# The Unit Economics of B2B Lending: Spreads, Losses and Margins

Lending looks simple from the outside: take money at 9%, lend it at 16%, keep the difference. In reality, the gap between a profitable lending book and a slow-motion disaster hides in four line items — cost of funds, credit losses, operating costs, and capital. Anyone evaluating a B2B financing platform (as an investor, vendor partner, or borrower wondering why fees are what they are) should understand this arithmetic. Here it is, in plain terms.

The Four Building Blocks

1. Cost of funds

Every lender borrows before it lends. Banks use deposits (cheap, maybe 4–6% in India); NBFCs use bank borrowings, bonds, and securitisation (typically 8–12% depending on their credit rating and market conditions). A platform financing SaaS contracts through an NBFC partner inherits that cost. When RBI tightens or bond markets wobble, cost of funds rises — and every basis point comes straight out of margin.

2. Yield — what the borrower pays

The interest or fee charged to the buyer. For short-tenure B2B instalments, this is often quoted as a flat fee (say 2–3% of contract value for 12 months), which annualises to a higher effective rate. The yield must cover all costs plus profit, so it reflects the riskiness of the borrower segment: a GSTIN-verified mid-size distributor with clean filings prices very differently from an unregistered trader.

3. Credit losses

The killer line item. Lenders measure it as a percentage of the loan book:

  • Expected loss = Probability of Default × Loss Given Default × Exposure.
  • Well-underwritten B2B short-tenure books might target 1–3% annual credit losses. Consumer BNPL books have run far higher when underwriting got loose — a lesson the industry learned expensively in 2021–23.

A useful rule of thumb: every 1% of unexpected credit losses can wipe out several years of expected profit, because net margins in lending are thin (often 2–4% of assets). This is why underwriting discipline isn't a compliance nicety; it's the business model.

4. Operating costs

Origination (sales, distribution), underwriting (data, models, manual review), servicing (collections, mandates, support), and compliance. Digital-first models attack this hardest line item through automation: instant GSTIN-based approval collapses underwriting cost, and UPI Autopay mandates with smart retries collapse servicing cost. This is precisely the design logic behind platforms like KredFlow — the economics of small-ticket B2B financing only work if the cost to deliver a ₹2 lakh instalment plan is a few hundred rupees, not thousands.

Putting It Together: An Illustrative P&L

For a ₹100 crore B2B instalment book (hypothetical, representative numbers):

| Line | Amount |

|---|---|

| Yield earned | ₹16 crore |

| Cost of funds (10%) | −₹10 crore |

| Credit losses (2%) | −₹2 crore |

| Operating costs (2.5%) | −₹2.5 crore |

| Credit cost of capital | −₹0.5 crore |

| Pre-tax margin | ₹1 crore (1% of assets) |

That 1% is why scale matters so much in lending. Doubling the book roughly doubles the profit with modest cost increases — but a 2% surprise in credit losses erases the entire year.

Levers That Actually Move the Needle

Improve underwriting (lower PD)

Better data — GST filing patterns, bank-flow volatility, vendor-specific contract performance — reduces defaults more cheaply than any pricing change. Data proximity is the single biggest structural advantage an embedded platform has.

Recover more when things go wrong (lower LGD)

Mandates that retry automatically, early-warning signals from missed payments, and quick restructuring keep losses contained. Collections is where LGD is won or lost.

Cut servicing cost

Automated collections, self-serve portals, and clean escalation flows. Every rupee saved here drops to the bottom line at nearly 100% margin.

Recycle capital faster

Shorter tenures mean the same rupee of capital supports more lending per year, raising return on equity. Twelve-month SaaS instalments recycle faster than five-year equipment loans.

Why This Matters to Vendors and Buyers

If you're a SaaS vendor offering financing to buyers, the unit economics explain the design you'll encounter:

  • The vendor is paid upfront because the lender's margin depends on volume through a book it can underwrite — not on vendor credit risk.
  • Buyers pay a transparent fee because the lender must cover a real cost stack; "free" financing usually hides the cost in inflated prices.
  • Approval criteria are data-driven because the entire profit pool lives in keeping credit losses near 2%, not 6%.

Lending is a business of thin margins, fat tails, and compounding data advantages. The platforms that survive India's next credit cycle will be the ones whose spreads, losses, and operating costs were designed to coexist — from day one, at every ticket size.