A development finance institution that wants to reach small and medium-sized enterprises in Central Africa almost never reaches them directly. It works through a partner bank or microfinance institution, to which it extends a credit line, a portfolio guarantee, or both.
The architecture is proven. The ARIZ guarantee covers up to 50 % of a loan or portfolio extended to small and medium-sized enterprises, and up to 75 % for microfinance institutions; the EURIZ guarantee covers 50 to 70 % of facilities granted to micro, small and medium-sized enterprises. The Choose Africa initiative has committed 3.5 billion euros to African start-ups and small and medium-sized enterprises. In 2026 alone, transactions of this type were signed with NSIA Banque Bénin, FBNBank Senegal and Ecobank Senegal.
So the architecture is in place, the capital is available, and the credit risk is largely covered. One question remains that the structure does not address, and it is the one that decides the outcome: once the line is signed, who produces the files?
The risk your guarantee does not cover
On this channel, credit risk is not your principal risk. Your guarantee absorbs it in whole or in part, and your partner bank carries the remainder. Your principal risk lies elsewhere, and nothing covers it: it is the risk of non-deployment.
A line that is never drawn is a silent failure. It produces neither loss nor incident; it produces an absence. The capital remains available, the partner remains solvent, the annual report stays clean — and the intended impact does not exist. This kind of failure never surfaces in a risk dashboard, because risk dashboards measure what goes wrong, not what fails to happen.
The evidence on obstacles to SME lending in Africa all points the same way. Asked why they do not lend more to this segment, African banks cite first the lack of bankable projects, at 40 %, and insufficient collateral, at 37 %.
These two answers are not symmetrical, and that is the whole point. Your guarantee addresses the second. It does not address the first.
Why the file is missing, precisely
In the CEMAC zone, four companies out of five operate outside official records, without certified accounts or auditable financial statements.
It is worth measuring what that sentence implies for a credit committee. It does not say these companies are badly run, or unprofitable. Many are neither. It says they are unassessable as they stand: there exists no set of documents from which an analyst can establish repayment capacity by a method defensible before a committee.
A relationship manager at a partner bank then faces a choice no credit line resolves. Either they decline, and the line is not drawn. Or they reconstruct the client's accounts themselves, and spend time their cost base does not support — average processing time for an SME credit file runs into tens of days, often sixty to ninety, and most of it is consumed by documentary back-and-forth.
The consequence is mechanical and it is poorly understood: the partner bank is not rationing your line for lack of appetite. It is rationing it on the cost of appraisal. With a fixed analysis budget, it serves first the files that are cheapest to assess, which are rarely the ones your line targets.
What technical assistance addresses, and what it does not
The standard answer to this observation is the technical assistance component attached to the line. It is useful, and it is mis-calibrated for this particular problem.
Technical assistance strengthens the partner institution: credit team training, redesign of origination procedures, scoring tools, information systems. All of it is necessary, and all of it acts on the bank's capacity to process a file — that is, on what happens once the file exists.
The bottleneck sits upstream. It sits with the company, in the production of the file itself: accounts restated to prevailing standards, a cash plan that holds, documented and enforceable collateral, a financing need consistent with the real operating cycle. No amount of credit-team training produces those documents, because they are not produced by the bank.
Between a credit line and a financeable company there is therefore a link nobody owns: the one that makes the file assessable. Not the development finance institution, which has no operational presence on the ground. Not the partner bank, whose cost base does not support that work. Not the company itself, which has neither the skill nor the resource.
The context has hardened since April
Two recent developments make this missing link more critical than it was a year ago.
On 2 April 2026, the Monetary Policy Committee of the BEAC suspended refinancing operations for medium-term credit intended to support productive investment. The Governor confirmed on 29 June that the suspension answered a request from the International Monetary Fund, whose position is that the mechanism should be phased out. Banks in the zone therefore lose one of the few arrangements that allowed them to match medium-term lending with a corresponding resource — at the precise moment they are expected to deploy medium-term lines.
COBAC regulation R-2025/02, in force since 1 January 2026, separately raises the minimum share capital of banks from 10 to 25 billion CFA francs, with a first step at 14 billion on 31 December 2026. An institution absorbed by its own recapitalisation does not allocate additional analytical resource to a segment that is expensive to assess.
In short: the constraint on deploying your lines is not easing, it is tightening.
What makes a line deployable
An intermediated channel produces impact when four conditions hold at once. The first two belong to the financial structure, and they are generally well handled. The last two belong to execution, and they are rarely equipped.
- Risk is credibly shared with the bank. That is what the guarantee does, and it is done.
- The resource is available at the right maturity. That is what the line does, and it is done.
- A flow of identified, pre-qualified companies exists, matching the eligibility criteria — sector, size, ticket, use of funds. This is almost never organised: it is left to the partner bank's branch network, whose business is not to prospect a segment it cannot assess.
- Each file reaches committee in a state that permits a decision. This is the deciding point, and it is the one nobody formally owns.
Conditions 3 and 4 are handled on the ground, in local language and local law, company by company. They are handled neither from a European head office nor from a dashboard.
The question to put to your next portfolio review
It is simple, and your drawdown rate already answers part of it:
Across our active intermediated lines in the CEMAC zone, what share of the committed amount has actually been disbursed to the final beneficiary, to date? And for files that did not complete, was the cause a risk refusal — or a file that was never finished?
If most non-completions fall into the second category, then the problem is neither the capital, nor the guarantee, nor the partner. It is the missing link, and it is addressed upstream of the credit committee, not by reinforcing it.
That is also the good news: it is the only one of the four factors that requires no additional funds, no renegotiation and no new partner.
Published 14 August 2026. The guarantee instruments and transactions cited are verified as at that date.
