Sizing the public money a project needs before a lender will touch it
The situation
A water supply project in an emerging market charges too little to repay a loan on its own. Public money has to fill the gap before a lender will look at it.
How much public money is usually set by habit: a share of cost that worked on the last project.
Too much crowds out lenders who would have come in anyway. Too little, and the project stalls at the lender’s door.
This is our own tool, not a client project. We built it because the question comes up on every such project, and the usual answer is a guess.
Why it is hard
The common method picks the grant first, then checks whether a loan works on the rest. That gets the order backwards.
A lender sizes the loan to the cash the project earns, with a cushion on top. An equity investor puts in only what their required return can justify.
Public money should fill whatever is left, and nothing more.
Three things make it harder than it looks:
- A single year in which the project loses money rules out a loan entirely. That is a grant case, not a blended one, and no lever changes it.
- Asking lenders for a thinner cushion adds debt, but takes cash away from equity. The gap closes by far less than the extra debt.
- Real and nominal figures get mixed. A tariff held flat in real terms, discounted at a nominal rate, understates what the project can carry.
The approach
The tool builds the cash flow from what a sponsor already knows. That means connections, the tariff, how much is collected and what it costs to run.
Then it sizes each layer in turn, in the order money actually arrives:
- The loan. Each year’s payment is that year’s cash divided by the lender’s cushion. The loan is what those payments repay.
- The equity. The cash left after the loan, and all the cash once it is repaid, valued at the investor’s required return.
- The gap. The cost, less the loan and the equity. This is the public money the project needs.
- The levers. A longer loan, a cheaper loan, cheaper equity and a higher tariff, each tested on its own against the gap.
It also prices the ways public money can come in. A grant, a junior loan, a loan guarantee and a first-loss layer each close the gap at a different cost.
The tool ranks them by how much private money each one brings in for every unit of public money spent.
Before any terms are agreed, it gives a range across typical terms. A sponsor knows the size of the gap on day one.
A worked example
A synthetic project costs 50 million to build. It serves 40,000 connections at 10 a month, and costs 1.6 million a year to run.
| Step | How | Millions |
|---|---|---|
| Revenue a year | 40,000 connections at 10 a month | 4.8 |
| Running costs a year | flat | 1.6 |
| Cash to repay a loan | revenue less running costs | 3.2 |
| Loan payment it supports | cash divided by a cushion of 1.3 times | 2.5 |
| Loan | those payments over 18 years at 8 percent | 23.1 |
| Equity | cash left, valued at 15 percent a year over 30 years | 5.9 |
| Private money | loan and equity | 29.0 |
| Public money needed | cost of 50 less private money | 21.0 |
The loan and the equity together raise 29.0 million. That leaves 21.0 million for public money, 42 percent of the cost.
The financing levers barely move it. A loan five years longer, a point cheaper, or equity three points cheaper each close between 1.7 and 2.0 million.
A tariff a fifth higher closes 8.7 million on its own. At 14.83 a month, 1.5 times today’s tariff, no public money is needed at all.
The thinner cushion shows the trap. Cutting it from 1.3 times to 1.2 times adds 1.9 million of loan.
Yet the gap closes by only 0.7 million. The equity investor loses the cash the lender takes.
On this project, the gap is a tariff problem, not a financing problem.
Where it breaks
- It never asks whether people can afford a higher tariff, or whether raising it is politically possible.
- Tax is charged without allowances for depreciation or interest. Early tax is overstated, so the gap comes out a little too large.
- Running costs stay flat while connections build up, and there is one cash flow case with one required return.
- A currency test shows the damage of a devaluation, but it does not re-size the loan to match.
If you are asked how much public money a project needs, start with the tariff, not the loan.