Credit Analysis and Underwriting for Unsecured Personal Loans
How credit analysts underwrite workplace-banking loans — the business-rules engine (PEP, CRB, debt-income, retirement age), bank-statement and net-pay checks, enhanced due diligence and the limits on varying a loan.
Credit analysis is the point in workplace-banking lending where everything gathered upstream — the application, the payslip, the documents, the external searches — is weighed against the bank's appetite and turned into a decision. For unsecured check-off (payroll-deduction) loans, that decision rests on affordability and conduct rather than collateral, so the discipline of the analysis matters a great deal. Within the Creodata Workplace Banking platform, credit analysis and approval are stages 6 and 7 of a thirteen-stage workflow, each gated by role-based authorisation and supported by a deterministic business-rules engine. This article walks through how an analyst actually reviews a file, what the rules check, where enhanced due diligence applies, and how the loan is sized to the affordability the file can demonstrate.
The split-screen review
By the time a file reaches the Credit Analyst, the application has already passed through pre-sale calculation, customer application, document collection, sales-manager review and document verification. The analyst does not start from a blank page. Instead, the platform presents a split screen: the full application details — applicant, contact, employment, loan and any buy-off details — sit beside the results of the external checks and the preliminary business-rule outcomes. The intention is to remove the need to switch between systems and to let the analyst compare what the customer declared against what the searches returned, line by line.
The external results shown alongside the application come from the platform's integration layer, each cached for 24 hours so that re-opening a file does not re-trigger every call:
- IPRS — national ID or passport verification, with bio-data auto-populated for the applicant.
- CRB — Credit Reference Bureau data covering performing and non-performing accounts and any arrears.
- KRA — tax PIN verification.
- Comply Advantage — AML, sanctions and politically-exposed-person (PEP) screening.
Presented next to these are the preliminary business-rule results, so the analyst can see at a glance which deterministic checks have passed and which need attention before forming a recommendation. The mechanics of how each external check is wired in — the adapters, retries and circuit breakers — are covered in compliance checks in loan origination.
The deterministic business-rules engine
A recurring question from lenders is whether the platform "scores" credit with a model. It does not. The engine is deterministic: a defined set of rules, each with a clear pass or fail outcome, applied to the data in front of the analyst. This is a design choice, not a limitation — an unsecured workplace loan is approved or declined against policy thresholds the bank can articulate and audit, not against an opaque statistical output. (Machine-learning-driven screening belongs to a different domain; for ongoing monitoring and risk modelling depth, see the AML compliance platform and the customer risk-assessment model.)
The rules the engine evaluates include:
| Rule | What it tests |
|---|---|
| PEP check | Whether the applicant is a politically exposed person, flagged from screening |
| Minimum take-home | Net pay must remain at or above KES 50,000 after the new instalment (a bank-configurable policy figure that applies even where there is a buy-off) |
| Retirement age | Loan maturity must not fall within three months of the applicant's retirement |
| Debt-income ratio / one-third | Affordability against the debt-to-income ratio and/or one-third of basic salary |
| CRB rules | Treatment of performing, non-performing and arrears positions |
| Employment-contract maturity | Tenure must not run within three months of contract expiry |
| Existing facilities | Existing loans and obligations are accounted for |
| Net-pay verification | Payslip net pay reconciled against bank-statement salary credits |
Because the figures are deterministic, two analysts reviewing the same file with the same data reach the same rule outcomes. The minimum take-home of KES 50,000 is the clearest example: it is enforced even where a buy-off is involved, and any loan being taken over must itself be included in the affordability picture. The way payslip allowances and deductions feed that affordability calculation is set out in the loan affordability check from a payslip.
Reading the bank statement
The business rules establish whether a loan fits on paper; the bank statement tells the analyst how the customer actually behaves. Net-pay verification — reconciling the payslip's net figure against the credits landing in the account — is the formal rule, but a competent analyst reads further. The statement review typically considers:
- Salary credits — are they present, regular and consistent with the declared employer and net pay?
- Turnovers — the overall level of activity through the account.
- Account conduct — how the account has been operated over the period under review.
- Standing instructions — committed outflows that reduce genuinely disposable income.
- Unpaid items and excesses — returned items or unauthorised overdrawn positions that signal strain.
- Arrears — missed or late obligations, read together with the CRB position.
For a customer who is new to the bank, salary-credit statements are an explicit documentation requirement — the original latest three months, stamped and signed, with a call-back to authenticate them. For an existing customer, separate statements are not needed where salary credits are already evident in Finacle. The full set of items the analyst expects to see, and how it differs for new versus existing customers, is detailed in the loan application documents checklist.
Sizing the loan to affordability
Credit analysis is where the requested loan meets the affordability the file can actually demonstrate. The maximum borrowing capacity is fixed by the affordability check and the minimum take-home of KES 50,000, which holds even where a buy-off is involved; the question at this stage is what the analyst does when the amount applied for sits above that ceiling.
When a customer applies for an amount the assessment cannot support — for example, where the resulting instalment would push net pay below the KES 50,000 floor — the file does not simply pass through. The analyst's available decisions are to comment, forward or recommend, return the file with a reason, or decline it with a reason; only the Credit Analyst and the Credit Approver can record an Approve decision. The outcome is that the loan that proceeds is the one the payslip and statement can carry, not the headline figure first requested. Because the analysis settles the affordable amount before the file moves on, the equated instalment that flows into check-off booking and disbursement reflects a genuinely affordable position rather than an inflated one.
Enhanced due diligence and employee verification
Beyond the documentary and statement checks, the platform applies enhanced due diligence (EDD) through employee verification — confirming with the employer, via an agreed mode, that the applicant is genuinely employed on the stated terms. The confirmation is recorded: the platform captures the mode used (a recorded telephone line or email), the person contacted, and the date, time and number.
Where this verification sits in the workflow depends on the employer:
- Government ministries are exempted from the verification at the credit stage, because sales managers and relationship managers already perform the call-back earlier in the process.
- Private entities receive employee verification as enhanced due diligence at the credit stage, in addition to the standard checks.
This split avoids duplicating a control that has already been carried out for government schemes while ensuring private-sector files carry the additional layer of scrutiny. EDD here is specific to confirming employment; for the broader treatment of when and how enhanced due diligence is applied across a customer relationship, see the enhanced due diligence guide and, again, the compliance checks overview. The analyst also confirms KYC compliance as part of the same review.
From recommendation to approval
Credit analysis ends with a recommendation, not a final sign-off. The Credit Analyst and the Credit Approver are the only two roles in the workflow permitted to take an Approve decision; every other role can comment, forward, recommend, return with a reason or decline with a reason. Once the analyst has completed the review — including any reduction to the requested amount — the file is recommended to the Credit Approver, who applies the bank's approval limits before the loan can proceed.
From there the path depends on the loan type. New loans and top-ups move to check-off booking and then loan booking; buy-offs route to loan booking and a buy-off clearance step. Each transition is authorised by role and recorded against a stage. The full stage-by-stage map, including the decline and return reasons available as a managed dropdown, is set out in the workplace-banking loan workflow. Credit analysis is one chapter of a longer story, and the complete guide to workplace banking places it in context alongside origination, booking and disbursement.
Frequently asked questions
Does the platform use AI to score credit?
No. Credit decisions on the Workplace Banking platform are driven by a deterministic business-rules engine, not a machine-learning model. The engine applies a defined set of checks — PEP screening, the minimum take-home of KES 50,000, the retirement-age rule, debt-income and one-third-of-basic ratios, CRB rules, employment-contract maturity, existing facilities and net-pay verification — each producing a clear pass or fail outcome that an analyst can see and an auditor can trace. The same data yields the same result every time. Machine-learning techniques are used elsewhere, for AML monitoring and screening on the separate AML compliance platform, but not for sizing or scoring an unsecured workplace loan.
How is the loan amount determined at credit analysis?
The amount is settled against affordability rather than the headline figure first requested. The affordability check totals the allowances on the customer's payslip, subtracts every deduction for the selected employer's scheme, and applies the minimum take-home of KES 50,000 — a bank-configurable floor that holds even where a buy-off is involved, with any loan being taken over included in the calculation. Where the requested amount cannot be supported, the analyst can return or decline the file with a reason; only the Credit Analyst and Credit Approver can approve. The result is that the instalment flowing into loan booking and disbursement reflects genuine repayment capacity.
When is enhanced due diligence required at the credit stage?
Enhanced due diligence in the form of employee verification is applied at the credit stage for applicants from private entities. It involves confirming employment with the employer through an agreed mode — a recorded telephone line or email — with the platform recording the person contacted and the date, time and number. Applicants from government ministries are exempted at credit, because sales managers and relationship managers already perform the employer call-back earlier in the process, so the control is not duplicated. This keeps scrutiny proportionate while ensuring private-sector files carry the additional layer of verification on top of the standard IPRS, CRB, KRA and screening checks.
To see how credit analysis fits within the wider origination, booking and disbursement flow, explore the Workplace Banking product page or book a demo.