Debt sizing is the discipline of turning property cash flow, collateral value, and loan terms into a debt amount the asset can carry through a reasonable downside. Coverage ratios are the checks on that decision – simple fractions that tell you when cash flow stops covering the lender’s required payments and reserves.
Debt sizing in real estate private equity (REPE) models is frequently reduced to a single “max LTV” or “min DSCR” rule. That’s neat, but it’s not how loans behave in real life. The binding constraint moves as the business plan moves, and the documents decide what “coverage” even means.
The question an investment committee should ask is not, “How much leverage can we get?” It’s, “What leverage should we assume will stay financeable when the easy assumptions fail, while we still keep the steering wheel?” Get that wrong and you don’t merely give up return. You give up choices, usually when choices are most valuable.
Definitions, scope, and boundary conditions that make sizing real
Debt sizing is the lender- and sponsor-driven determination of (i) maximum initial principal, (ii) future draw capacity, if any, (iii) amortization and maturity profile, and (iv) the liquidity structure – reserves, cash management, and traps. Sizing is constrained by underwriting metrics and by the loan documents: cash sweeps, cure rights, transfer restrictions, and lender consent rights.
Coverage ratios measure the relationship between a cash flow proxy and a financing obligation proxy. The common set is familiar, but the definitions vary enough to matter.
- DSCR: Debt Service Coverage Ratio is NOI divided by scheduled debt service for a period. In U.S. real estate, DSCR may use “stabilized” NOI for term loans and in-place NOI for acquisition or bridge loans. What matters is what the agreement calls NOI and what it includes in “debt service.”
- ICR: Interest Coverage Ratio is NOI or EBITDA divided by interest expense only. It’s most relevant when loans are interest-only and repayment depends on sale or refinance.
- Debt yield: Debt yield is NOI divided by loan balance. It is not the lender’s yield. It’s a leverage-invariant cushion metric that ties directly to collateral value under a cap-rate framework.
- LTV: Loan-to-Value is loan balance divided by appraised value or “lender value.” LTV is only as good as the appraisal definition and the cap-rate assumptions underneath it.
- LTC: Loan-to-Cost is loan balance divided by total project cost, with “cost” defined by negotiation and often haircut in underwriting.
Debt sizing is not a math contest to maximize IRR. A sponsor can usually lever more by paying a higher all-in rate, accepting tighter covenants, or giving away control. Instead, size to the weakest link in the plan, which is often liquidity and timing rather than long-run value.
Boundary conditions also bite. A model can show an acceptable DSCR and still be unfinanceable because of sponsor concentration limits, property type restrictions, tenant rollover, market blacklists, or title and environmental exceptions. Lender credit committees do not underwrite your base case. They underwrite the story that goes wrong and ask whether the loan survives it.
Incentives are predictable. Senior lenders like early-warning metrics that are hard to massage, such as debt yield or springing cash management. Sponsors like forward-looking definitions, including DSCR on stabilized NOI, flexibility to spend on leasing, and cure rights. Mezzanine and preferred equity care most about enforcement timing, intercreditor standstills, and who controls the collateral when the plan slips.
Where coverage ratios affect the capital stack, not just proceeds
Coverage ratios sit under three linked decisions: quantum, shape, and control. You can think of these as the practical outputs of a capital stack discussion.
- Quantum: How much debt the asset can support under the lender’s version of NOI, rates, and haircuts.
- Shape: Interest-only versus amortizing, fixed versus floating, and what hedging is required.
- Control: The operational trade-offs, including lockbox, cash sweeps, leasing consents, transfer limits, and remedies.
Two loans can show the same leverage and pricing and still behave very differently. One may have hard sweeps that trap cash at the first stumble. Another may allow reserve builds and sponsor cures that keep the property operating. Coverage ratios are the language used to negotiate those differences, and the cash waterfall is where they show up.
How debt sizing changes by product: senior, mezz, pref, and construction
Senior loans: the constraint is usually a stack of metrics
Senior mortgage loans are usually constrained by a stack of LTV, DSCR, and debt yield. In volatile rate periods, floating-rate senior often binds on stressed DSCR and the borrower’s ability to buy and maintain an interest rate cap. The cap is not a nice-to-have. It is often the cost of admission.
Mezzanine: paperwork often beats “coverage”
Mezzanine debt sizes to combined loan-to-value (CLTV) and to the attachment point cushion under a stressed cap rate, meaning how much value is left after paying off the senior. Mezz “coverage” tests exist, but the practical constraint is intercreditor mechanics: payment blockage, standstill length, purchase options, and cure rights. When trouble comes, paperwork decides who moves first. For a deeper overview of this layer, see mezzanine financing.
Preferred equity: model cash after senior and reserves
Preferred equity can look like mezz economically but often carries more flexible current-pay terms, including PIK toggles. The relevant coverage becomes cash after senior obligations and required reserves, plus the sponsor’s ability to defer distributions without triggering control shifts.
Construction: liquidity, completion support, and peak balance
Construction loans size primarily to LTC, budget contingency, sponsor net worth and liquidity tests, and completion support. Coverage ratios usually appear later as conversion tests, reserve release tests, or conditions to refinance into a term loan.
Across all of these, the modeling job is simple to say and hard to do: match your ratio calculations to the definitions in the documents. A generic DSCR built off a model NOI line can be comforting and useless at the same time.
From model NOI to lender-defined cash flow (the line item that breaks models)
REPE models typically compute NOI as revenue minus operating expenses, before capex, leasing costs, and debt service. Lenders often run their own version of NOI. Common adjustments include minimum vacancy and credit loss even if the rent roll is full, market rent and downtime assumptions on rollover, TI/LC haircuts, exclusion of non-recurring income, and standardized management fees.
On the obligation side, “debt service” in covenant land may include more than interest and scheduled principal. Many agreements treat required reserves as part of the debt service burden for covenant testing, including taxes, insurance, replacement reserves, and tenant-related reserves. If DSCR is tested after reserves, the sponsor’s distributable cash can be far lower than the model’s free cash flow. That reduces flexibility at exactly the wrong time.
A workable model carries at least three cash flow lines: (i) property NOI, (ii) lender-defined net cash flow for covenants, and (iii) distributable cash after required reserves and sweep mechanics. If you only carry one, you’re choosing to be surprised later.
What the core ratios really measure (and what they hide)
DSCR is intuitive because it asks whether the property can pay the loan this period. However, DSCR carries definition risk and timing risk. Definition risk comes from what counts as NOI and what counts as debt service. Timing risk comes from the way DSCR improves mechanically with interest-only periods or delayed amortization, even while refinance risk piles up at maturity.
In a model, DSCR should be tested under the projected rate path, a lender-style stressed rate, and a refinance or exit-rate assumption at maturity. For floating-rate loans, DSCR is often more sensitive to rates than to NOI. Put that sensitivity on the page where the committee can see it, and consider pairing it with a stress testing view that isolates the driver.
ICR isolates rate burden by focusing on interest only. That’s useful in bridge and transitional deals where principal is expected to be taken out by refinance or sale. The catch is that ICR can look fine while liquidity is tight because reserves, capex, and rollover costs absorb the cash that keeps the plan on schedule.
Debt yield is blunt and hard to game. It ignores interest rates and asks, “How much cash flow sits on top of the lender’s principal?” Lenders like it as a backstop when DSCR could be satisfied by moving to interest-only or assuming friendlier rates. Debt yield also links to value: if value equals NOI divided by a cap rate, then debt yield approximates the cap rate at which the lender would be at full value coverage. When uncertainty rises, lenders raise required debt yields to compensate.
LTV depends on value, and value is an estimate. Appraisals can move with sentiment, which makes LTV pro-cyclical. The document definition matters: “as-is” versus “as-stabilized,” one appraisal versus the lower of two, and whether the lender can refresh value tests after closing. Those choices drive cash sweep triggers and mandatory paydowns, not just initial proceeds.
LTC is central in construction because cost overruns hurt lenders in a one-way fashion. “Total cost” is negotiated. Land basis, capitalized interest, developer fees, contingency, and soft costs may be included or excluded. In stress, lenders often haircut soft costs and fees and focus on what can be sold and what must be paid.
How lenders actually size: the constraint stack approach
Term sheets show one leverage point. Credit memos usually run a constraint stack and take the lowest result. For a stabilized senior loan, the stack often includes maximum LTV on lender value, minimum DSCR on underwritten NOI and stressed rates, minimum debt yield on underwritten NOI, sponsor and deal concentration limits, and cash management or reserve requirements that reduce net cash flow. If your model only backs into leverage from one metric, it will miss the true binding constraint and misstate close certainty.
Rate risk and hedging: model extensions like a separate loan
Floating-rate debt needs a stressed coverage view, not a single DSCR number. Lenders often require an interest rate cap, and the model should show the strike, the term, the premium, and whether the cap survives the full extension period.
Here is a useful rule of thumb that is often missed in REPE models: treat an extension option as a new underwriting event with new hedging economics. In practice, cap replacement cost can become a real liquidity event right when NOI is still stabilizing and lenders are least forgiving.
Run three cases and put them side by side: a current-rate case (forward curve or house view), an underwriting stress case (lender stress rate plus spread), and an extension case (loan extends and the cap is replaced at then-prevailing pricing). Compute coverage under each and map it to covenants and refinance proceeds. The point is not precision. It’s to surface the terms that create forced actions.
Covenants: ratios only matter because they trigger outcomes
Coverage ratios matter because they activate lender rights. The same DSCR shortfall can be a mild nuisance or a full cash lock, depending on the waterfall and cure mechanics.
- Cash sweeps: Excess cash is used to pay down principal instead of being distributed or reinvested.
- Springing lockbox: The lender takes control of cash receipts once a trigger is hit.
- Reserve top-ups: The borrower must fund reserves, which can reduce covenant cash flow and create a feedback loop.
- Mandatory paydowns: LTV breaches can force paydowns even if payments are current.
- Control rights: Leasing, capex, distributions, and transfers can require tighter lender consent.
In REPE, the pain often arrives before any payment default. Cash traps reduce distributions, strain sponsor liquidity, and can force equity raises at unattractive moments. A model that treats covenant breach as a binary “default” misses the slow squeeze that erodes flexibility quarter by quarter.
Construction and transitional assets: liquidity and timing run the show
During construction or heavy value-add, NOI is weak or negative, so lenders focus on LTC, as-complete value, interest reserve sizing, sponsor completion support, and draw mechanics with inspections and retainage. The modeling hazard is circularity: capitalized interest grows the loan balance, which can breach LTC or as-complete LTV unless the sponsor contributes more equity. If you don’t model peak balance and fee timing, you will understate required equity and overstate runway.
When the project stabilizes, DSCR and debt yield tests often show up as conditions to convert, release reserves, or allow distributions. Stabilization is usually a covenant definition, not a leasing team’s optimism. Treat it like a legal milestone, because it is one.
Waterfalls: “available cash” is not distributable cash
Debt sizing and coverage live inside the priority of payments. A typical lender-controlled waterfall takes gross receipts and pays taxes and insurance, required reserves, senior interest, senior principal or swept paydown, then mezz if permitted, then approved reimbursements, and only then equity distributions, usually conditioned on no default and covenant compliance.
That ordering creates feedback loops. A reserve increase can reduce net cash flow, which reduces DSCR, which triggers a sweep, which reduces liquidity further. Model the loop. Then ask a blunt question: can the business plan operate for multiple quarters in a swept state without starving leasing and capex execution? If the answer is no, the leverage is too high or the terms are too tight.
Documents and net proceeds: model the definitions, not the headlines
Coverage ratios are legal definitions. The minimum document set tells you what to model: the loan agreement for covenants and triggers, the note for payment terms, the mortgage and assignment of rents for remedies and cash flow control, the cash management agreement for lockbox mechanics, guaranties for completion and carveouts, intercreditor for mezz dynamics, and hedging documents for cap economics and collateral terms. Estoppels and SNDAs matter because they affect rent enforceability and lender rights against tenants.
Fees and “net proceeds” are the quiet leverage. Size to net proceeds and peak balance, not headline principal. Upfront fees reduce cash available and raise effective leverage because equity must cover more of the purchase or budget. Ongoing fees can be treated as operating expenses or included in covenant debt service calculations. Exit fees and prepayment penalties affect refinance and sale options. Hedging premiums can be paid upfront or financed. Financing them increases balance and can tighten LTC or LTV.
Practical kill tests and what committees should demand
A sponsor should be able to reject a leverage level quickly with a few screens. The fastest screens are designed to force clarity on timing, control, and refinance survivability.
- NOI drawdown: Can the deal handle a 10-20% NOI shortfall for four quarters without triggering sweeps that halt leasing and capex execution?
- Refinance reset: What happens if the refinance market demands a higher debt yield or lower LTV than your exit assumes?
- Extension liquidity: Does the rate cap cover the full extended term, and if not, do you have liquidity to replace it?
- Rollover alignment: Are reserves and TI/LC funding aligned with rollover timing?
- Value refresh risk: Can the lender refresh appraisals at will, and can you fund a mandatory paydown if value drops?
- Cure realism: Do you have a cure path that is rational and funded, not wishful?
If any answer is weak, lower leverage, change the structure, or reprice the equity. Hoping is not a strategy.
A decision-ready debt sizing package shows a sizing grid under LTV, DSCR, and debt yield using lender-defined cash flow and stress rates; a monthly liquidity bridge through stabilization including reserves, TI/LC, capex, and interest reserve mechanics; a covenant dashboard showing headroom and the cash impact of each trigger; a conservative refinance proceeds analysis including extension fees, replacement hedging, and transaction costs; and sensitivities that isolate the real drivers: NOI, cap rate, interest rate, and timing delays. If you want the sensitivities to read cleanly, borrow the discipline of separating inputs and drivers used in sensitivity analysis.
Closing Thoughts
Debt sizing is not a single number. It is a set of constraints you agree to live with. Coverage ratios are the interface between the property and the documents, and the documents decide who controls the cash when performance slips.