Underwriting the Downside: Financial Due Diligence for Private Lenders

A private lender's return is capped at the coupon. It does not share in the equity upside if the business outperforms, but it absorbs the loss if the business fails. That single asymmetry reshapes the entire purpose of financial due diligence: a lender is not underwriting how big the company can become, but the probability it is not repaid and what it recovers if things go wrong.

Key takeaways

• Credit diligence is asymmetric and pessimistic. Because the lender's return is capped at the coupon, the dominant question is the probability of loss and the recovery in default — not the size of the upside.
• Cash conversion of EBITDA is the centre of gravity. What matters is whether adjusted EBITDA becomes cash that can service debt after capex, tax, and working-capital swings — not the headline earnings figure.
• DSCR and FCCR are the serviceability tests, but neither has a universal formula. Both are defined per credit agreement, and the definitions are where diligence earns its keep.
• Maintenance covenants are the heart of downside protection — and "private credit means strong covenants" is no longer blanket-true. The picture is size-driven and eroding in the upper-middle market.
• Cash-flow lending and asset-based lending are distinct diligence workstreams. One underwrites going-concern enterprise value; the other underwrites a dynamic borrowing base of liquid collateral.

The asymmetry that defines credit diligence

A private-equity buyer underwrites the upside. It owns the equity, its return is uncapped, and so its diligence asks how big this can get and what it is worth. It builds base and bull cases, scrutinises growth, margin, and scalability, and treats the balance sheet primarily as an input to the price bridge.

A lender or private-credit fund underwrites the downside. Its return is capped at the coupon, so the worst case is not a lower return — it is a loss of capital. The dominant question becomes: what is the probability I am not repaid, and what do I recover if things go wrong? Credit diligence is therefore pessimistic by design. It stress-tests cash generation, liquidity, leverage capacity, collateral, and recovery in a default scenario. The bull case is largely irrelevant; the floor of performance through a cycle is everything.

This reframes the same underlying FDD data. Where an equity buyer fixates on growth and margin trajectory, a lender cares far more about three things: the cash conversion of EBITDA — does adjusted EBITDA actually turn into cash to service debt, after capex, tax, and working-capital movements? — the floor of performance through a downturn, and the enforceability and recovery value of collateral. The same quality-of-earnings file is read from the opposite end.

The positioning of the lender in the capital structure governs the analysis. A first-lien senior-secured lender sits at the top: in a restructuring it recovers ahead of mezzanine and equity. If enterprise value in a downside scenario still exceeds the senior debt, the senior lender is made whole. So credit diligence obsesses over where value breaks in a stressed enterprise-value scenario, and whether the lender's claim sits above that break point. Everything else — the ratios, the covenants, the collateral mechanics — is in service of that single question.

Dimension	PE / equity buy-side	Lender / private credit
Primary question	How big can this get, and what is it worth?	What is the probability of loss, and what do I recover?
Earnings focus	Growth, margin, scalability	Cash conversion, sustainability of the floor
Forecast use	Base / bull case for valuation	Downside / stress case: can it service debt in a recession or at higher rates?
Balance sheet	Net debt for the price bridge	Liquidity runway, undrawn revolver, collateral and recovery
Key ratios	EV/EBITDA, IRR, MOIC	DSCR, FCCR, Net Debt/EBITDA, ICR, LTV / borrowing base
Protection	Reps and warranties, escrow	Maintenance covenants, security and lien position, intercreditor terms
Worst case	Lower return	Capital loss — hence first-lien senior-secured positioning

Serviceability: the DSCR test

The headline serviceability metric is the debt service coverage ratio (DSCR). In its corporate, cash-flow form:

DSCR = (EBITDA − unfinanced capex − cash taxes) / (interest + scheduled principal amortisation)

A simpler textbook form is EBITDA / total debt service. The defining feature that separates DSCR from a plain interest-cover ratio is that the denominator includes principal amortisation, not just interest. The lender is asking whether cash covers the full debt service obligation, principal included — because that is what actually has to be paid.

DSCR is a multiple. A ratio of 1.0x means cash exactly covers debt service with no cushion; below 1.0x means cash flow does not cover the obligation. Typical minimum thresholds vary by asset class: commercial real estate generally sits at 1.20x–1.35x; project and infrastructure finance at 1.30x–1.50x; SBA and small-business lending around 1.15x. Since 2008 the 1.25x–1.35x band has become the general "new normal" floor. In real-estate and project structures the numerator changes shape — DSCR is computed as net operating income (NOI) over total debt service, and in project finance the numerator is cash flow available for debt service (CFADS) — but the principle is identical.

A worked DSCR example

Consider a borrower with adjusted EBITDA of $10.0m. Maintenance capex is $1.5m, cash taxes $1.0m, cash interest $4.0m, and mandatory amortisation is 1% of a $40m term loan, or $0.4m.

DSCR = (10.0 − 1.5 − 1.0) / (4.0 + 0.4) = 7.5 / 4.4 = 1.70x

The business generates $1.70 of free cash for every $1.00 of debt service. Comfortable — at today's rates.

Now run the floating-rate stress test that private-credit diligence runs as a matter of course. Hold operations flat but assume rates rise so that cash interest climbs to $6.0m:

DSCR = 7.5 / 6.4 = 1.17x

The same business, unchanged operationally, has moved into covenant territory purely on rate movement. This is why, in a rising-rate environment, coverage covenants matter more than leverage covenants: leverage is a snapshot of the balance sheet, but coverage is what actually fails first when the cost of debt rises against a fixed cash-generating engine.

The wider family of ratios

The interest coverage ratio (ICR) is EBITDA / cash interest. It is simpler than DSCR because it ignores amortisation, which makes it the natural fit for cash-flow leveraged structures with light or bullet amortisation, such as a Term Loan B.

Leverage is Net Debt / EBITDA, where net debt is total debt less cash. It is the single most-watched leverage metric and is often layered into senior (first-lien) net leverage and total net leverage. Typical mid-market private-credit leverage runs around 3.5x–4.5x in the lower mid-market and 5.0x–5.5x for upper mid-market buyouts. US leveraged buyouts averaged roughly 4.9x in 2024, and leverage maintenance covenants — where they exist — average around a 4.4x threshold, set with headroom above the opening level.

The fixed-charge coverage ratio (FCCR) is broader than DSCR: it captures all fixed obligations, including operating leases and sometimes preferred dividends. Critically, there is no single standard formula — FCCR is defined per credit agreement. A common cash form is:

FCCR = (EBITDA − capex − cash taxes) / (cash interest + mandatory debt amortisation)

A worked FCCR example

Take a borrower with EBITDA of $20m, capex of $2.5m, cash taxes of $5m, cash interest of $2.25m, and mandatory amortisation of $4m.

Covenant-adjusted cash flow is 20 − 2.5 − 5 = $12.5m. Total fixed charges are 2.25 + 4 = $6.25m.

FCCR = 12.5 / 6.25 = 2.0x

Typical minimum FCCR covenants run around 1.0x–1.25x, sometimes as high as 1.5x. The FCCR is the most common single financial covenant in the lower-mid-market and in asset-based lending, frequently structured as a springing covenant. The Wall Street Prep primer on the FCCR is a useful reference on how the components vary in practice.

A standing warning for any reader auditing these calculations: DSCR and FCCR have no universal formula. Whether the numerator uses maintenance capex or total capex, cash or accrual taxes, how operating leases are treated, and which tranches of amortisation count — all vary by agreement. The number is only as good as the definition behind it, and the definition lives in the credit agreement, not in a textbook.

Covenants: the early-warning system

If serviceability ratios tell a lender whether the business can pay today, covenants determine what the lender can do tomorrow. The critical distinction is between maintenance and incurrence covenants.

Maintenance covenants must be satisfied on an ongoing, periodic basis — usually tested quarterly — regardless of whether the borrower does anything. A typical formulation: total net leverage must not exceed 5.0x at each quarter-end. They are the lender's early-warning system. A breach, even while payments are current, is a technical default that brings the borrower to the table. That is what allows the lender to intervene, reprice, demand a cure, or accelerate. This is the heart of private-credit downside protection: the ability to act on deterioration before a missed payment, not after.

Incurrence covenants, by contrast, are tested only when the borrower takes a specified action — incurring new debt, paying a dividend, making an acquisition. They restrict actions, not ongoing performance. The protection they offer is far weaker: a borrower can deteriorate for years without tripping anything, provided it takes no triggering action.

A springing covenant is a maintenance covenant that only "springs" into testing when a condition is met — classically a revolver springing FCCR or leverage covenant tested only when revolver utilisation exceeds a threshold, often around 35–40% drawn. It is common in otherwise covenant-light structures as a way to protect the revolving lenders without imposing a full maintenance regime on the whole facility.

The cov-lite reality — and where the traps are

The market context matters. In broadly-syndicated and institutional leveraged loans — the Term Loan B market — cov-lite is now dominant. It accounted for roughly 20% of leveraged loans in 2007, more than 86% of loans outstanding by 2021, and about 93% of US institutional leveraged-loan volume issued in 2024. Here "cov-lite" has a precise meaning: no maintenance covenants — only incurrence covenants. The LSTA's loan-market covenant trends track this shift in detail.

Private credit and direct lending were historically the opposite of cov-lite. Maintenance covenants and lender control were precisely the edge these funds sold to their LPs. But that edge is eroding as funds chase larger deals. Roughly 40% of upper-middle-market private-credit deals in 2024 were cov-lite. As S&P Global Ratings documents in its analysis of how loose maintenance covenants permeate private credit, the dilution is real and size-driven. Below roughly $50m–$100m of EBITDA, deals almost always carry at least one maintenance covenant — commonly total-leverage and/or FCCR. Above roughly $100m of EBITDA, and in large-cap unitranche, cov-lite is increasingly the norm.

Four oversimplifications are exactly where a careless reader gets caught:

1. "Private credit means strong covenants" is no longer blanket-true. It is segment-dependent, and the protection is eroding fast in the upper-middle market and large-cap. The right question is always which segment, which size band.
2. DSCR and FCCR have no universal formula. Both are definitionally per-agreement. State the definition; never assume it.
3. "Cov-lite" does not mean "covenant-free." It means no maintenance covenants. Incurrence covenants remain, and there is usually a springing revolver covenant protecting the revolving lenders. A cov-lite loan is not an unprotected loan; it is a loan with a slower trigger.
4. Adjusted-EBITDA gaming is a credit issue, not just a price issue. Expanding EBITDA add-back definitions in modern credit agreements mean covenant EBITDA can drift well above audited EBITDA. For a buyer, an inflated add-back is a question of price. For a lender, it directly loosens every leverage and coverage covenant calculated off that number — which makes the add-back definition itself a core diligence item.

Two lending models, two diligence workstreams

Not all lending underwrites the same thing. The split between cash-flow lending and asset-based lending (ABL) determines the entire shape of diligence.

Cash-flow lending

In cash-flow lending the loan is sized and serviced off EBITDA and free cash flow. Collateral typically takes the form of a blanket lien, but recovery ultimately depends on the enterprise value of the business as a going concern. The loan is sized via leverage — Net Debt / EBITDA — and tested via coverage: DSCR, FCCR, ICR. This is the dominant private-credit, unitranche, and Term Loan B model. The diligence centre of gravity is quality of earnings, cash conversion, and the durability of EBITDA through a cycle, because there is no liquid asset pool to fall back on — only the value of the enterprise itself.

Asset-based lending

ABL underwrites something different. The loan is sized and governed dynamically by a borrowing base — a fluctuating cap equal to the value of eligible liquid collateral, recalculated frequently, often monthly or weekly. In its standard form:

Borrowing Base = (Eligible AR × AR advance rate) + (Eligible Inventory × Inventory advance rate) − Reserves

Typical advance rates run around 80%–85% of eligible receivables and 50%–60% of inventory. Inventory rates are lower and far more variable: finished goods advance higher than work-in-progress, which advances higher than raw materials, and perishable, custom, or specialised inventory attracts steep haircuts or outright exclusion.

A worked borrowing-base example

A borrower has $10m of eligible accounts receivable and $20m of eligible inventory.

AR: $10m × 85% = $8.5m Inventory: $20m × 60% = $12.0m Borrowing base = $8.5m + $12.0m = $20.5m

That $20.5m is the maximum the borrower may draw — not a static commitment, but a number that moves every time the collateral pool moves. The Wall Street Prep borrowing-base reference sets out the mechanics in further detail.

What ABL diligence actually tests

ABL is a fundamentally different diligence workstream from cash-flow quality of earnings:

• Eligibility and ineligibles. Receivables more than 90 days past due, intercompany balances, foreign or government obligors, contra accounts, and disputed invoices are excluded before the advance rate is applied. The advance rate is meaningless until the ineligibles are stripped out.
• Concentration limits. A single large customer is capped — for example, no obligor representing more than 15%–20% of the base — so that the collateral is not hostage to one relationship.
• Dilution. The rate of credit notes, returns, and write-offs against AR. High dilution forces lower advance rates or additional reserves, because it signals that face-value receivables overstate collectible cash.
• Reserves. Lender deductions for rent and landlord-lien risk, unpaid taxes, and accrued payroll — claims that would rank ahead of, or alongside, the lender's recovery on the collateral.
• Field exams and NOLV appraisals. Periodic field examinations and inventory net-orderly-liquidation-value appraisals establish what the collateral would actually fetch in a controlled wind-down — a workstream with no real equivalent in cash-flow diligence.

ABL suits asset-rich, cyclical, lower-margin, or volatile-earnings businesses — distribution, manufacturing, retail — where EBITDA is an unreliable basis for credit but receivables and inventory are genuinely liquid. It is generally light on financial covenants, often carrying only a springing FCCR tied to availability, because the borrowing base itself is the discipline: as collateral shrinks, so does the permitted draw, automatically and continuously.

Jurisdiction: the US and UK/European picture

The documentary and structural conventions differ across jurisdictions, and the differences matter for diligence.

US deals are built on LSTA precedents; UK and European deals on LMA precedents. LMA documents have historically been somewhat more lender-protective, though European leveraged loans have converged toward US-style cov-lite over time. Cov-lite originated, and remains most extreme, in the US institutional Term Loan B market, where it now exceeds 90% of issuance. UK and European leveraged loans followed, but mid-market and private-credit deals on both sides retain more maintenance covenants — governed by the same size-based split that applies in the US.

ABL is a deeper, more mature product in the US than in the UK and Europe. Its relative prevalence reflects differences in security and enforcement: the US relies on the Article 9 UCC regime, while the UK uses fixed and floating charges and qualifying-floating-charge administration. ABL is growing in the UK and Europe but remains less established than across the Atlantic.

One further US development is reshaping intercreditor diligence: liability-management exercises (LMEs) — uptier and drop-down transactions that can subordinate or strip value from existing lenders. As a result, lenders increasingly diligence the credit agreement's own loopholes — anti-cooperation and pari-passu protection provisions — as carefully as they diligence the borrower's financials. The document itself has become a source of risk, not merely a record of terms.

Frequently asked questions

Why does financial due diligence for lenders focus on the downside rather than the upside?

Because a lender's return is capped at the coupon. It does not benefit if the business outperforms, but it bears the loss if the business fails. The risk is therefore asymmetric: the most a lender can do is be repaid in full, while the most it can lose is its capital. Diligence concentrates on the probability of non-repayment and the recovery in a default scenario — stress-testing cash generation, liquidity, leverage, collateral, and where value breaks in a downside enterprise-value case — rather than on growth potential.

What is the difference between DSCR and the fixed-charge coverage ratio?

DSCR measures whether cash covers debt service — interest plus scheduled principal amortisation — and its defining feature is that it includes principal, not just interest. FCCR is broader: it captures all fixed obligations, including operating leases and sometimes preferred dividends. Neither has a universal formula; both are defined per credit agreement, so the treatment of capex, taxes, leases, and which amortisation counts must be read from the document rather than assumed.

Does "cov-lite" mean a loan has no covenants?

No. Cov-lite means no maintenance covenants — covenants tested periodically regardless of borrower action. Incurrence covenants, which restrict specific actions such as new debt or dividends, remain in place, and there is usually a springing revolver covenant protecting the revolving lenders. The practical consequence is a slower trigger: without maintenance covenants, a borrower can deteriorate for an extended period before the lender gains a contractual seat at the table.

When does a lender use asset-based lending instead of cash-flow lending?

ABL fits asset-rich, cyclical, lower-margin, or volatile-earnings businesses — distribution, manufacturing, retail — where EBITDA is an unreliable basis for credit but receivables and inventory are liquid. The loan is governed by a borrowing base that recalculates frequently against eligible collateral, so the permitted draw shrinks automatically as collateral shrinks. Cash-flow lending, by contrast, sizes the loan off EBITDA and depends on going-concern enterprise value for recovery, and is the dominant model for unitranche and Term Loan B structures.

How DiligenceForge supports credit teams

DiligenceForge is built for the way private lenders actually underwrite: downside-first, with the focus on cash conversion and serviceability rather than the growth story. It is designed so that every figure traces back to its source, so a credit committee can see exactly how an adjusted-EBITDA number, a coverage ratio, or a borrowing-base input was derived — and confidentiality is treated as a first-order requirement throughout.

If you are a private lender or private-credit team and want to see how it fits your process, you can request access to explore it with your own deals.