Documentation and covenants

Total net leverage covenant

A maintenance test capping net debt as a multiple of adjusted EBITDA, measured quarterly on a last-twelve-months basis with pro forma effect for acquisitions and disposals. It is the covenant that most often steps down over the life of the loan.

Field	Value
Formula	Net leverage = (Total debt - min(unrestricted cash, netting cap)) / Adjusted EBITDA_LTM <= L_max. EBITDA headroom = 1 - (Net debt / L_max) / EBITDA_current
Worked	Debt 220, unrestricted cash 15 with a 15 cap, EBITDA 40: (220 - 15) / 40 = 5.125x against a 6.00x test
EBITDA headroom	Breach when EBITDA < 205 / 6.00 = 34.17, so EBITDA can fall 14.6 percent
Debt headroom	At constant EBITDA, net debt can rise to 6.00 * 40 = 240, giving 35 of additional net debt capacity
Step-down effect	A step-down to 5.50x at the next test date cuts EBITDA headroom to 1 - (205/5.50)/40 = 6.8 percent without any change in performance

• Express headroom as a percentage EBITDA decline rather than as turns. Turns of headroom are not comparable between a 3.0x credit and a 6.0x credit: half a turn of cushion on 6.0x is an 8 percent EBITDA decline, and on 3.0x it is 14 percent.
• Step-downs are the mechanism by which a covenant tightens without a negotiation. A package with generous opening headroom and aggressive step-downs is a package that will be amended, and the amendment is the lender's second bite.
• The cash netting cap is worth checking. Uncapped netting lets a borrower draw its revolver, hold the proceeds, and improve its reported net leverage - the drawn debt and the held cash cancel while liquidity risk has increased.
• Because the denominator is LTM and pro forma, a covenant computed on the acquisition-date pro forma figure can be met by an acquisition that has not yet contributed a single quarter of actual results.

Interest coverage ratio

Adjusted EBITDA divided by cash interest expense for the period. In a floating-rate structure it is the covenant that responds to rate moves rather than to operating performance, which makes it the binding test in a rising-rate environment even where leverage is unchanged.

Field	Value
Formula	ICR = Adjusted EBITDA_LTM / Cash interest expense_LTM >= ICR_min. Breakeven all-in rate = EBITDA / (ICR_min * Total debt)
Worked	EBITDA 40, debt 220 at a 9.00 percent all-in rate so cash interest = 19.80: ICR = 40 / 19.80 = 2.02x against a 1.50x test
EBITDA headroom	Breach when EBITDA < 1.50 * 19.80 = 29.70, a 25.8 percent decline
Rate headroom	Breakeven interest = 40 / 1.50 = 26.67, implying an all-in rate of 26.67 / 220 = 12.12 percent. The rate can rise 312 bps.
Both at once	A 10 percent EBITDA decline to 36 cuts the rate headroom to 36/1.50/220 = 10.91 percent, or 191 bps. The two sensitivities are multiplicative, not additive.

• Stating the covenant as a breakeven interest rate rather than a ratio is what turns it into a usable risk measure, because the base rate is observable and the forward curve is quoted. A borrower whose breakeven rate is inside the forward curve has already breached; it just has not been tested yet.
• Whether the definition says cash interest or total interest decides whether a PIK toggle cures the covenant. On a cash-interest definition, capitalising interest raises the ratio - the covenant improves as the credit deteriorates.
• Hedging changes the answer and is frequently excluded. If a required interest rate cap is in place, the covenant should be computed on the hedged rate, and the document should say which.
• Interest coverage and leverage bind at different points in the cycle. Leverage binds when EBITDA falls; coverage binds when rates rise. A package with only a leverage covenant has no protection against the second.

Fixed charge coverage ratio

A cash-flow coverage test measuring EBITDA less non-discretionary cash uses against fixed obligations, being at minimum cash interest and scheduled amortisation and often also rent, taxes, and dividends. It is the standard test in ABL and in cash-flow-poor credits.

Field	Value
Formula	FCCR = (Adjusted EBITDA - capex - cash taxes) / (Cash interest + scheduled amortisation + other fixed charges) >= FCCR_min
Worked	EBITDA 40, capex 6, cash taxes 3, cash interest 19.80, amortisation 2.20: (40 - 6 - 3) / (19.80 + 2.20) = 31 / 22 = 1.41x
Capex sensitivity	Cutting capex to 2 raises the ratio to 35 / 22 = 1.59x. The borrower can cure the covenant by underinvesting.
Maintenance-only variant	Where the definition uses maintenance capex rather than total capex, a growth-capex programme no longer affects the ratio - and neither does cancelling it

• Capex in the numerator is the reason FCCR is a weak covenant in a deteriorating credit: the cheapest way to satisfy it is to stop investing, which improves the ratio and degrades the asset. A lender who wants the covenant to mean something specifies maintenance capex or a floor.
• Springing FCCR tests in ABL facilities are triggered by an availability threshold. Because availability falls as the borrowing base shrinks, the test appears at exactly the moment the borrower's cash flow is worst - which is the design, but it also means the covenant produces defaults rather than early warnings.
• FCCR is the most definition-dependent of the standard covenants. Whether rent, preferred dividends, cash taxes, and the revolver's average outstanding balance are inside it varies more between documents than the test level does.

Covenant-lite

A structure in which the term loan has no maintenance financial covenant. Where a revolver exists it usually carries a springing covenant tested only when revolver utilisation exceeds a threshold, so the term lenders benefit from it only indirectly and only while the revolver is drawn.

Field	Value
What remains	Incurrence-based tests on debt, liens, restricted payments and investments, plus reporting and negative covenants
What is absent	Any quarterly test that a decline in performance can fail, and therefore any lender-side trigger that does not depend on borrower action
Springing covenant	Typically tested when revolver utilisation exceeds 35 to 40 percent of the commitment, for the benefit of the revolving lenders

• The consequence of covenant-lite is not a higher default rate; it is a later default date. Problems surface at a liquidity event or a maturity rather than at a quarterly test, by which point enterprise value has had longer to fall below the debt. Empirically the pattern shows up as unchanged or higher recoveries per default and a longer period of deterioration beforehand.
• A springing covenant for the revolver's benefit gives the term lenders nothing, because a borrower approaching the trigger simply repays the revolver from cash. The trigger is under the borrower's control.
• Covenant-lite and covenant-loose are not the same. A cov-loose package has a maintenance covenant set so wide it cannot be breached without a catastrophe. That is still better than none, because the covenant creates a reporting and certification obligation and a defined default event that can be leveraged.
• In direct lending the meaningful question is not whether a maintenance covenant exists but whether the lender group is small enough to act on it. A covenant held by two lenders is a negotiating right; the same covenant across sixty holders is a coordination problem.

EBITDA addbacks and adjustments

Items added to reported EBITDA under the credit agreement's Consolidated EBITDA definition, converting an accounting figure into the contractual figure every ratio is computed against. The definition is negotiated and is not an accounting standard.

Field	Value
Formula	Adjusted EBITDA = Reported EBITDA + addbacks, subject to a cap. Where the cap is a percentage p of adjusted EBITDA: A <= p*(EBITDA_reported + A), so A_max = p*EBITDA_reported / (1 - p).
Worked, cap on adjusted	Reported 32, cap 25 percent of adjusted: A_max = 0.25 * 32 / 0.75 = 10.67. Adjusted EBITDA = 42.67. Net leverage on 205 of net debt = 4.805x.
Worked, cap on unadjusted	A_max = 0.25 * 32 = 8.00. Adjusted EBITDA = 40.00. Net leverage = 5.125x.
Value of the denominator	0.320 turns of leverage, from the same stated 25 percent cap. The word 'adjusted' inside the cap is worth about a third of a turn here.
General identity	A cap of p on adjusted EBITDA is equivalent to a cap of p/(1-p) on reported EBITDA. A 25 percent cap on adjusted is a 33.3 percent cap on reported.

• The identity p on adjusted equals p/(1-p) on reported is the single most useful piece of arithmetic in covenant review, and the caps are almost always quoted without saying which denominator applies. At a 35 percent cap the difference is larger still: 53.8 percent of reported.
• Every ratio in the package - leverage, coverage, incremental capacity, restricted payment capacity, and the excess cash flow sweep - runs off the same adjusted number. A wide EBITDA definition loosens all of them simultaneously, which is why negotiating the definition dominates negotiating any single test level.
• Run-rate synergy addbacks have three governing parameters: the cap, the lookforward window in which the savings must be realised, and whether any third party must certify them. A generous cap with a long window and no certification requirement is an unpoliced number.
• Addbacks compound through acquisitions. A target's LTM EBITDA is itself an adjusted figure, and adding it pro forma imports its adjustments into the group's. A serial acquirer's adjusted EBITDA can drift a long way from any cash number without a single individually unreasonable addback.

Equity cure right

A right for the sponsor to remedy a financial covenant breach by contributing cash equity within a stated period after the test date. The economics turn entirely on how the cure amount is applied: credited to EBITDA for the covenant calculation, or applied to reduce debt, or both.

Field	Value
Formula	EBITDA-credit cure: X = Net debt / L_max - EBITDA_actual. Debt-paydown cure: X = Net debt - L_max * EBITDA_actual.
Worked, EBITDA credit	Net debt 205, covenant 6.00x, actual EBITDA 32: required cure X = 205/6.00 - 32 = 34.17 - 32 = 2.17
Worked, debt paydown	Same facts: X = 205 - 6.00 * 32 = 205 - 192 = 13.00
Ratio	13.00 / 2.17 = 6.0x. The debt-paydown formulation requires almost exactly L_max times as much cash as the EBITDA-credit formulation.
General identity	X_paydown = L_max * X_EBITDA. The multiple is the covenant level itself.

• The identity X_paydown = L_max * X_EBITDA is the whole negotiation, and it is rarely stated. At a 6.00x covenant the two formulations differ by a factor of six in the cash the sponsor must write. A sponsor who agreed to an EBITDA-credit cure has bought the covenant back for a sixth of the price.
• An EBITDA-credit cure that persists in the calculation for the following three quarters cures four tests with one cheque, because the LTM figure carries the credited amount forward. Whether the credit applies to that quarter only or to every LTM calculation including it is a term worth more than the cap on the number of cures.
• Caps are usually expressed as a maximum number of cures, a maximum in consecutive quarters, and an aggregate amount. A package permitting cures in non-consecutive quarters only is materially tighter, because a deteriorating credit fails consecutively.
• Whether cure proceeds may be netted against debt for the leverage calculation, and whether they count in the following period's cash netting, are separate questions from how the cure is applied. Documents get this wrong often enough that it is worth reading rather than assuming.

Maintenance versus incurrence covenants

A maintenance covenant is tested on a schedule regardless of what the borrower does; failing it is an event of default. An incurrence covenant is tested only when the borrower proposes a specified action; failing it means the action is not permitted, with no default.

Field	Value
Maintenance	Quarterly test. Deterioration alone causes a default. Curable by an equity cure.
Incurrence	Tested at the moment of incurrence, on a pro forma basis. Deterioration alone causes nothing.
Pro forma testing	Incurrence tests are computed as if the transaction had occurred at the start of the LTM period, which imports the pro forma EBITDA of anything acquired with the proceeds

• An incurrence covenant is a leak-prevention device, not a monitoring device. It stops value leaving; it does not tell the lender that value is falling. A package with only incurrence tests gives the lender no scheduled moment at which the borrower must come and ask for something.
• Because incurrence tests are run pro forma for the transaction, a debt-financed acquisition can satisfy a leverage-based incurrence test that the group would fail on its actual results, provided the target's adjusted EBITDA is large enough. The test is passed by the transaction it is meant to constrain.
• The practical asymmetry is timing. A maintenance covenant hands the lender a negotiation while enterprise value still covers the debt. Every restructuring outcome depends more on when the conversation started than on the terms it started under.

Baskets and incremental debt capacity

Exceptions permitting debt, liens, investments, restricted payments, and asset transfers that the negative covenants otherwise prohibit. Capacity comes in three forms: fixed dollar amounts, grower amounts scaling with EBITDA or assets, and ratio-based capacity available only if a pro forma test is met.

Field	Value
Formula	Day-one incremental capacity = Free-and-clear (greater of fixed amount and x percent of EBITDA) + Ratio debt (L_ratio * EBITDA - existing debt of that ranking) + Available amount + reclassified capacity
Worked, free-and-clear	Greater of 40 and 100 percent of EBITDA (40) = 40, incurrable without meeting any ratio test
Worked, ratio debt	First-lien ratio test 4.00x, existing first lien 120, EBITDA 40: capacity = 4.00 * 40 - 120 = 40
Worked, combined	40 + 40 = 80 of day-one incremental. Pro forma total debt = 220 + 80 = 300, or 7.50x total leverage against a 6.00x maintenance covenant.
The gap	Incurrence capacity of 80 exists inside a document whose maintenance covenant permits only 35 of additional net debt. In a cov-lite structure with no maintenance test, nothing catches the difference.

• The worked example is the point: free-and-clear capacity plus ratio debt can exceed what the maintenance covenant would allow, and the two provisions are negotiated by different people at different points in the document. Aggregate the baskets and compare the total against the maintenance test before signing.
• Grower baskets scale with EBITDA, so a wide EBITDA definition enlarges every basket in the document at once. They also generally do not shrink: most are drafted as the greater of a fixed amount and a percentage, so a fall in EBITDA leaves the fixed floor intact.
• Reclassification permits capacity used under one basket to be re-designated to another once headroom reopens, freeing the original basket for reuse. Without an anti-reclassification provision, aggregate capacity is a stock that refills rather than a budget that depletes.
• The most consequential basket in a liability-management context is not a debt basket at all. It is investment capacity into unrestricted subsidiaries, because that is the pipe through which collateral leaves the credit group.

Source: LSTA model credit agreement provisions address incremental facilities, MFN, and basket construction.

Available amount (builder basket)

A cumulative, growing pool of capacity for restricted payments and investments, built primarily from retained net income from a build-up date, plus equity contributions and returns on prior investments, less amounts already used. Unlike a fixed basket it grows with performance.

Field	Value
Formula	Available amount = 50 percent of cumulative Consolidated Net Income from the build date (or a fixed starter if CNI is negative) + specified equity contributions + returns on and dispositions of investments previously made using the basket - amounts utilised
Worked	Cumulative CNI since the build date 60, so 50 percent = 30. Plus specified equity contributions 10. Plus 5 returned from a prior basket investment. Less 12 already used. Available = 30 + 10 + 5 - 12 = 33.
Conditions to use	Typically no event of default, and for restricted payments a pro forma leverage test; for investments, frequently no ratio condition at all

• The asymmetry between the restricted-payment condition and the investment condition is where value leaves. A builder basket that requires a leverage test for a dividend but not for an investment permits the same cash to be moved into an unrestricted subsidiary and then used from there.
• The starter amount matters more than the builder in the early years, because cumulative CNI is small or negative for a newly levered borrower. Documents commonly grant a fixed starter precisely so the basket is usable from day one, which makes the 50 percent CNI construction a description of the later years only.
• Consolidated Net Income is itself a defined term, adjusted in the same direction as Consolidated EBITDA. The builder grows on the adjusted figure, not the reported one.
• Returns on investments feeding back into the basket create a revolving pool: capital deployed under the basket, returned, and redeployed does not consume permanent capacity. Cumulative gross usage can far exceed the basket's stated size.

Unrestricted subsidiary designation

A designation removing a subsidiary from the credit group, so that its assets are no longer subject to the covenants, guarantees, or liens, and its results no longer count in the financial definitions. It is the mechanism on which the J. Crew and drop-down transactions depend.

Field	Value
Formula	Designation consumes investment capacity equal to the fair market value of the designated subsidiary's assets, charged against the general investment basket, the available amount, or a ratio-based investment basket
Capacity required	Sum of the general investment basket, the available amount, ratio investment capacity, and any similar-business or joint-venture basket, measured at fair market value at the time of designation
Worked	IP with a fair value of 45 transferred using a 20 general investment basket plus 33 of available amount: 53 of capacity against a 45 requirement, so it is permitted with 8 to spare
Consequence	The subsidiary's EBITDA leaves the covenant calculations while the debt stays, so leverage rises; documents therefore usually require pro forma covenant compliance after designation

• Designation is not a breach and does not require lender consent in most documents. It is a permitted use of investment capacity, which is why the response was to restrict the assets that can move rather than to prohibit the designation.
• The valuation of the transferred asset is usually determined by the borrower's board acting in good faith. Where the asset is intellectual property with no market comparables, the capacity consumed is effectively self-assessed.
• A borrower can transfer assets to a non-guarantor restricted subsidiary instead, which does not require investment capacity at all in many documents and achieves much of the same result. Restricting only unrestricted subsidiaries leaves that route open, which is what the Chewy-style transactions used.

Uptier exchanges and sacred rights

An uptier transaction is an amendment, passed by the required lender majority, permitting new super-priority debt, followed by an exchange in which the consenting majority moves its own holdings into that priority position and leaves the non-consenting minority subordinated. Sacred rights are the provisions requiring more than a majority to amend.

Field	Value
Formula	A group holding just over the required-lender threshold - commonly more than 50 percent of the outstanding loans - can amend everything that is not a sacred right, including provisions protecting the other side of that threshold
Conventional sacred rights	Reductions in principal or interest, extensions of maturity, and changes to the pro rata sharing and payment waterfall provisions typically require all-lender or each-affected-lender consent
The gap the transactions used	Lien subordination and the release of collateral were commonly amendable by majority, and the 'open market purchase' exception to pro rata sharing permitted a non-pro-rata buyback
Drafting response	Add lien subordination and any non-pro-rata debt purchase or exchange to the sacred rights; delete or define the open market purchase exception; require each-affected-lender consent for anything altering relative priority

• The arithmetic is the whole vulnerability: 50.1 percent of a class can act against 49.9 percent of the same class, and the sponsor chooses which lenders to invite into the majority. Position size relative to the required-lender threshold is therefore a risk factor independent of credit quality.
• This is why direct lenders holding a whole facility, or a club small enough to act together, price a genuine structural advantage over a broadly syndicated position of the same seniority. It is not a liquidity premium; it is a control premium.
• The litigation has not settled a single national rule. Courts have reached different conclusions on whether an open-market-purchase exception covers a negotiated exchange and on the scope of the implied covenant of good faith. Protection has to come from the drafting, not from an expectation of how a court will read a document.
• A minority-protection provision is only useful if the holder can monitor it. Knowing the composition of the lender group, and whether a cooperation agreement is forming, is the practical defence.

J. Crew blocker and dropdown protections

Drafting provisions preventing the transfer of material assets - characteristically intellectual property - out of the credit group where the lenders' liens and guarantees reach them. The name refers to the 2016 transaction that transferred brand IP to an unrestricted subsidiary using investment-basket capacity.

Field	Value
J. Crew blocker	Express prohibition on transferring material intellectual property, or specified assets, to any unrestricted subsidiary or non-guarantor, regardless of available basket capacity
Guarantor coverage test	A requirement that guarantors represent a minimum percentage of consolidated EBITDA and assets, tested and cured by adding guarantors
Anti-dropdown	Prohibition on transfers to non-guarantor restricted subsidiaries, closing the route that does not require investment capacity
Designation conditions	No designation of an unrestricted subsidiary while a default exists, or where the available amount would be used, or without pro forma covenant compliance

• A J. Crew blocker that covers only unrestricted subsidiaries is incomplete, because the same assets can move to a restricted non-guarantor. The blocker and the guarantor coverage test have to work together; either alone leaves a route open.
• The definition of 'material intellectual property' does the work. Where it is limited to registered trademarks, a transfer of the operating know-how, customer data, or a licence over the same brand can achieve the economic result without touching the covered asset.
• The strongest version is not a prohibition on transfers but a requirement that any asset leaving the group either remain subject to the lien or be sold for cash consideration at fair value applied to prepay the loans. That converts an asset-stripping route into a mandatory prepayment.
• Blockers are now common in newly drafted documents and absent from many outstanding ones. For a secondary purchase, the presence of a blocker is a credit term, and its absence is a discount.

Excess cash flow sweep

A mandatory prepayment of a percentage of the borrower's excess cash flow for each fiscal year, with the percentage stepping down as leverage falls and with deductions for voluntary prepayments and permitted capital expenditure already made.

Field	Value
Formula	Sweep = s(L) * ECF - voluntary prepayments during the period, where ECF is broadly EBITDA less cash interest, cash taxes, capex, scheduled amortisation and working capital increases, and s(L) steps down with net leverage
Worked	ECF 12, sweep rate 50 percent at leverage above 5.00x: sweep = 6.00, less 2.00 of voluntary prepayments already made = 4.00 payable
Step-down	50 percent above 5.00x, 25 percent between 4.00x and 5.00x, zero below 4.00x. At 4.50x leverage the same 12 of ECF sweeps 3.00 rather than 6.00.
De-levering effect	A 4.00 sweep on 220 of debt with 40 of EBITDA takes net leverage from 5.125x to (205 - 4)/40 = 5.025x, a tenth of a turn

• The sweep is the only contractual mechanism by which a bullet term loan amortises in a good year, and the step-downs are calibrated so that it stops exactly when the credit is improving. As a de-levering tool it delivers a fraction of a turn a year at best.
• Deducting voluntary prepayments from the sweep means a borrower can pay early at a moment of its choosing and receive full credit against a payment it would have owed anyway. Where the deduction is dollar-for-dollar with no discount, the sweep is optional in timing if not in amount.
• ECF definitions include a working capital adjustment, which makes the sweep procyclical in the wrong direction: a growing borrower consuming working capital reports low ECF and sweeps little, while a shrinking borrower releasing working capital reports high ECF and sweeps into a declining business.
• Check whether the sweep is applied pro rata across tranches or can be directed. In a structure with a first-out and last-out split, who receives the sweep is worth more than its size.