Episode 75 · Fintech · 52 min

The dead asset on every SME's books

KredX's claim is that India's 60 million SMEs are already sitting on the credit they need: an accepted invoice, parked as a dead asset while a large buyer takes 90 days to pay. Sell it at a discount and a founder who could produce four times a year can produce 365 — and because that invoice is really the buyer's liability, an investor gets to price Infosys, not the small supplier. Seven years and 300-plus anchor businesses in, the harder claim is the second one: bank is not required, banking is.

MK
Manish Kumar and Devang Mundra
Founder & CEO; CTO, KredX · with Vishal Krishna
The dead asset on every SME's books — episode thumbnail
51:46
Said in this episode
▶ 0:00
60 million
SMEs in India
Vishal's opening figure, immediately hedged on air as 'depending on whose data you access' — the constant is that they need an operating system for their finances.
▶ 0:44
300+
Businesses KredX works with
Cited in the intro and again as the anchor base whose payment behaviour is predictable enough to underwrite against.
▶ 26:15
80–90%
Share of invoice value funded upfront
The remaining 10–20% is released only when the anchor pays, keeping the SME's skin in the game and its collection follow-up intact.
▶ 19:11
20%
Return that made an early investor call back three times
His words, quoted by Manish: friends were in a five-to-six percent asset class. Returns are known upfront like a fixed-income product, and are not a promise of future yield.
▶ 40:17
160 / 60
Total team, and the tech team inside it
Sixty of KredX's roughly 160 people are engineers — Vishal's point being that on-demand infrastructure has replaced the need for a very large organisation.
▶ 46:28
~50
Parameters in the in-house risk score
Adopted after an externally AA-rated anchor defaulted during the IL&FS crisis and took a wholly dependent vendor with it; the figure was stated in passing and the caption is imprecise.
The brief

The argument in sixty seconds

Manish Kumar's claim is that Indian SMEs are less short of credit than short of a way to sell what they already own. An accepted invoice is a receivable — a dead asset sitting on the books while a large buyer takes 60 or 90 days to pay — and because that same invoice is a liability on the buyer's books, selling it transfers a blue-chip's credit quality rather than the small supplier's. That arbitrage is what turns invoice discounting into an asset class instead of a loan, and it is why KredX went to family offices, treasuries and HNIs before it went to banks: retail capital is the hard side to build, and once bankers had their own money on the platform they became the champions who brought their institutions in. Factoring itself is more than a century old; the contrarian bet was that it could be done without the relationship, the location and the emotion — on data alone. Devang Mundra's half of the argument is that there was nothing to copy, so the first product was a digital experience stitched out of DocuSign and Google Sheets while the real stack got built, and the underwriting had to read the seller, the buyer, the relationship and the macro at once — a lesson learned expensively when an externally AA-rated anchor defaulted in the IL&FS year and took its single-customer supplier down with it. The wider stakes sit outside the company: banks will not touch anything below AA, SME-to-SME credit still cannot prove a transaction isn't phony, and until contracts are reliably enforceable, foreign money stays away and the cost of financing stays high.

Worth your time if you are

SME founders waiting 90 days on a large buyer
HNIs and family offices hunting fixed-income yield
Fintech operators building underwriting from thin data
Bankers wondering what is left of the branch
Product leaders selling to customers who aren't tech-savvy
Episode map

Where the conversation travels

Every block is a chapter, coloured by what it's about. Click any of it to jump straight to that minute on YouTube.

01Cold open: 60 million SMEs, one cash cycle 0:00 Vishal frames the problem — large companies live off SME credit, SME founders drown in the cash cycle, and banks want relationships nobody has — before introducing KredX's Manish Kumar and Devang Mundra and a platform already working with 300-plus businesses. 02The dead asset, and the arbitrage inside it 1:32 Manish's founding theory: banks ask for your house, eyes, ears and kidneys as collateral, yet the SME already holds an asset it never sells — the accepted invoice — and because that receivable is the large buyer's liability, discounting it prices the buyer's credit, not the supplier's, which is what makes it an asset class rather than century-old factoring done over a handshake. 03A digital experience before a digital product 5:43 The first MVP was a bid-and-auction exchange that died on the time value of money — nobody wants to wait for a bid — so the team scrapped it and ran on DocuSign, Google Forms, Sheets and Drive for five or six months, buying engineering the runway to build the real thing. 04Nothing to copy, and nobody tech-savvy 7:57 Devang describes building without a cookie-cutter reference anywhere in the world, for sellers who are not always tech-savvy and investors who need the models visible before they will trust the asset — while the team learned, transaction by transaction, how people feel when money arrives on the 70th day instead of the 67th. 05Retail capital first, banks second 13:31 Manish's rule for doing anything new in finance — change the source of capital, the source of data, or the product itself — leads to the counter-intuitive sequencing: build the hard retail side first, and let family offices, treasuries and bankers investing their own money become the champions who pull institutions in. 06Cutting the bill-rediscounting middlemen out 15:30 A big slice of what retail already owns through mutual funds is this same discounted paper, repackaged through the bill-rediscounting chain, so KredX's pitch is to delete two intermediaries and let an investor build the portfolio directly — 20% one anchor, 30% another, 5% a startup. 07The investor who felt like a genius 18:13 An early investor phones back three times after his first repayments — the last time to say his friends are in a five or six percent asset class while he is making twenty, and can he refer a friend — which is the network effect KredX then had to systematise into a published internal score and anchor-level notification preferences. 08Why e-commerce became the favourite sector 22:10 E-commerce became the most-subscribed sector because it was the one segment still paying through COVID, and because sale seasons make marketplaces shorten supplier credit from about 45 days to 30 to keep volumes flowing. 09Onboarding by API, underwriting in-house 23:32 Sellers onboard through DigiLocker, GST, bank-statement and accounting-system feeds, a proprietary model and in-house underwriting team read the seller, the buyer, the relationship and the macro together, and ERP integrations then let each new order flow in, get discounted and reconcile itself. 10With recourse: 80% now, 20% on payment 25:45 KredX funds 80 to 90 percent of invoice value and holds the rest back so the SME keeps skin in the game and keeps chasing its own buyer — because a financier knocking on the anchor's door damages the very relationship the credit depends on. 11Real transaction or phony? 27:28 Asked why Indian SMEs never reach US scale, Manish points at two missing stacks — data good enough to tell a real transaction from a phony one, and legal recourse worth using — noting banks won't touch anything below AA while cash-flow triggers watch whether the money that should arrive each day actually does. 12ITFS, account aggregator, and foreign money 31:25 Devang argues the value of government-backed rails like ITFS, GST e-invoicing and account aggregator is as much regulatory as informational: foreign investors funding a five-trillion-dollar economy need clarity about the avenues, not just richer data. 13Four pieces of a B2B transaction 34:40 Manish splits every B2B transaction into payables and receivables, the payment itself, post-payment reconciliation and filing, and financing — KredX covers all four because the only reason an SME shares its data is a lower cost of payment and of capital, which in turn needs enforceable contracts before international lenders will price India cheaply. 14Bank is not required, banking is 38:34 On why they don't chase an NBFC licence: banks exist because there was no internet, and the generational ask has moved from a branch in the village to an ATM to simply being available — so KredX would rather hand a bank the technology to open a current account on the fly than replicate the branch. 15Saudi, Singapore, and the worst NPA 43:20 The same instrument travels — Saudi is live, Egypt and the UAE next, with conversations in Bangladesh, Singapore and Africa, allowing for Sharia-compliant tweaks — and the hardest lesson came in the IL&FS year, when an externally AA-rated anchor defaulted along with a vendor wholly dependent on it, after which KredX dropped external ratings for its own multi-parameter score. 16Red teams, flat org, favourite movies 46:50 Devang treats credit risk like security — internal and external audits, red and blue teams gaming the platform, a culture where a new joiner can question the founder — before the closing round on daughters, mindless action movies and When Genius Failed.
Takeaways

Ideas to carry out of this hour

01

The receivable is an asset the SME never thought to sell

Banks answer the SME credit gap with collateral — the house, and everything short of a kidney. Manish's reframe is that the business already holds an asset it treats as dead: the accepted invoice, rolling over every 60 or 90 days. His worked example is deliberately small — ten lakh from friends and family, goods billed at twelve lakh on 90-day terms, the invoice sold at eleven — and the punchline is throughput, not margin: instead of producing four times a year you can produce 365 times, at a lower percentage but far higher absolute profit.

02

You are not lending to the supplier; you are pricing the buyer

When a small vendor bills a large company, both sides make an entry — a receivable for one, a payable for the other. So when the vendor sells the receivable, it is also selling the large company's liability, priced at the vendor's cost of funds. That mismatch is the arbitrage that makes invoice discounting attractive as an asset class rather than charity toward small business, and it is the whole reason an HNI will look at the paper at all.

03

To do something new in finance, change capital, data or the product

Manish's test: if you are not different on your source of capital, your source of data, or the product itself, you are repackaging a bank. A genuinely new product is rare, and any data edge gets competed away as the ecosystem digitises — which left capital. Seven years ago there was very little data outside the banks, so KredX built where it could be structurally different, on where the money comes from.

04

Build the hard side of the market first

Retail and institutional capital move inversely, Manish argues — when one arrives the other leaves — so a platform needs both to stay complementary. Institutional money is the easier side to raise, so KredX deliberately began with the harder one: treasuries, family offices, funds and HNIs. The payoff was distribution nobody budgeted for. Bankers who put personal money on the platform watched it come back and then volunteered to plug their own bank or NBFC in.

05

Sell a digital experience before you have a digital product

The first build was a bid-and-auction exchange, tested over ten or fifteen trades, and it failed on a human fact: people who want money want it now, and people who want to deploy want to deploy now. Rather than stall, the team ran the business on DocuSign, Google Forms, Sheets and Drive — no product, but an experience customers already read as a leap forward — which bought engineering five or six months to build the real stack while requirements changed daily.

06

Leave the buyer-seller relationship alone — that is the product design

KredX lends with recourse and funds only 80 to 90 percent of invoice value, holding the balance until the anchor actually pays. The retained slice is skin in the game, but the deeper point is that the SME, not the financier, keeps chasing collection: a lender knocking on the buyer's door is a stranger endangering the commercial relationship the credit rests on. Verification stays light — a contact-point check that the company exists — with everything else digital.

07

Bank is not required, banking is

Asked repeatedly why they don't become an NBFC, Manish gives the thesis instead: banks were invented because a person with surplus money in one city had no way to find a borrower in another, and technology has since removed that constraint. The customer ask has migrated from a branch in the village, to an ATM, to simply being available. So the ambition is to hand banks the rails — current account, financing, invoice discounting, payables and receivables done on the fly — rather than rebuild the branch with a licence.

08

Credit risk behaves like security, so run red teams against yourself

The formative loss was an anchor rated AA by an external agency that defaulted in the IL&FS period, dragging down a supplier almost entirely dependent on it; KredX dropped external ratings afterwards for an in-house score across roughly fifty parameters. Devang's framing is that credit and risk are a lot like hacking — every safeguard you build meets people who are immensely creative — which is why the company runs internal and external audits and red and blue teams against its own platform, and quotes James Dyson on staying permanently dissatisfied.

The numbers, drawn

What the episode measures

Every figure below was said on air — timestamps included, caveats kept.

Conversation share

portion of the hour spent on each theme
Credit & lending · 24%Payments & fintech · 15%Data & digitisation · 14%Savings & wealth · 13%Product strategy · 11%Regulation & policy · 9%
Credit & lending24%
Payments & fintech15%
Data & digitisation14%
Savings & wealth13%
Product strategy11%
Regulation & policy9%
Computed from the chapter map of this episode.

What selling the invoice does to a year

production cycles per year
Waiting 90 days to b4Discounting every in365
Manish's illustrative arithmetic on air, not company data: ten lakh of family capital, goods billed at twelve lakh on 90-day terms turns roughly four times a year; sell each invoice at eleven lakh and you can rebill daily — lower margin, higher absolute profit.▶ 3:30

Yields, as described on air

% return
Large bank savings r2A friend's usual ass6The app rate Manish 8Early invoice-discou20
Four numbers said in different parts of the conversation and not directly comparable: the 2% and 8% come from Manish's hypothetical about a two-lakh-employee bank, the 5–6% and 20% from an early investor's phone call. Tenor and risk differ, and none of these is a quoted product return.▶ 19:11
Worth keeping

Lines that stay

You go to banks and banks will say: give me your house, give me your eyes, ears, kidney — whatever you can give me — and then we'll finance you.

— Manish Kumar ▶ 2:01

What I said is, we will give people a digital experience — we may not have a digital product.

— Manish Kumar ▶ 7:12

My friends are putting money in a five, six percent asset class and I'm making twenty percent. I'm feeling like a bloody genius.

— Manish Kumar, quoting an early investor ▶ 19:11

Bank is not required, banking is.

— Manish Kumar ▶ 38:49

Credit and risk is a lot like hacking. As soon as you build protections and safeguards, people are immensely creative — so we can never be satisfied.

— Devang Mundra ▶ 46:57
Clips that travel

Short on time? Start here

SME founders waiting 90 days on a large buyer

The dead asset, and the arbitrage inside it

The whole thesis in four minutes: collateral versus receivables, four production cycles a year becoming 365, and why selling your invoice sells the buyer's credit.

1:46 → 5:43 · 4 min ▶ Watch clip
Fintech operators building underwriting from thin data

Change the capital, or you're a bank in a wrapper

Manish's three-way test for doing anything new in finance, and why KredX built the harder retail side first — then watched bankers become its salesmen.

13:31 → 15:30 · 2 min ▶ Watch clip
HNIs and family offices hunting fixed-income yield

The investor who felt like a genius

The three-phone-call anecdote that explains the platform's referral engine, followed by how those preferences hardened into a published score and anchor-level alerts.

18:13 → 22:10 · 4 min ▶ Watch clip
Product leaders selling to customers who aren't tech-savvy

With recourse, and hands off the relationship

The most practical stretch: fund 80–90 percent, hold the rest, and never let the financier be the one knocking on the buyer's door.

25:45 → 27:28 · 2 min ▶ Watch clip
Bankers wondering what is left of the branch

Bank is not required, banking is

The answer to why KredX won't chase an NBFC licence, told through three generations of what customers asked a bank for: a branch, an ATM, availability.

38:34 → 40:30 · 2 min ▶ Watch clip
Glossary

The jargon, unpacked

Invoice discounting
Selling an accepted invoice before its due date for slightly less than face value, so the supplier gets cash now and the funder collects the full amount when the buyer pays.
Account receivable
Money a business is owed for goods already delivered — on KredX's telling, a dead asset sitting on the books for 60 to 90 days until someone buys it.
Anchor
The large buyer at the other end of the invoice, whose credit quality is what the investor is really pricing; KredX says it works with 300-plus of them.
With recourse
The structure where the SME remains liable if its buyer fails to pay — here reinforced by funding only 80 to 90 percent of the invoice and releasing the rest on payment.
Bill rediscounting
Described on air as the older bank route — banks discount bills, bundle them and sell them on to mutual funds, which pass them to retail investors; KredX's pitch is to remove those two layers.
Cash-flow underwriting
Watching whether money that should flow in and out of a business each day actually does, and firing a trigger when the expected amount arrives at a fraction of its usual level.
ITFS
The government-backed international trade financing framework the guests say KredX has signed up to as one of a few companies; the value they claim from it is regulatory confidence for foreign lenders as much as data.
Account aggregator
India's consent-based framework for sharing financial data between institutions — cited alongside GST and e-invoicing as the rails that let lenders build richer models on SMEs.
Connections

If this resonated, go here next

Full transcript

The whole conversation, searchable

207 segments

Auto-generated captions, lightly cleaned. Click a timestamp to open that moment on YouTube.