Capital stacks and blueprints
Illustrative educational model — not a real transaction
First-loss guarantee for household lending
An illustrative teaching model, not a real transaction. Assumptions, funding flows and risks are explicitly stated.
- Last reviewed
- Related chapter
- Philanthropy as Amplifier
Label: Illustrative educational model. This is a teaching example that generalises the pattern in the book's Chapter Four water-and-sanitation composite and in real guarantee programmes. The amounts are round numbers chosen for clarity. It does not describe any specific transaction, and no organisation should be named as a party to it.
Assumptions. A philanthropic guarantor; two commercial banks; a loan product (household water connection) with unknown repayment behaviour; a target of USD 50 million in lending over three years; a guarantee covering the first 20 percent of losses.
Problem addressed. Commercial banks will not lend to low-income households for water or sanitation because they have no repayment data and expect high default.
Layers and roles
| Layer | Amount | Provider | Role |
|---|---|---|---|
| Partial credit guarantee | Covers first 20 percent of portfolio losses, up to USD 10 million | Philanthropic guarantor | Absorbs first loss; capital stays on guarantor's balance sheet unless called. |
| Bank lending | USD 50 million over three years | Two commercial banks | Make the loans; carry losses above 20 percent. |
| Technical assistance | Say USD 1 million | Philanthropic grant | Product design, borrower education, data systems. |
| Borrowers | Loans of a few hundred dollars each | Households | Repay from household income. |
Funding flows. No cash moves from the guarantor at signing. Banks lend. TA grant funds product design and data.
Repayment flows. Households repay the banks. Banks recycle repayments into new loans. For covered portfolio losses, the guarantor pays first losses up to the USD 10 million cap; the banks carry losses above that cap.
Risk allocation. Guarantor: first 20 percent of losses. Banks: everything above. In the illustrative case, delinquency stays below 2 percent, the guarantee is never called, and the capital remains invested in the guarantor's endowment throughout.
What the philanthropic contribution enables. Fifty million of lending from a guarantee that is never drawn, plus the repayment data that lets the banks continue without a guarantee in year four.
Governance and accountability. The guarantee agreement should specify: the trigger (portfolio-level or loan-level losses), the cap, the term, reporting frequency, the data the banks must share, and the conditions under which the guarantee steps down as data accumulates.
Conditions under which the structure may not work. Banks lend carelessly because the guarantor carries first loss (moral hazard; mitigated by keeping the guarantee partial and portfolio-level). The product is one the banks would have offered anyway (no additionality). A macro shock pushes losses past 20 percent and the banks withdraw. The guarantee is never stepped down and becomes a permanent subsidy.
Why this is illustrative. Real household-lending guarantees vary widely in cap, trigger, and term. A practitioner designing one should start from the questions in the vehicle guide and from the documented structures (MedAccess, case 3, for how a guarantor accounts for exposure), not from these round numbers.
Sources
Educational material, not investment, legal or tax advice. Reported figures and review dates reflect the cited sources and editorial review; they are not live data.
