Women-Owned SMEs Are Getting Loans. Just Not Enough of Them.

Evidence from Ethiopia, Lebanon, Ecuador, Türkiye and India points to a stubborn flaw in women’s SME finance: borrower numbers can rise while usable capital remains scarce. The programmes that changed outcomes altered the lender’s rules, from collateral and guarantees to customer journeys and cash-flow assessment. For capital providers, the real test is whether those changes produce profitable portfolios after the cost of risk-sharing is included.

El Asmar had a business, not a charity case.

Her Lebanese company, Madera Creation, lacked the conventional credit history that can end a bank application before anyone has properly examined the firm. BLC Bank, a Lebanese commercial bank, lent it $50,000 through its WE Initiative. Madera Creation expanded to 20 employees.

That is the neat version of the story: a women-focused product, an entrepreneur and a loan that worked.

The interesting part is why it worked. BLC did not simply advertise to women. It offered collateral-free loans to businesses operating for at least two years, trained staff on bias, and changed its digital, advisory and service channels. The eligibility rules moved.

That distinction matters because the wider numbers contain an awkward contradiction. In the International Finance Corporation’s July 2026 survey of emerging-market financial institutions, women-owned SMEs accounted for 27% of SME loans by number but only 19% by dollar volume. Their average loan was 28% smaller. Their non-performing-loan ratio, the share falling behind on repayment, was 3.6%, against 3.8% for SME portfolios overall.

More accounts. Less money. Slightly better repayment.

So the question is not merely whether women can obtain loans. It is whether the loan can finance the next real commercial step: buying stock before a busy season, purchasing equipment, paying workers while a customer pays late, or delivering a contract.

The label is not the lending decision

The conventional response to women’s SME finance is understandable. If too few women receive formal credit, create more women-labelled credit lines, grants and preferential products.

Targeted outreach has a place. But a label cannot, by itself, change what happens when an applicant lacks property collateral, needs more time to repay, or has sales records rather than formal accounts.

Imagine a shop owner who can finance one shelf of stock but not a full sales cycle. She may show up every year in a lender’s report as a successful borrower. Her business may still be stuck.

This is where most people stop looking. They count borrowers rather than asking four less glamorous questions: Was the loan large enough? Was the term long enough? What proof did the lender accept? And after repayment, could the business borrow more?

Ethiopia provides the clearest answer.

Inter-American Development Bank. Photo: APK / Wikimedia Commons, CC BY 4.0.

Ethiopia changed the size of the door

The World Bank-backed Women Entrepreneurship Development Project, or WEDP, began in 2012. By March 2025 it had financed about 30,000 women, normally with loans averaging roughly $10,000 and terms of 24 to 48 months.

Earlier portfolio reporting recorded repayment of 99.1% and an average loan near $12,000, 870% above borrowers’ previous average. Participating institutions reduced collateral requirements from around 200% of loan value in 2013 to 125% in 2018.

The lenders also began accepting vehicles, inventory, personal guarantees, psychometric scores and cash-flow analysis. Cash flow means the money entering and leaving a business over time. For a firm with busy customers but little formal paperwork, that can be more revealing than a title deed it does not own.

WEDP moved lenders beyond group-loan logic by accepting more evidence of repayment capacity. Photo: Tsion Molla / Pexels, Pexels licence (free commercial use).

According to the World Bank’s March 2025 account, participating firms recorded approximately 30% higher profits and 50% more employees than comparable firms. WEDP’s decisive move was not creating another women’s lending lane. It enabled microfinance institutions, lenders usually associated with very small loans, to make larger individual loans under different appraisal rules.

Here is the twist: the borrowers did not become less risky because a programme called them women entrepreneurs. The lenders became better at seeing their capacity to repay.

There is an important limit to the evidence. The material reports gains during the project’s operation, but does not establish whether the profit and employment gains persisted after programme support ended. That is not a small missing detail. A durable lending model should still work when the project team, concessional support and special attention are gone.

Lebanon found that price was not the complaint

BLC Bank launched its WE Initiative in 2012 with IFC support after customer research. The findings were not chiefly about cheaper money. Women reported distrust, poor treatment and processes that did not fit their needs.

The bank responded with collateral-free loans for qualifying businesses, staff training, digital services and advisory support. Between year-end 2011 and 2015, its number of women SME borrowers rose 82%, while its outstanding women-SME portfolio rose 121%. Its total SME business grew 46% in borrower numbers and 71% in outstanding portfolio. Women’s non-performing-loan rate was 2.45%, below the bank-wide rate.

Those figures are portfolio comparisons, not a controlled experiment, and they predate Lebanon’s banking crisis. They cannot prove that each measure caused each result. Still, the pattern is hard to ignore. If women had mainly wanted a discount, changes to collateral and treatment would not have mattered so much.

They did.

For an operator, this changes the question to ask a lender. Do not stop at “Do you have a women’s product?” Ask whether inventory, sales records, vehicles or customer payments can support the application. Ask whether a spouse’s signature is required. Ask when repayment begins.

A loan is not useful because it is advertised to women. It is useful because its rules fit the business.

Ecuador raised the money after changing the rules

Banco Pichincha’s story is often told backwards. The headline is its $100 million gender bond, issued on March 24, 2022 and split equally between IFC and IDB Invest for investment and working-capital lending.

But the operating changes came first.

The Inter-American Development Bank’s 2012 announcement recorded that $12.8 million from a larger financing package was allocated to women-owned SMEs at Banco Pichincha. IDB Invest’s 2021 reporting, rather than the later bond announcement, records the relevant underwriting changes: applicants failing conventional minimum requirements could be assessed through psychometric testing, and products were introduced without a guarantee or spouse’s signature. IDB Invest said the women-MSME portfolio grew faster than Banco Pichincha’s overall MSME portfolio in every year of the operation.

Banco Pichincha changed rejection criteria before scaling wholesale capital, though borrower graduation remains the unanswered test. Photo: EEIM / Wikimedia Commons, CC BY-SA 4.0.

That platform later absorbed larger wholesale funding. Banco Pichincha’s 2023 annual report recorded $821.6 million outstanding across 101,551 loans to women-led microenterprises.

At first glance, that is the desired result: scale, a major bond and a large borrower base. Here is what the numbers do not tell you. The portfolio remains described as microenterprise lending. Did successful clients graduate into larger SME loans, equipment finance or contract-backed facilities? The published figures in this brief do not answer.

Put Ethiopia, Lebanon and Ecuador side by side and a pattern appears that none of the reports says outright: outreach gets borrowers to the door, but underwriting redesign decides whether the door opens wider over time.

In Türkiye, the blockage sat inside the bank

The first EBRD Türkiye Women in Business programme supplied €300 million through five banks, combining credit lines, technical assistance and first-loss risk cover. First-loss cover is a guarantee that absorbs an agreed initial share of losses if borrowers do not repay.

Women new to participating banks represented 42.2% of sub-borrowers. Average non-performing-loan ratios were lower for programme loans, while first-time lending was strongest among banks using first-loss protection. EBRD and Turkish central-bank analysis found that participating banks increased women’s portfolio share relative to non-participants.

The programme later expanded to €600 million. On June 11, 2025, the EBRD approved a loan of up to $70 million to Akbank, a Turkish bank, with targets for new borrowers and lending beyond major cities.

Why did the guarantee matter? The EBRD working paper study, *Gender Discrimination*, tested 334 Turkish loan officers. It found they were 30% more likely to require a guarantor when an otherwise similar application was presented as female-owned.

That is not an advertising problem. It is a decision problem at the moment credit is approved or denied.

A guarantee can alter the lender’s risk calculation while it learns whether its assumptions are wrong. Yet guarantees are not magic and they are not free. Someone bears their cost. The serious question is whether lower losses, wider lending and a growing portfolio produce enough return to justify that cost.

India found the barrier before scoring began

Aye Finance, an Indian lender to small businesses, built cluster-based underwriting using business-sector patterns, customer data and more than 500 additional data points. By early 2024, it had applied the method across more than 180 clusters and disbursed more than ₹100 billion, about $1.2 billion, to more than 800,000 customers lacking conventional records or collateral.

You might expect that to level the field.

It did not. A World Bank document on Aye Finance and CGAP, the financial-inclusion organisation that worked with the lender, reported that only 5.8% of SwitchPe inventory-finance users were women. In response, Aye and CGAP redesigned the customer journey: longer application lead times, visits around care obligations, fewer male-officer visits and gender-sensitisation training.

This is the corrective to a fashionable belief. Better data does not automatically make finance fairer. A model cannot assess a woman who never reaches it because of documentation, mobility, digital-access or household-decision constraints.

In India, the break appeared before assessment. In Türkiye, it appeared in the loan officer’s judgment. In Ethiopia and Ecuador, it appeared in what counted as evidence. In Lebanon, it appeared in collateral and treatment.

Different countries. Same bottleneck.

What investors should measure, not merely celebrate

Borrower counts are easy to present. Bond announcements are easy to celebrate. The less photogenic work is determining whether the lending model makes commercial sense.

For investors, banks and development-finance institutions, the essential measures are portfolio profitability, losses, average loan size, term, repeat borrowing and the cost of any guarantee. They should then examine risk-adjusted returns, meaning the return remaining after allowing for the risk of loans failing and the price of protection against that failure.

A low non-performing-loan ratio is encouraging. It is not the entire investment case. A lender can report healthy repayment while making loans too small to cover operating costs, relying on expensive support, or failing to retain borrowers as they grow. Equally, a first-loss guarantee can be rational if it unlocks a profitable lending segment that a bank had wrongly avoided.

GI Network’s view: Count the businesses that graduate, but also test whether the lender earns enough after credit losses and guarantee costs to keep funding that graduation.

Experienced capital providers should map where women leave the process. Are they screened out before applying? Required to provide guarantors more often? Approved for lower amounts despite comparable repayment? Does a guarantee change officer behaviour? Do repayment dates clash with inventory or customer-payment cycles?

That is a better diligence process than asking whether a facility carries a gender label. It is also relevant to the problem explored in why lenders can love your customers and still reject your business: a sound commercial opportunity can fail if it does not fit a lender’s internal rules.

The work before a new facility

GI Network would map the precise break in a women-SME lender’s capital pathway: customer origination, appraisal, collateral, term design, approval authority or graduation into larger facilities. We would test whether portfolio cash flows can support the proposed product, model losses and guarantee costs, align reporting with capital-provider requirements, and rehearse the investment-committee objections that a gender label cannot answer.

The aim is not simply to raise a larger funding pool. It is to establish whether the lender can deploy capital under rules that create viable, repayable and commercially sustainable growth loans. As with any capital raise, the deal often dies in the investment committee memo, not in the opening presentation.

Use the STEP test

The lesson from these five countries can be carried in four words: Size, Time, Evidence, Progression.

Size: Is the amount sufficient for the actual opportunity, not merely enough to create another loan account?

Time: Does the repayment schedule fit the inventory, equipment or customer-payment cycle?

Evidence: Can the lender recognise cash flow, stock, vehicles, transaction records, receivables, guarantees or other credible signs of repayment capacity?

Progression: After successful repayment, is there a clear route into larger and more formal capital?

Then add the investor’s final check: can this model earn its way without permanently depending on subsidised risk protection?

If a women’s finance programme cannot pass the STEP test, call it outreach, early support or inclusion. Those can all be valuable. But do not call it a growth-capital solution yet.