Sizing the Deal: Loan-to-Value, Coverage, and Covenants

Every data center loan is sized by two ratios, and if you learn only two numbers in infrastructure debt, learn these: loan-to-value and debt-service-coverage. The first controls how much you can borrow against what the asset is worth. The second controls how comfortably the asset's cash flow can pay the debt. Together they set the size of the loan, and everything else in the term sheet is commentary.

Loan-to-value is the ratio of the loan amount to the appraised value of the asset. A sixty-percent loan-to-value on a five-hundred-million-dollar campus means three hundred million of debt and two hundred million of equity underneath it. That equity is the lender's cushion: if the asset has to be sold in distress, value can fall by the size of the equity layer before the lender takes a loss. Lower loan-to-value means a thicker cushion and a safer loan, so lenders reward it with a lower interest rate. Higher loan-to-value means thinner protection, so it costs more and eventually becomes unavailable. Stabilized data centers with strong tenants often support loan-to-value in the region of sixty to seventy percent; speculative or single-tenant-concentrated assets get less.

Debt-service-coverage is the ratio of the asset's net operating income to its annual debt payments. A coverage ratio of 1.5 means the building throws off one and a half dollars of cash for every dollar of principal and interest due. Above one, the asset pays its debt with room to spare. At exactly one, every dollar of income goes straight to the lender with no margin for error. Below one, the asset cannot cover its own debt and the sponsor must inject cash or default. Lenders require a minimum coverage, commonly around 1.3 to 1.4 for a well-leased asset, and that minimum, applied to the asset's income, directly caps the loan size. Coverage, not appraised value, is often the true binding constraint, because a lender will not advance more debt than the rent can comfortably service no matter how high the appraisal.

These two ratios interact. In a low-interest-rate world, debt payments are small, so coverage stays high even at a large loan balance, and loan-to-value becomes the binding limit. When rates rise, debt payments swell, coverage tightens, and suddenly coverage becomes the constraint that caps the loan well below the loan-to-value ceiling. This is why rising rates shrink how much debt a given building can carry even when its value has not fallen: the same rent now services less borrowing. Sponsors who sized deals in a cheap-money era and must refinance into a costlier one confront exactly this squeeze.

Covenants are the ongoing rules that keep those ratios from drifting after closing. A coverage covenant requires the borrower to maintain, say, at least 1.25 times coverage tested quarterly; breach it and the lender gains rights, from trapping cash to declaring default. A loan-to-value covenant can force a paydown if the asset's value falls. Cash-sweep provisions divert surplus cash to prepay debt when coverage weakens. Distribution blocks stop the sponsor from taking money out while the asset is stressed. Reserve accounts pre-fund maintenance, taxes, and several months of debt service so a temporary hiccup does not trigger a default. Covenants are how a lender enforces discipline on an asset it does not operate.

Tenant credit sits underneath all of it. The same physical building supports very different debt depending on who signs the lease. A campus leased to a top-tier hyperscaler on a fifteen-year term is a near-bond, and lenders will stretch loan-to-value and accept thinner coverage because the income is so secure. The identical building leased to a weak or unrated tenant, or split among several short-term tenants, is riskier income, so lenders demand lower loan-to-value, higher coverage, and a higher rate. Lease term matters too: debt that matures before the lease does is far safer than debt that outlives its tenant's commitment and must be repaid into an uncertain re-leasing market.

Put the framework together and you can underwrite any deal on a napkin. Take the stabilized net operating income, divide by the required coverage ratio to find the maximum annual debt service, and translate that into a loan amount at the prevailing rate. Check that the result does not exceed the loan-to-value ceiling on the appraised value. The lower of the two answers is your loan. The rest, equity, is what the sponsor must supply. Master that calculation and you understand why some announced projects quietly never get financed: the rent simply will not service the debt the build requires, and no covenant package can fix an asset that does not earn enough.