The mathematics of credit. Pricing and unit economics.

Mejero Emmanuella
6 min readMar 8, 2024


Credit is a straightforward business. There’s just 2 things. The art and the mathematics.

The art is the relationships and that covers the macro. First you find and secure liquidity and cash supply. Next, you prospect and close customers that will likely not give you headaches. Then you give them money. Finally, you wake up every day and make libations to back your recovery strategy because, it’s not pretty otherwise. That’s the macro. It reads so straightforward, yet it is daunting. A full time job for multiple people that are dogged, determined, patient and great negotiators.

The micro are the numbers. How do you make enough money to survive. Think of the numbers in this way. You get money from Alpha at x cost and give that to Beta, Gamma and Delta at y cost. Whatever comes back to you has to be enough such that alpha gets their money at x cost, yet you can pay for z cost associated with collecting and giving the money, and still make you enough that is significant at the end of the year even though you will not get all your money back from Beta, Gamma and Delta.

I digress here to say that thinking that 100% of the money lent will be recovered is simply delusional. But being too comfortable with losing even the tiniest bit is detrimental.

The maths of it all. This will be lengthy, I warn.

Alpha hands you a $2 million liquidity contract.


  • Is this $2 million in tranches or fully available? If in tranches, what’s the structure?
  • Is this a credit line or a termed loan? If a credit line, is there a cost for when not drawn down and what’s the structure?
  • What volume can be currently process and what’s the ‘real’ projected growth?
  • What are the limitations of how this money can be disbursed according to Alpha’s terms and how does that affect the loan book structure?
  • What currency is the lending done in?
  • What’s the cost?

Before you read this and if you’ve never constructed debt portfolios, you might fall into the trap of never thinking that the first 4 terms matter. But consider;

  • If money is tranched, it can stunt your growth and make you lose good quality customers. If it is not tranched, you might not have the capacity to absorb it without recklessness.
  • If it’s a credit line and you don’t draw down a certain amount within a certain period, you might be required to pay a certain fee as the provider cannot have dead capital. If it’s termed debt, you will start getting charged from the very first day it hits your bank. Your deployment and pricing needs to be efficiently constructed for any of these scenarios.
  • If you don’t know how much you can absorb per time realistically you will run into capital flow needs or high cost.
  • If your client base vs the willing exposure of Alpha do not match, you are as good as having no customer and no capital.
  • If the currency you lend is the currency you borrow, this is straightforward. If it’s not, you must factor in currency volatility on both ends and how that applies to cost. Managing fx risk can simply be non-exposure at all, managed exposure by hedging or cost management by pricing; the options are listed from best to worst.
  • How much you are being charged. Your capital can come from an institution like Alpha but can also come from individuals/deposits and how much these cost matter in your pricing obviously.

You understand Alpha’s terms and sign the contract. Time for disbursement. How much do you charge customers?

It is not as simple as charging more than you were charged. Not only is it not that simple, how much you charge can directly impact the quality of your customers and loan book. A low fee is attractive to everyone but the higher the fee, the lesser the quality of customers you are attracting. You see, the more organized and reliable a person is, the more options they have. The highest tiers of customers are the ones the banks are begging to book and they are doing it at the best rate in the market. Those people and businesses have options. The lesser the likelihood of a person or entity to pay back, the less options they have, the more desperate they are and the more willing they are to accept ridiculous fees. My point is, having low fees doesn’t guarantee that all your customers will be high quality but the lesser your fees, the more you can compete for the higher quality.

So what goes into the cost

  • Cost of capital. This is obvious
  • Non performing loans. Uh-oh.
  • Cost of transactions. Someone has to pay for this.
  • Profit percentage. You do have to make some money.
  • Weighted inflation/value of money.

Cost of capital is pretty straightforward. If you are a bank or have the license and consent to receive and trade with customer deposits, you have the lowest cost of capital. For most commercial banks it is practically zero as they don’t give us any significant yield on our monies in the bank. For everyone else your capital will come at a price. Many times I have seen people talk about raising equity to lend. This is a naive venture; at best, you can do this in the periphery stage. The money you need to lend will keep growing larger as you grow and you can’t sell all of your business shares to secure capital for lending. In my opinion, the best alternative to operating like a bank is distributed capital cost. Low cost to mid cost multiple facilities where you are optimizing for cost, ease and supply. Dump high cost facilities. Don’t be desperate. Believe me, it’s never worth it.

Non performing loans (NPL) will come. It is accessory to the business. Do not be naive. Your aim is to optimize for the best quality of customer, manage the distribution within your loan book so that you are not over exposed and set up repayment strategies that help you to catch it early and keep it low. But it will be there and you will have to pay for it. And need to set a good buffer for it. Depending on how high risk your customers are, you will have to set a good annual percentage. Un-collateralized micro loans over digital platforms for instance, are really high risk; and from the data I’ve seen of a company with over 300,000 customers, you can expect anywhere from 40% to 65% loss of NPL (number, not value) in a month. You have to factor in that volume of principal lost, plus cost-of-principal loss into your pricing.

Transactions come with different costs and is unique to your application, underwriting and processing structure. Let’s say you have originators or a sales team that gets paid on commission, or you have to conduct KYC or a fresh underwriting for each loan request, these will all incure costs. There might also be insurance, bank processing or other type of cost that go with every transaction. The bottom line is that someone will have to pay for these. Typically, it will not be baked into interests as it is hard to do, it will instead be a one time separate processing or management fee.

Finally, you can make some money. You are not too greedy; you project to process $18 million this year and 10% of that is enough for you.

But the money you make in January doesn’t have the same value you make in December, it is depleting at an estimated 1.83% monthly. So you have to adjust for inflation. You can do some refined maths to sustain the value of your 10% or you can do rough maths and settle for 12%.

You add all of your cost, and get the annualized price. Other factors that can influence this price are; the point you capture your fees and pricing, the degree of liquidate-able security you acquire and how often your principal repayments are scheduled. One of the most important lessons I’ve learnt in lending is that when and how you can capture money is as important as how much you are charging.

The trouble with late fees is that they must be high enough that they terrify the borrower from incurring, but not too high that they discourage them totally whenever it is incurred. But you should charge it. You also can’t reduce late fees at any point because it signals that you are fine with people paying late.

Depending on what your annualized fee is, your longest maturity period and amount limit will be influenced. If your annualized fee is say 300%, you will mostly lend within a couple of months as anything higher would seem ridiculous, yet you will be limited to micro loans because any higher will have you termed ‘Shylock’ — there’s no evidence that you are not.

I always say there is money to be made in credit; ethically even. But there’s a lot of numbers and factors to keep your eye on. A lopsided loan book and you are now the debtor.

I hope you enjoyed reading this as much as I enjoyed writing it and I hope it brought some perspective and clarity to you.