Key Commercial Loan Terms and Definitions

Knowing what DSCR, LTV, debt yield, and a covenant actually mean is table stakes in commercial real estate finance. What separates a competent lender from a vulnerable one is understanding that each of these terms names a test the borrower has to keep passing for as long as the loan exists, not a box checked once at closing. Every quarter after that is a fresh exam, and the borrower doesn't get to skip it just because they aced the first one.

That reframing matters more in 2026 than it has in years, because the calendar is doing something unkind to CRE lenders. Nearly $875B in CRE loans mature in 2026, with hundreds of billions more in the years that follow, and that's happening while borrowing costs stay elevated and lenders are now cherry-picking only the most disciplined borrowers to refinance. A lender facing that kind of maturity wall doesn't have the luxury of treating DSCR as a number on an origination memo. It has to treat DSCR, LTV, debt yield, and every covenant in between as live signals that need watching for the full life of the loan, because the borrowers coming back to the table this cycle are being sorted by how disciplined their numbers have stayed.

Read together, the terms stop looking like vocabulary and start looking like what they are: a system built to catch trouble before it becomes a missed payment.

DSCR: the primary lender covenant metric and its manipulability

Debt service coverage ratio gets top billing in nearly every CRE loan covenant package, and for good reason: it answers the single most direct question a lender can ask, which is whether the property's income covers its debt payments with room to spare. But knowing the formula tells you almost nothing about where DSCR can go wrong, and a ratio can be technically correct while still hiding a weakening property.

Start with how the number gets built. Both halves of the fraction, the income side and the debt side, can move for reasons that have nothing to do with how the property is actually performing. An underwriter working through a stack of financials by hand can miss those anomalies, where a system built specifically to flag that kind of pattern catches it automatically. On the other side of the ratio, debt service isn't fixed either: it moves with the interest rate and amortization schedule attached to the loan, so a borrower can push DSCR higher by refinancing or restructuring terms without the property generating one more dollar of income. Lenders have started layering a debt yield covenant on top of DSCR because debt yield doesn't care what the interest rate is.

A timing problem sits underneath the DSCR calculation: it usually gets calculated off trailing twelve-month financials, and T-12 numbers can swing hard on a single event. A large unexpected repair, a tenant rolling down to a lower rent, a jump in insurance premiums: any one of these can drag T-12 NOI low enough to push DSCR under the covenant threshold well before the next quarterly test date actually fires. The property's actual financial condition doesn't wait for the calendar.

DSCR behaves like a trend, not a switch, so a ratio declining steadily over three straight quarters is telling a story well before it crosses whatever line the loan agreement draws, even though most covenant testing treats it like a switch anyway. Watching a risk develop, instead of finding out about it only after the threshold's already been crossed, is the whole difference between a monitoring program that works and one that's just filing paperwork after the fact, as CRE risk management frameworks describe the problem.

LTV, debt yield, and NOI: what each metric measures that the others cannot

DSCR tells a lender whether cash flow covers debt payments today. It has nothing to say about what the collateral is worth, and it can be gamed by restructuring debt terms rather than improving the property. That's the gap LTV and debt yield exist to close, and net operating income is the number both of them depend on.

Loan-to-value divides the loan balance by the property's current appraised or marked value, and unlike a one-time underwriting metric, LTV gets monitored for the life of the loan, not just at closing. That ongoing tracking exists because of a blind spot DSCR simply can't cover: a property's value can fall sharply from cap rate expansion or a repricing submarket, with the loan balance and the borrower's payment history completely untouched. Agency multifamily programs typically cap stabilized CRE loans at up to 80% LTV.

Debt yield solves a different problem: it divides NOI by the loan amount, and by design, it ignores both the interest rate and the amortization schedule entirely. Because debt yield can't be improved by refinancing or stretching out amortization, it forces the loan to stand or fall on property-level income relative to principal, which is a cleaner read on collateral risk than a ratio that moves with financing terms. Debt yield is harder to manipulate than DSCR, and that difficulty is the entire reason lenders bother requiring it.

NOI sits underneath both ratios and produces the real work done by each. NOI feeds directly into both DSCR and debt yield, so if NOI is calculated inconsistently, every number built on top of it inherits that inconsistency. Getting NOI calculated the same way, loan after loan, analyst after analyst, is the foundation the whole covenant stack rests on. A lender can have airtight DSCR and debt yield covenants written into every loan agreement in the portfolio, and still be exposed if the NOI feeding those formulas isn't built the same way twice.

The full covenant stack: financial, operational, and reporting obligations lenders must track

DSCR, LTV, debt yield, and NOI get most of the attention because they're numeric and they show up on every term sheet, but a covenant is a broader thing than a financial ratio. A covenant is any contractual obligation the borrower has to satisfy for the life of the loan, and a missed reporting deadline carries exactly the same legal weight as a DSCR breach: both are technical defaults, and both can trigger the same consequences. It means a borrower who's never come close to tripping a financial covenant can still land in default over a late rent roll.

The covenant stack breaks into three categories, and each demands a different way of watching it. Those obligations can't be checked just by reading a borrower's submitted financials. They require looking at property-level operating data and, in some cases, third-party sources to confirm what's actually happening on the ground, since an operator has every incentive to describe occupancy more generously than reality supports.

Reporting covenants round out the stack, and they come with hard deadlines: rent rolls, trailing twelve-month statements, audited financial statements, insurance certificates, construction progress reports, all due on a schedule spelled out in the loan agreement. If a borrower misses one of those deadlines, the loan is technically in default even if the property is performing beautifully and every payment has landed on time. That's the part of the covenant stack most likely to get treated as paperwork rather than risk, and it's exactly where that assumption causes problems.

Covenants also split along a different axis: affirmative covenants require the borrower to do something (deliver a quarterly statement, keep insurance in force), while negative covenants restrict what the borrower can do without the lender's sign-off (take on more debt, sell the asset, change who owns it). A lender has to actually detect that a triggering event happened, often through borrower disclosure or third-party data, rather than simply checking a box that a form arrived on time.

Cash management triggers sit on top of all of this as a kind of early consequence layer. Springing cash sweeps and lockbox provisions activate once DSCR or debt yield drops below a defined level, usually below the formal covenant threshold itself, and often right around the same time as the formal breach. A borrower can be drifting toward a cash sweep trigger without having technically breached anything yet, so watching the trend line, not just the threshold, catches a problem early instead of discovering it after the lockbox has already activated.

Where covenant monitoring breaks down

For a lot of portfolios, the covenants are usually fine. What fails is the system that's supposed to track them, and that failure traces back to three specific, repeatable structural causes, not to careless staff or unusual borrowers.

The first is a data lag built into DSCR itself. Trailing twelve-month financials often arrive weeks after the period they're supposed to cover, and when they arrive late, or arrive but never get pulled into the system that actually runs the calculation, the DSCR number on file simply doesn't update. The lender definitely doesn't.

The second failure mode is inconsistency. When DSCR gets calculated by hand across a portfolio, different analysts apply different interpretations of which income lines count, how to normalize unusual expenses, and how to treat rent concessions, so the same financial statement can produce different DSCR numbers depending on who ran it, creating both credit risk and regulatory exposure when examiners review the same loan file. That's not just a credit risk. It becomes a regulatory problem the moment an examiner pulls the same loan file and asks why last quarter's number doesn't match this quarter's methodology.

The third failure is routing: an alert that fires correctly but lands in the wrong inbox, or sits unread in a shared mailbox nobody actually checks, has functionally never fired at all. That kind of failure doesn't scale gently. At a portfolio of 100 loans or more, every additional loan is another chance for an alert to get misrouted, and the odds compound rather than stay flat.

Firms processing loan volume at real scale, including institutional lenders like KeyBank and asset managers like JLL, operate at a size where that kind of manual leak is a mathematical certainty if the monitoring layer doesn't scale with the portfolio.

What gets missed in all of this is the most useful intervention point in the entire covenant lifecycle: the pre-watch list. A pre-watch list flags loans for closer attention before any formal covenant trigger fires, based on early warning signals surfacing at the tenant, property, or market level. A borrower sitting on a pre-watch list still has real options on the table: sourcing new equity, renegotiating a lease, getting ahead of a problem proactively. By the time a loan reaches the formal watch list, those options have mostly narrowed down to workout paths that cost both sides more. Risk management frameworks built around CRE lending point to the pre-watch list as the stage where the most productive interventions actually happen, and building one depends on continuous monitoring, since the signals that populate it tend to appear two to six quarters before the formal trigger ever fires. If the lender misses that window, it is left reacting to a default instead of heading one off.

What regulators expect lenders to demonstrate about covenant monitoring

None of this stays a private risk-management preference once an examiner walks in the door. Bank examiners don't stop at confirming that covenants exist in the loan agreement. They look at whether covenant testing runs on a system or gets handled case by case, whether exceptions get logged and resolved rather than quietly forgotten, and whether the lender can produce documentation on demand instead of needing days to reconstruct it from scattered files.

That standard has a concrete shape. A monitoring program built to hold up under examination needs to produce a timestamped audit trail for every single covenant test run, one that ties each DSCR calculation back to the specific trailing twelve-month line items that drove the result. It needs to produce that documentation in exportable form, without someone having to manually reconstruct the math after the fact, which is the bar the OCC and FDIC hold lenders to and the bar a spreadsheet-based process structurally cannot clear once a portfolio reaches real scale. A spreadsheet can store a number. It can't reliably reproduce, months later, how that number got built, which is the thing an examiner is going to ask for.

The regulatory bar isn't static, and 2026 is raising it. The OCC's Fall 2025 Semiannual Risk Perspective called out weakening credit performance in certain CRE segments, with multifamily singled out specifically. A lender that can't show, loan by loan, how each covenant test was run and what it found is carrying an examination finding waiting to happen, putting covenant monitoring in the same category as capital adequacy or loan loss reserves: a documented, examined obligation a lender has to be able to prove on request.