth was the first word out of everyone's mouth. Everything else—risk, quality, sustainability—was organised around it. Risk teams weren't there to question the direction; they were there to make the journey possible. 

In Nairobi, that had flipped. 

Growth was still on the agenda, but it had moved downstream. The questions coming first were different: 

• How resilient is the book? 

• Can we see stress before it shows up on the balance sheet? 

• Will this portfolio still look healthy twelve months from now? 

I flew back to India and spent the next few days catching up on the week's news. The RBI circular. The latest microfinance data. 

And they were all saying exactly the same thing. 

The RBI drew a line that was already being crossed 

The RBI on June 15, 2026, issued the Reserve Bank of India (Commercial Banks — Responsible Business Conduct) Second Amendment Directions, 2026. 

It formally banned dark patterns in digital interfaces — design and UX techniques that steer customers toward actions they didn't choose. It mandated explicit, recorded consent before any product sale, made the default option "No," and said that where mis-selling is established, the lender refunds the entire amount, with no carve-outs for channel, agent, or DSA.

Reading through the news, what struck me was how familiar most of it felt. Pre-selected checkboxes. Insurance bundled into loan flows without clear disclosure. Consent screens built to be clicked through, not read. 

None of this was particularly unusual. In many organisations, it was just how the journey worked. 

This isn't a newly discovered problem. Most people in the industry would recognise these practices immediately. It's converting a known, tolerated practice into an enforceable liability. The product team that designed for conversion, the DSA that got paid per signup, the compliance function that signed off on "technically, the checkbox was there" — they're all in a different position now. Refund liability travels backward through the origination chain. This dynamic, where distribution stops being a free-for-all and starts carrying legal consequence, is exactly what Rajat unpacked in RBI's new sales rules: What changes when distribution stops being free-for-all. The loan UX, as he put it, just became a board-level issue. 

For a CRO reviewing the existing book, any product distributed through a third-party channel that leans on pre-selected consent deserves a closer look. The ability to query that exposure in plain language — the way Sentinel AI now allows through its MCP-ready interface — matters more in an environment where the regulator can ask the same question at any time. 
 
The microfinance recovery that isn't straightforward 

On the flight back, I pulled up the CRIF MicroLend Lite June 2026 report. And one number made me put my phone down for a moment. 

Disbursals dropped to ₹20,239 crore in April from ₹29,543 crore in March. And yet, portfolio quality improved. 

Across the top 10 states, active loan volumes declined month-on-month. Yet PAR 31-180 improved across each of them. That's an unusual combination. Lending activity slowed, but portfolio quality continued to improve. It suggests the sector is becoming more selective about where growth comes from. 

It connected directly to what I'd been hearing all week. The lenders I met in Africa weren't just optimising for volume. They were grappling with a harder question: Will this growth still look healthy twelve months from now? 

That's one reason many lenders keep a close eye on microfinance. Stress tends to surface here earlier than it does in larger-ticket retail portfolios: 

• Borrower margins are thinner, so a single income shock translates into a missed EMI faster than it would for a salaried urban borrower. 

• Repayment cycles are shorter, meaning delinquency becomes visible in weeks, not quarters. 

• A borrower who is 1-30 days past due today can slip into the 31-90 bucket surprisingly quickly. 

What the MFI book does today, unsecured personal loans and small-ticket MSME portfolios tend to do several months later, just less visibly and with less warning. 

What stood out in this report was that portfolio quality improved even as lending activity slowed. Across the industry, there seems to be a growing focus on the quality of the book rather than the pace at which it grows. That selective approach — doing more with an efficient, well-instrumented origination process rather than a wider, faster one — is increasingly where the advantage lives. 

I spent all of last week debating this very point, only to find the exact same story waiting for me in the data on my flight home.

Growth, with conditions 

There's a version of what happened this week where you read two separate stories: a consumer protection circular and a sector data report. Tidy, unrelated, each filed in its own column.

But that's not how the lenders I sat with in Nairobi were reading their environment. They weren't separating the regulatory from the portfolio. They were treating it all as the same question, asked through different channels: how do you keep growing when the room for error is smaller than it used to be? 

Everyone was still talking about growth. They just weren't starting there anymore. 

The answer, in every market where I've seen it work, isn't better forecasting. It isn't faster origination or a more aggressive acquisition engine. It's visibility — knowing what's actually in your book before the dashboard tells you, catching the stress that's building rather than the stress that's already arrived. (Related: How Superflows puts your risk team in control of verification for digital lending). 
 
The shift in how lenders are approaching onboarding tells the same story. The move from branches to bots isn't just a cost play — it's an information play. Digital-first journeys create the data trails that make early stress detection possible. What Cars24's 80% reduction in loan TAT demonstrated wasn't just speed; it was that a tighter, more instrumented process produces better data, which produces better risk decisions downstream. 

When growth was abundant, speed was an advantage. In a tighter environment, visibility may matter more. 

After all, portfolios don't break when risk appears. They break when risk goes unnoticed. 

Reading list 

• Indian cenbank tightens rules to curb mis-selling by lenders, bans 'dark patterns' 

• MicroLend Lite, June 2026 

• PSBs may lose the edge over private peers as tightening liquidity, rising credit costs may impact growth 

• Oil retreat hands RBI an assist in boosting rupee's near-term outlook 

• Personal loans from 8.95%: Compare June 2026 rates of HDFC Bank, ICICI Bank, Axis Bank and other top private lenders 
  
  

Thank you for reading. If you liked this edition, forward it to your friends, peers, and colleagues. You can also connect with me on X here and follow FinBox on LinkedIn to get the latest updates.   

Cheers, 
Mayank