Case 004Unit economicsHard

Kistvik Credit makes 30-day pay-later loans of Rs 8,000 for a 3% fee. With its funding cost, defaults and servicing costs, does each loan make money, and at what default rate does it break even?

The situation

Kistvik Credit lets shoppers pay for purchases 30 days later. The average loan is Rs 8,000. The merchant pays Kistvik a 3% fee on each loan at checkout, so the fee is earned whether or not the shopper repays. Kistvik borrows the money it lends at 13% a year.

Four per cent of loans default, and collections recover 20% of a defaulted loan. Acquiring the shopper plus servicing and collections cost Rs 110 a loan on average. The company is raising a Series B on the strength of fast loan growth.

Your task

Build the profit on one loan. Is Kistvik making or losing money per loan, what default rate would it need to break even, and what would you ask the founders?

Quick check

Before working it: does each Rs 8,000 loan make or lose money?

Worked solution

Try it on paper, then open one step at a time.

30-second answerThe answer to give first

Each loan loses about Rs 213, and Kistvik breaks even only if defaults fall below about 0.7%. The 3% fee brings Rs 240. Funding a month at 13% costs about Rs 87, expected credit loss is Rs 256, and servicing is Rs 110. Credit loss alone exceeds the fee, so growth in loan volume grows the loss. The business needs repeat borrowers, a higher fee or far better underwriting.

Step 1How do you build the profit on a single loan?

Picture a friend who lends Rs 8,000 to shoppers for a month, charges the shop Rs 240 for the favour, borrows the Rs 8,000 from a bank herself, and pays a collector to chase late payers. Her profit is the fee less three costs. Every lending business reduces to four lines per loan: income, the cost of the money, the expected loss, and the cost of running the loan. Compute each in rupees, not percentages, so they can be added.

Fee: 3% of Rs 8,000 is Rs 240. Funding: 13% a year for one month is 13% divided by 12 on Rs 8,000, about Rs 87. Expected credit lossThe average loss per loan from defaults: the chance of default times the share of the loan that is not recovered, times the loan size.: 4% times 80% not recovered times Rs 8,000, Rs 256. Acquisition and servicing: Rs 110. The loan loses about Rs 213, and the credit loss on its own is larger than the entire fee.

One Rs 8,000 loan, from fee to result, in rupees+240Fee earned-87Funding cost-256Credit loss-110Acquisitionand servicing-213Result per loan0
On one Rs 8,000 loan Kistvik earns a Rs 240 fee, then pays about Rs 87 to fund it, expects Rs 256 of credit loss and spends Rs 110 on acquisition and servicing, a loss of about Rs 213 per loan.
Step 2At what default rate does the loan break even?

Take everything except credit loss: Rs 240 less Rs 87 less Rs 110 leaves Rs 43 to absorb defaults. Each defaulted loan costs Rs 6,400 after the 20% recovery. Rs 43 over Rs 6,400 is a break-even default rate of 0.68%, against the 4% Kistvik actually sees, so defaults would have to fall by about five sixths. Unsecured consumer credit at under 1% default is prime-customer territory, not a pay-later book built for reach.

The relationship
d∗=240−86.7−1108,000×(1−0.20)=43.36,400≈0.68%d^{*} = \frac{240 - 86.7 - 110}{8{,}000 \times (1 - 0.20)} = \frac{43.3}{6{,}400} \approx 0.68\%
240the merchant fee, 3% of Rs 8,000
86.7one month of funding at 13% a year
110acquisition and servicing per loan
6,400loss on a defaulted loan after 20% recovery
What it says in wordsThe break-even default rate is the money left after every other cost, divided by what one default costs.
Profit per loan against the default rate: the room for error is under 1%+40-100-2000Break-even: 0.68% defaultsKistvik at 4%: -Rs 2130%1%2%3%4%5%Share of loans that default (20% recovered)Rs per loan
Kistvik's profit per loan falls by Rs 64 for every point of default rate, crossing zero at about 0.68% and reaching a loss of about Rs 213 at the 4% default rate it actually runs.
Step 3What could rescue the model, and what would you ask?

Three levers, each with a number. Break-even at today's defaults needs a fee of about 5.7%, which merchants are unlikely to pay. Repeat use helps: if most of the Rs 110 is acquisition and a shopper borrows six times a year, the per-loan cost falls sharply, but only if repeat borrowers default less, not more. Ask for defaults by cohort and by loan number, because a pay-later book is only as good as its repeat customers.

Close with the view: on these numbers loan growth multiplies the loss, so the Series B is funding a subsidy, not a business. A yes needs evidence that second and third loans default at under 1%, or a plan to lift the fee by selling merchants something they value, such as higher basket sizes.

Where candidates lose it

The trap is comparing the 3% fee with a 13% funding rate and concluding the fee is too low, or too high, without converting both to the same period. Thirty days at 13% a year is about 1.1%, not 13%.

The second is ignoring recoveries or treating the default rate as the loss rate. A 4% default rate with 20% recovered is a 3.2% loss, Rs 256 on the loan, and that single line decides the case.

What the interviewer asks next

  • If the shopper, not the merchant, paid the fee only on repayment, how would the numbers change?
  • Kistvik says defaults on repeat loans are 1.5%. What share of loans must be repeats for the book to break even?
  • How would a rise in funding cost to 16% move the break-even default rate?

Company names and figures are illustrative.