The missing layer in the lending market (2): Borrower decision infrastructure

By John Pabi NTIFO, Product Architect, Credit and Digital Lending

Ghana’s credit ecosystem has invested heavily in approving loans. It has invested almost nothing in helping borrowers decide whether and which loan to take. That asymmetry has a measurable cost.

Ghana’s credit infrastructure is, in structural terms, almost entirely one directional. Significant institutional investment has been made in tools that help lenders assess borrowers: credit bureaus, income verification systems, collateral registries, scoring models, and now increasingly, mobile data derived creditworthiness assessments. What does not exist, in any systematic or accessible form, is the equivalent infrastructure on the other side of the transaction. There are no tools that help borrowers assess loans before they take them.

This asymmetry is so deeply embedded in the architecture of the lending market that it rarely gets named as a problem. It is simply the way lending works: lenders decide whether to approve; borrowers decide whether to accept. The second decision is treated as trivial, as though the existence of an offer were sufficient evidence of its suitability. In practice, this assumption causes a great deal of preventable harm.

The Approval Bias in Credit Infrastructure

When the Credit Reference Bureau system was established in Ghana, it represented a genuine advance in market infrastructure. XDS Data, Ghana’s first credit reference bureau, became fully operational in 2010, followed by Dun & Bradstreet Ghana receiving its full license in 2012. Lenders gained access to consolidated repayment histories, outstanding debt obligations, and default records across participating institutions. The system works within the parameters it was designed to serve.

Those parameters are, however, exclusively lender facing. A borrower approaching a microfinance institution today can obtain their own credit report, but only as a document attesting to their creditworthiness in the eyes of lenders. There is no complementary infrastructure that generates, for the borrower, an assessment of whether a specific proposed loan is affordable given their specific income profile, existing obligations, and planned use of funds.

What borrowers lack at point of borrowing

• Micro borrowers in urban Ghana classified as over-indebted based on repayment sacrifices: 30%
• Micro borrowers in Accra who struggle with at least one repayment instalment: 74%
• Reduction in borrowing when total cash cost was shown upfront rather than as APR: 11%
• Ghanaians with sufficient financial literacy to understand loan interest calculations: 32%

The consequences of this information asymmetry become visible in repayment data. Ghana’s NPL ratio of 20.6% at the end 2023 substantially overstates lender risk and understates borrower stress.

Why borrowers do not self assess

The absence of pre loan self-assessment among Ghanaian borrowers is not primarily a financial literacy problem, though financial literacy gaps are real. It is a structural problem: the tools required for meaningful self-assessment do not exist in accessible, borrower facing form.

To properly evaluate a loan offer, a borrower needs to be able to answer four questions. First, what is the true cost of this loan, not the stated monthly rate, but the effective annual percentage rate inclusive of all fees, charges, and the method of interest calculation. Second, what will my actual monthly cash outflow be, and does that outflow fit within my realistic income projection for the loan period? Third, given my existing obligations, what is my genuine debt service capacity? Fourth, if my income falls below its current level, at what point does this loan become unserviceable?

The loan approval process confirms that a borrower is lendable. It says nothing about whether borrowing is wise. These are entirely different questions, and conflating them is the market’s most consequential design omission.

Three Components of Borrower Decision Infrastructure

What would it look like to build the missing layer? Drawing on the literature in behavioural finance, consumer credit design, and emerging practice in East and Southern African fintech markets, three functional components stand out as essential.

True cost affordability scoring

The first component is an affordability assessment that calculates the effective cost of a loan, not the stated rate, and maps that cost against the borrower’s actual income profile. This is distinct from a credit score, which assesses historical repayment behaviour. An affordability score assesses forward looking cash flow viability for a specific loan product at a specific moment.

The input required are stated monthly income, existing monthly debt obligations, essential household expenditure, and the proposed loan’s full repayment schedule. The output is not a binary approval signal but a cash flow map: how much disposable income remains after debt service in each repayment period, and what income reduction would push the borrower into stress?

Kenyan fintech Pezesha has deployed income buffer metrics that flag loans where repayment obligations risk exceeding a borrower’s realistic net income. This approach has drawn attention from impact investors and development finance institutions precisely because it addresses the affordability question that conventional credit scoring ignores. The methodology is transferable to the Ghanaian context with appropriate calibration for local income distribution patterns.

Loan regret indicators

The second component addresses a question that conventional credit underwriting ignores entirely: at the time of application, does the borrower show signs of decision making patterns associated with future regret? Behavioural finance research identifies several reliable pre decision indicators: borrowing under immediate social pressure without a clear repayment plan; taking a larger loan than initially intended because a larger amount was offered; inadequate understanding of the loan’s true cost; and a repayment schedule that requires the borrower to change their financial behaviour rather than accommodate their existing one.

These indicators can be surfaced through brief structured pre loan interactions. These are not gatekeeping mechanisms, but reflective prompts. Research from the Financial Diaries project, which tracked detailed financial flows of low income households in Ghana, South Africa, and Tanzania, found that borrowers who engaged in even a brief structured reflection before borrowing were significantly more likely to take appropriately sized loans and report satisfaction with the decision six months later.

Pre loan simulation tools

The third component is the most technically straightforward and arguably the most immediately impactful: a simulation tool that shows a borrower, before they sign, exactly what their financial position will look like at each point in the loan’s life. This is not a repayment schedule. Every loan contract already contains one. It is a cash flow simulation that overlays the repayment schedule on a realistic projection of the borrower’s income, flags the months where repayment will be most stressful, and presents the total cost of the loan in absolute cedi terms rather than percentage rates.

Research by Bertrand and Morse, published in the Journal of Finance, demonstrated that presenting total cost in cash terms produces significantly more accurate borrower understanding of loan cost and meaningfully changes borrowing behaviour. The finding has been replicated in multiple low income consumer credit contexts.

Applied example: simulation tool output

Borrower profile: Informal caterer, Kumasi. Monthly income: GH₵2,400 average, ranging from GH₵1,600 (low months) to GH₵3,200 (peak months: December, Easter). Proposed loan: GH₵8,000 at 6% per month flat, 10 months.

What the stated terms suggest: Monthly repayment of GH₵1,280. Repayment to income ratio: 53%. Feasible in peak months, infeasible in three projected low months.

What a simulation tool surfaces: Total repayment GH₵12,800. Effective APR approximately 108%. In three low income months, the repayment will require drawing down savings or borrowing. Alternative: GH₵6,000 loan with same structure. Total repayment GH₵9,600, feasible across full income range.

The borrower takes GH₵6,000. The simulation did not prevent borrowing; it prevented overborrowing.

Why the market has not built this

The absence of borrower decision infrastructure in Ghana’s lending market is not an accident of oversight. It is, at least in part, a rational consequence of incentive structures. Lenders are compensated on origination volume and interest income. Tools that cause borrowers to borrow less, or to choose a different product, reduce both. Without regulatory pressure or competitive differentiation incentives, the market will not spontaneously build tools that work against its immediate commercial interests.

This is precisely the gap where regulatory intervention, civil society tooling, or well positioned fintech entrants can create disproportionate value. The Bank of Ghana’s Consumer Protection Department has signaled interest in mandatory pre loan disclosure standards. Borrower decision tools would provide the mechanism through which such disclosures could become genuinely useful rather than compliance formalities.

The third article in this series looks forward at the emerging technology and product design approaches that could make borrower first lending not just theoretically possible but commercially viable. It also examines the conditions in Ghana’s evolving fintech landscape that make now the moment to build them.

NOTES & SOURCES

• XDS Data (now part of TransUnion Ghana) became operational 2010; Dun & Bradstreet Ghana received full operational license 2012. Bank of Ghana, Credit Reporting Act, 2007 (Act 726).
• Bank of Ghana, Financial Stability Review 2023. NPL ratio December 2023: 20.6%. For borrower-side stress data see Collins, D. et al., Portfolios of the Poor, Princeton University Press, 2009; CGAP Financial Diaries with Smallholder Families, 2021.
• Pezesha Financial Services. Company described in CGAP FinTech for Financial Inclusion publications, 2022-2023.
• Collins, D., Morduch, J., Rutherford, S., Ruthven, O., Portfolios of the Poor: How the World’s Poor Live on $2 a Day, Princeton University Press, 2009. West Africa field data: Kaffenberger, M. et al., CGAP Financial Diaries with Smallholder Families.
• Bertrand, M. and Morse, A., ‘Information Disclosure, Cognitive Biases, and Payday Borrowing,’ Journal of Finance, 66(6), 2011.

John is a Product Architect, Credit and Digital Lending

Post Views: 40