Debt Service Reserve Account: a Worked Example

Project finance, infrastructure and single-asset property loans are repaid from one asset's cash and nothing else. There is no wider business to carry a bad quarter, and the sponsors are usually not obliged to add money. That cash is also uneven: a hydro scheme depends on rainfall, a wind farm on wind, a building on its tenants, a plant on staying in service. When the cash falls short on a payment date, a missed payment is an event of default after a short grace period, and the parties are negotiating under pressure.

A reserve puts a cushion between a weak period and that default. It is a liquidity tool, not a solvency one: it pays the bank on time while a temporary shortfall corrects itself, and it buys a few months of thinking time when the shortfall is not temporary.

The look-forward period is set deal by deal: the next six months, the next twelve, the next payment, or a fixed sum. Because it is recalculated each period, the required balance falls as the debt amortises, and the excess is released. Where repayments are sculpted, debt service changes from period to period and the reserve moves with it.

The money can come from three places. It can be paid in at financial close, as part of the amount borrowed or of the sponsors' contribution. It can be built up from the first months of operating cash, which leaves the lenders unprotected until it is full. Or the agreement may accept a standby letter of credit or guarantee from a bank meeting a stated rating, which the lenders draw on instead of the cash. Facility agreements commonly require a letter of credit to be replaced or cash-funded if its issuer is downgraded or the letter nears expiry. Cash in the account usually earns less than the debt costs, and that gap, the negative carry, is why sponsors prefer a letter of credit.

A run-of-river hydro scheme, constructed for this page like every figure in it, has borrowed US$36m repaid in level half-yearly payments of US$2.70m. Its facility agreement requires a reserve equal to the next six months of debt service, funded from the loan at financial close. Cash in the account earns 3.5%; the loan costs 7%. Distributions to the sponsors are allowed only while twelve-month debt service cover is at least 1.20 times. The rates and the 1.20 times belong to the example and are not offered as market norms.

The bank was paid in full both times and no default arose. But the lenders now know something the payment record does not show: in a dry season the scheme's own cash met only 70% of what it owed. The US$0.10m left after the refill stays in the project, because twelve-month cover of 1.02 times is below the 1.20 times the distribution test requires.

The reserve's real measure is time. At a shortfall of US$0.80m a half-year, a full reserve covers three such periods, about eighteen months. If the turbines stopped altogether, it would cover one payment. That is the question a lender asks of any reserve: how long would it last in the shortfall this asset is likely to have?

The two are confused because both are reserves inside a project loan. They belong to different stages and work in opposite ways. An interest reserve belongs to the build. It is a line in the development budget that meets the loan's interest before there is any income, normally added to the balance as it goes, and the budget assumes every dollar of it is spent. A debt service reserve belongs to the years after completion. It is real cash, sitting against future repayments, and the plan is that nobody touches it. Drawing an interest reserve is the plan. Drawing a debt service reserve is news.

Nor, in many agreements, does the reserve change the cover ratio: cash available for debt service is measured without counting withdrawals from the reserve, so a draw cannot make a weak period look like a strong one. The ratio reports the asset. The reserve reports how long the lenders can wait for it.

DEFINED IN ONE PARAGRAPH → · CASH FLOW AVAILABLE FOR DEBT SERVICE → · THE LOAN LIFE COVER RATIO →

The reserve is argued twice. At the term sheet, sponsors push for a shorter look-forward period, a letter of credit in place of cash, or a reserve built from early cash flow. Lenders weigh how volatile the asset is and how quickly a shortfall could be fixed. The Basel Committee's slotting criteria for grading project finance loans make the same distinction: a longer than average coverage period, with every reserve fully funded in cash or by letters of credit from highly rated banks, counts towards a strong grade, while a shorter period with reserves funded from operating cash flows points to a weak one.

Once the asset is operating, a draw is reported to the lenders, usually stops distributions until the account is full again, and opens the conversation the reserve was bought for. Is this a dry year, a repairable outage, a tenant replaced? Or is it a permanent fall in what the asset earns? A reserve that is drawn and refilled every few periods can hide a slow decline for years, which is why a lender reads the loan life cover ratio beside it. Near maturity the reserve is released, and agreements often let it meet the final payments.

DISTRIBUTION LOCK-UP → · DEBT SERVICE COVERAGE RATIO →