T&S Asset Management

Lesson 1 of 8

What a lending system must do, and where companies underestimate it

A lending system is more than an application screen and a credit decision. This lesson sets out the six functions every loan passes through, why most of the work comes after the money is paid out, the areas that are easy to underestimate, what supervisors learned about data and reporting, and the questions to settle before choosing a system.

Published 2 November 2026 · Listen to this lesson

Six functions, one loan

Whatever the product, instalments at a shop, an advance on wages, finance for a dealer or a supplier, every loan passes through the same functions. A working description, in the order a loan meets them:

  • Origination. Receives the application, identifies the customer, records consent and collects the data the decision needs.
  • Credit decision. Applies the lender's rules and scores and returns approve, decline or refer, with a limit and a price, and records which version of the rules was used.
  • Disbursement. Shows the borrower the terms, records acceptance and sends the money to the right account.
  • Repayment and collections. Keeps the schedule, receives payments, matches them to loans, and follows up arrears.
  • Accounting. Posts every movement, interest, fees and provisions to the general ledger, the company's main accounting record.
  • Reporting. Produces reports for management, the board, auditors, the credit bureau and the regulator.

A company can build each function, licence it or have a partner run it, but no loan can skip one. Lesson 1 of the course on how a digital lending platform is built follows the same parts in more technical detail.

Most of the work comes after the money is paid out

A demonstration of a lending system usually shows the application and the decision, because they happen once and are easy to see. The rest of the loan's life is longer. Each instalment creates a due date, a payment to match, possibly a reminder, a late charge or a request to change the terms, and each month ends with accruals and reports.

Example, with illustrative numbers: a company makes 10,000 loans a year, each repaid in 12 monthly instalments. It takes 10,000 credit decisions, but its system must handle 120,000 due dates and a comparable number of payments to receive and match, plus every reminder, change and correction. The decision is the step most often discussed; the instalments are where most of the daily work happens.

Repayment events over the life of a book = number of loans × number of instalments per loan (before reminders, changes and corrections).

Where companies usually underestimate

Five areas are easy to underestimate because they are hard to see in a demonstration and only appear once loans are live:

  • Matching payments. Payments arrive late, early, in part, twice, or without a usable reference, and each one must still be posted to the right loan.
  • Changes after disbursement. Early settlement, restructuring, waivers of fees and corrections must change the loan without erasing what it was before.
  • Accounting and month-end. The total of loan balances must agree with the general ledger every day, and provisions must be calculated from the same data.
  • Reporting. Regulators, auditors, funders and the credit bureau each ask for their own reports, in their own formats and on their own dates.
  • Existing loans. A company that already lends must move its current loans, with their history, into the new system and prove that the balances agree.

The Basel Committee on Banking Supervision's 2021 revised principles for the sound management of operational risk list regular verification and reconciliation of transactions and accounts among the traditional internal controls. In a lending system, reconciliation is not an extra feature; it is how the company knows that its loan book is right.

Data that adds up: what supervisors learned

Reporting is only as good as the data each function captures. The Basel Committee's principles for effective risk data aggregation and risk reporting, published in January 2013 and known as BCBS 239, start from the observation that the financial crisis that began in 2007 revealed that many banks, including global systemically important banks, were unable to aggregate risk exposures and identify concentrations fully, quickly and accurately. Their implementation is meant to strengthen banks' risk data aggregation capabilities and internal risk reporting practices, and so their risk management, in particular at global systemically important banks.

The same point applies to provisions. The Basel Committee's 2015 guidance on credit risk and accounting for expected credit losses asks banks to have a credit risk assessment and measurement process that provides a strong basis for common systems, tools and data to assess credit risk and to account for expected credit losses. A non-bank lender is not bound by these bank standards, but its auditors and funders will ask the same question: can the numbers in the report be traced back to individual loans?

Questions to settle before choosing a system

Before comparing suppliers or designs, the project team can answer a short set of questions in writing. They become the requirements against which every option is judged:

  • Which products, in which markets, in which currencies, and at what volume in the first year and in three to five years?
  • Which of the six functions will the company run itself, and which will a partner run?
  • Which reports are required, by whom and how often, including those for the regulator and the credit bureau?
  • How will repayments arrive, and who will match the ones that cannot be matched automatically?
  • Are there existing loans to move, and how will the balances be proven to agree after the move?

Lesson 2 uses these answers to compare building, licensing and outsourcing.

Sources