DSCR, LLCR and PLCR: What Each One Measures, and Why One of Them Is Sometimes Redundant
Here is a result that is not widely remarked on and is worth building a post around.
On a loan sculpted to a constant target DSCR, the loan life coverage ratio equals that target exactly. Not approximately — exactly. On the schedule from the sculpting post, sized at 1.30×, the LLCR at financial close is 1.300000.
That is not a coincidence and it is not a modelling artefact. It follows directly from what both ratios are: if debt service in every period is CFADS divided by 1.30, and the outstanding balance is the present value of that debt service stream, then the present value of CFADS divided by the balance is 1.30 by construction.
So on a perfectly sculpted loan, LLCR adds nothing. Which is exactly what makes it informative — because real loans are never perfectly sculpted, and the gap between LLCR and DSCR measures precisely how far the schedule departs from the shape the cash flow wants.
This post covers what each of the three ratios measures, why they diverge, what the divergence means, and which of them belongs in a covenant.
ℹ️ Note: Conventions vary more than covenant levels do — particularly on the discount rate and whether reserve balances are included. Two lenders quoting different LLCR requirements may be measuring different things.
What Does Each One Measure?
Three different questions about the same debt.
| Ratio | Question | Type |
|---|---|---|
| DSCR | Can the project pay this period's debt service? | Flow, per period |
| LLCR | Is there enough value over the remaining loan life to repay the balance? | Stock, whole-of-loan |
| PLCR | Is there enough value over the remaining asset life to repay the balance? | Stock, whole-of-asset |
The distinction between the first and the other two is the one that matters. DSCR is a liquidity test. LLCR and PLCR are solvency tests — they ask whether the asset is worth more than the loan, not whether this year's cash covers this year's payment.
DSCR: The Period Test
DSCR(t) = CFADS(t) ÷ Debt service(t)
It shows whether the project can cover its debt service in a single period from operating cash flow. It is what sizes the facility under the coverage constraint, what the covenants test, and what triggers lock-up.
Around 1.30× is the standard comfort level for a contracted renewable project, with the range and the drivers covered in the first post of this series.
Its weakness is that it is blind to everything outside the period. A project with a one-off maintenance year can breach a DSCR covenant while being entirely sound, and a project with strongly front-loaded cash flow can pass every annual DSCR test while accumulating a balance it will not be able to repay.
LLCR: The Stock Test
LLCR = PV(CFADS over remaining loan life) ÷ Outstanding debt
It measures how many times the discounted cash flow over the remaining loan life can repay the outstanding balance. Discounting is typically performed at the weighted average cost of debt specified in the term sheet — not at the equity discount rate, and not at a project WACC.
Two convention questions move the number materially and should be settled before any level is agreed.
The discount rate. Using the cost of debt is standard and is what makes the identity above hold. Using anything else breaks the relationship to DSCR and makes the ratio harder to interpret.
The reserve balance. Many definitions add the debt service reserve account balance to the present value of cash flow, on the basis that it is cash available to repay. That addition raises the ratio, by more in the early years when the reserve is large relative to the remaining cash flow.
Indicative covenant levels vary widely — a rule of thumb of 1.7× appears in some general infrastructure guidance — but the level matters less than the definition. A 1.5× LLCR including the DSRA and a 1.5× LLCR excluding it are different requirements.
PLCR: The Whole-Asset Test
PLCR = PV(CFADS over remaining project life) ÷ Outstanding debt
Identical to LLCR except that the cash flow runs to the end of the asset's life rather than to loan maturity. It therefore captures the tail — the years after the debt is repaid — and is always higher than LLCR where those years have positive cash flow.
On the worked schedule, with seven post-maturity years at 60% of final-year CFADS:
LLCR = 1.300
PLCR = 1.419
-----
Gap = 0.119
That gap is the tail, and it is the most useful thing PLCR produces. It says what the years beyond the loan are worth relative to the debt — which is the lender's ultimate recovery cushion if the schedule has to be restructured and extended.
Minimum PLCR covenants around 1.20× appear in practice, and lenders often use it alongside DSCR and LLCR as an additional sizing constraint. Its role in sizing is usually secondary: it binds only where the loan is short relative to the asset and the tail is large.
Why Does the Identity Matter?
Because it tells you what LLCR is actually detecting.
If a loan is sculpted so that DSCR equals the target in every period, then LLCR equals that target too, at every test date. The ratio contains no independent information.
So whenever LLCR differs from DSCR, something has prevented the schedule from following the cash flow. The usual causes:
- A level repayment schedule against a non-level cash flow.
- A minimum amortisation floor overriding the sculpt in some periods.
- A bullet or balloon at maturity.
- A cash sweep that has accelerated repayment.
- Actual performance diverging from the forecast the sculpt was built on.
Take the level-repayment version of the same loan. Debt service is flat at $16,858,864 and CFADS declines:
| Year | DSCR | LLCR | Outstanding |
|---|---|---|---|
| 1 | 1.542 | 1.535 | $175,879,531 |
| 6 | 1.467 | 1.485 | $144,981,881 |
| 11 | 1.395 | 1.439 | $102,649,424 |
| 18 | 1.300 | 1.385 | $15,829,919 |
Two things are visible that a sculpted loan would not show.
DSCR falls and LLCR falls more slowly. The period ratio deteriorates as cash flow declines against fixed debt service; the whole-of-life ratio is an average and smooths it.
They cross. In year 1 DSCR is above LLCR; by year 6 LLCR is above DSCR. The crossover is the point at which the remaining schedule stops being over-covered on a whole-of-life basis and starts being under-covered in the near term. It is a genuinely useful early warning and it is invisible if only one ratio is tracked.
The sign of the gap is the diagnosis, and it is worth stating as a rule because it generalises beyond this example.
LLCR below DSCR means the near term is better covered than the remainder — cash flow is front-loaded relative to the repayment schedule, or the schedule is back-ended. The risk sits at the far end, and the question to ask is whether the balance outstanding in the late years is supportable.
LLCR above DSCR means the remainder is better covered than the near term. Something is squeezing the current period — a maintenance year, a soft price period, a step-up in debt service — against an asset that is fundamentally sound over the loan. The question is liquidity, not solvency, and the answer is usually a reserve rather than a restructuring.
Those are two entirely different problems requiring two entirely different responses, and a covenant package reporting only DSCR cannot distinguish them. That is the practical case for tracking both, and it does not depend on either ratio breaching anything.
Forward-Looking, Backward-Looking, or Both?
A convention question that decides when a breach actually happens, and it is settled in the credit agreement rather than by any principle.
Backward-looking (historic) DSCR tests the twelve months just ended: did the project cover its debt service from the cash it actually generated? It is factual, unarguable, and late — by the time it breaches, the damage has occurred.
Forward-looking (projected) DSCR tests the twelve months ahead on the current forecast. It is timely and it is an opinion, which means it is negotiable, and a borrower with discretion over the forecast has discretion over the covenant.
Most agreements use both, and the drafting question is what each one triggers. The usual and sensible arrangement is that the historic ratio drives the consequences — lock-up, default — while the forward ratio drives obligations to notify, to produce a remediation plan, or to fund a reserve.
LLCR and PLCR are inherently forward-looking, since both are present values of future cash flow. That is worth noticing: a covenant package that tests only historic DSCR and forward LLCR has one backward-looking test and one whose inputs the borrower prepares.
Two practical consequences.
The forecast underlying a forward test needs a defined basis. Whose assumptions, updated how often, approved by whom. A forward-looking covenant against an undefined forecast is unenforceable in any useful sense.
Test dates matter as much as levels for seasonal assets. A project with a strong summer and a weak winter will show very different twelve-month ratios depending on whether the test date falls in March or September. Quarterly testing on a rolling twelve-month basis largely removes the problem; annual testing on a fixed date does not, and sponsors have been known to prefer whichever date flatters.
Where Does the DSRA Fit?
In two places, which is why its treatment needs to be explicit.
The debt service reserve account holds cash — typically six or twelve months of debt service — available to meet payments if operating cash flow falls short. It is a liquidity buffer, and its primary job is to prevent a temporary shortfall becoming a default.
In the ratios it appears, or does not, depending on convention.
In DSCR it normally does not. The period ratio is meant to measure whether operating cash flow covers debt service. Adding a reserve drawing to the numerator would make the ratio pass in exactly the circumstances it exists to detect. Where a drawing occurs, the ratio should show the shortfall and the reserve should separately show the cover.
In LLCR and PLCR it commonly does. Both are solvency measures asking whether available value covers the outstanding balance, and cash in a reserve account is unambiguously available value. Including it is defensible; excluding it is more conservative; and the two produce noticeably different numbers in the early years when the reserve is large relative to the discounted cash flow remaining.
The point to insist on is consistency across the covenant package. A model that includes the DSRA in LLCR and a term sheet that quoted an LLCR level assuming exclusion have not agreed on anything, and the gap will not be discovered until a test date is close.
Why Do Ratios Stop Meaning Anything?
Because the numerator is negotiable, and a covenant is only as strong as the definition underneath it.
The first post in this series made the point about CFADS being a defined term rather than an accounting fact. It applies with more force here, because a covenant converts that definition into a legal trigger.
The recurring pattern is a definitional arms race. A borrower seeks add-backs — extraordinary items, one-off maintenance, a bad-weather adjustment, the release of a reserve. Each is individually arguable. Collectively they can lift a ratio far enough that it no longer measures the thing it was set to measure, and the covenant passes in circumstances that would plainly have breached the version originally negotiated.
Two defences are worth knowing.
Cap the add-backs, in aggregate and individually. A covenant permitting unlimited add-backs subject to reasonableness is not a covenant. A cap of a stated percentage of CFADS, with a list of permitted categories, is.
Test the covenant against the sizing case. The ratio in the credit agreement should measure the same quantity the facility was sized on. Where they differ — and they often do, because sizing happens in a model and covenants are drafted in a document — the covenant is testing something the structure was never calibrated against.
The blunt version: if the borrower can move the ratio more than the business can, it is not a covenant.
Which Belongs in a Covenant?
Different ones, for different purposes, and the choice is not arbitrary.
DSCR belongs in the lock-up and default tests. It is a liquidity measure, it responds quickly, and it is what actually determines whether the project can pay. A lock-up triggered by DSCR stops distributions at the moment cash cover thins, which is what a lock-up is for.
LLCR belongs in the additional-debt and restructuring tests. It is a solvency measure and it is the right question to ask before permitting more leverage or agreeing an amendment: is the remaining value sufficient for the balance?
PLCR belongs in sizing and in workout analysis. It is where the tail enters, and it answers the question a lender asks when a schedule needs extending — is there enough asset life left to repay from?
Using DSCR alone is the common failure and it produces two specific blind spots. A project can pass every annual DSCR test and have an LLCR below one, if the cash flow is front-loaded and the balance is not amortising fast enough. And it can breach a single-year DSCR because of a one-off maintenance event while being entirely solvent on both stock measures — which is why well-drafted agreements distinguish a lock-up from a default, and the next post is about exactly that.
How Do You Build These in Excel?
As three separate calculations with the conventions made explicit, plus the identity as a self-test.
The three ratios
DSCR(t) = CFADS(t) / Debt_Service(t)
LLCR(t) = ( NPV(Debt_Rate, CFADS from t+1 to maturity) + DSRA_Balance(t) )
/ Outstanding_Debt(t)
PLCR(t) = ( NPV(Debt_Rate, CFADS from t+1 to end of project life) + DSRA_Balance(t) )
/ Outstanding_Debt(t)
Note that all three use the same discount rate — the cost of debt — and that the DSRA term is a switch, not a constant. Build it as a flag so the same model can produce both conventions.
The self-test
IF the schedule is sculpted to a constant target:
LLCR(0) should equal Target_DSCR to within rounding
Put that check in the model. If a loan is supposed to be sculpted and LLCR at close does not equal the target, the sculpt is wrong — and this test catches it faster than reading the amortisation schedule.
The gap analysis, which is the real output
LLCR − DSCR = how far the schedule departs from the cash flow shape
PLCR − LLCR = what the post-maturity tail is worth, per dollar of debt
On the worked cases: the sculpted loan has an LLCR−DSCR gap of zero, and a PLCR−LLCR gap of 0.119. The level loan has an LLCR−DSCR gap that starts at −0.007 and reaches +0.085 by year 18.
Publish both gaps rather than three raw ratios. A reader given three numbers has to reconstruct the relationship; a reader given the gaps has the diagnosis.
ℹ️ Note: Where LLCR is used as a covenant, the definition — discount rate, DSRA inclusion, whether it is tested forward-looking or on actuals — belongs in the credit agreement with the same precision as the CFADS definition. A 1.50× covenant against an undefined LLCR is not a covenant, it is a future dispute.
To build the three ratios with convention switches, the sculpt self-test and the gap analysis, prompt Dezzmond with your cash flow profile and amortisation schedule.
What Do Modellers and Lenders Actually Check?
- Is LLCR discounted at the cost of debt, and is the DSRA included or excluded?
- Does LLCR at close equal the target DSCR on a sculpted facility? If not, the sculpt is broken.
- Where do DSCR and LLCR cross, and what does that say about the repayment shape?
- What is the PLCR−LLCR gap, and therefore what is the tail worth?
- Is the lock-up tested on DSCR and additional debt on LLCR, rather than one ratio doing both jobs?
- Are the ratios tested forward-looking, backward-looking, or both?
- Do two lenders quoting different LLCR levels actually mean the same ratio?
Frequently Asked Questions
What is the difference between DSCR and LLCR?
DSCR is a flow measure for a single period — can the project pay this year's debt service. LLCR is a stock measure over the whole remaining loan — is the present value of future cash flow sufficient to repay the outstanding balance.
Why does LLCR equal the target DSCR on a sculpted loan?
Because debt service in each period is CFADS divided by the target, and the outstanding balance is the present value of that debt service stream. Dividing the present value of CFADS by the balance therefore returns the target exactly.
What discount rate should LLCR use?
The weighted average cost of debt from the term sheet. Using an equity discount rate or a project WACC breaks the relationship to DSCR and makes the ratio much harder to interpret.
What does PLCR add?
The tail. PLCR runs the cash flow to the end of the asset's life rather than to loan maturity, so the gap between PLCR and LLCR measures what the post-maturity years are worth per dollar of debt — the lender's cushion if the loan has to be extended.
Should DSCR be tested historically or on a forecast?
Usually both, with the historic ratio driving lock-up and default and the forward ratio driving notification and remediation. A forward test is only meaningful where the forecast basis — whose assumptions, updated how often, approved by whom — is defined.
Should the DSRA be included in the ratios?
Normally not in DSCR, which is meant to measure whether operating cash flow covers debt service. Commonly yes in LLCR and PLCR, which measure available value against the balance. What matters is that the model and the term sheet use the same convention.
Can a project pass DSCR and fail LLCR?
Yes. Front-loaded cash flow with slow amortisation can clear every annual DSCR test while leaving a balance the remaining cash flow cannot repay. That is precisely the case a stock measure exists to catch.
Closing: Three Ratios, Two Questions, One Diagnosis
The coverage ratios are usually taught as a list and reported as a table of three numbers. They are more useful understood as two questions and a residual.
The two questions are liquidity and solvency: can it pay now, and is it worth more than what it owes. DSCR answers the first. LLCR and PLCR answer the second over two different horizons.
The residual is the interesting part. On a perfectly sculpted loan, LLCR contains no information beyond the target it was sculpted to — which means every deviation of LLCR from DSCR is telling you something specific about how the repayment schedule differs from the shape of the cash flow. A level schedule against declining cash flow, a minimum amortisation floor, a bullet, a sweep, or simple underperformance: each leaves a distinct signature in that gap.
Reporting three ratios hides that. Reporting the two gaps — LLCR minus DSCR, and PLCR minus LLCR — makes the structure legible in a way that no amount of staring at the raw numbers achieves.
The next post takes the ratios into the credit agreement: how covenant testing actually works, what a lock-up does that a default does not, and why the distinction between them is the most consequential drafting in the document.
Sources: Forvis Mazars — LLCR Explained: A Project Finance Lender's Key Metric · LexisNexis — Project Finance Financial Covenants · The Financial Modelling Podcast — Project Life Cover Ratio · MMCG — Coverage Ratios in Real Estate Finance: DSCR, LLCR and PLCR