Bank lending ratios: what a banking syndicate
really looks for in your accounts
before it lends

Every leverage ratio is built on EBITDA. Yet the EBITDA a credit analyst uses is almost never the figure in your tax return. Five adjustments are made as a matter of course.

Non-recurring items are stripped out. A capital gain on a disposal, an insurance pay-out, a lawsuit won, a one-off grant: all of it comes out. Conversely, a one-off restructuring cost is added back — but the analyst will be quicker to remove income than to add back a cost.

The owner-manager's pay is normalised. If you pay yourself EUR 50,000 where a salaried chief executive would cost EUR 180,000, the difference is deducted. This is the adjustment that surprises owners most, and the one that most often worsens the leverage of family- and owner-managed businesses.

Leasing is restated on both sides. Under French accounting rules (French GAAP), a finance lease does not appear as a liability on the balance sheet. The bank treats it as what it is in economic terms: debt. It adds the outstanding principal to borrowings and the depreciation component back to earnings.

Related-party transactions are scrutinised. Rent paid to the owner's property company (SCI), services invoiced by a holding company, purchases from a sister company: each is brought back to its market value.

Acquisition pro forma figures are capped. If you consolidate a company acquired part-way through the financial year, the analyst uses its annualised EBITDA but applies a haircut to the announced synergies — often half, sometimes all of them until they have actually been delivered.

This is the headline ratio — the one that decides whether an application stays within conventional bank lending or moves to other instruments. Thresholds observed in the French market, excluding leveraged buyout structures:

The often-quoted 4x threshold is really the ceiling of acceptability rather than the covenant. The standard leverage covenant set by a French banking syndicate is generally 3.5x, with a step-down profile: 3.5x at the year-end of Y+1, 3.0x in Y+2, 2.5x in Y+3. That step-down is the real point of negotiation — a flat covenant leaves you headroom that a stepped profile takes away year after year.

Leveraged structures follow different norms. On an SME leveraged buyout (LBO), the structure usually sits between 3.5x and 4.5x total debt, with an equity contribution of 30–50% of the purchase price. The levels reached before 2022 are rarely seen today.

Negotiate the definition before you negotiate the level. A 3.25x covenant on a narrow definition is more comfortable than a 3.75x covenant on a broad one.

Gearing compares net debt with equity. It answers a simple question: who is carrying the risk?

The usual benchmark is 1.0. Above 1.5 the application becomes tight; above 2.0 a conventional banking syndicate pulls out in most situations. A French variant deserves attention: the equity ratio (taux d'autonomie financière), i.e. equity divided by total assets. Below 20% it is a clear warning sign; between 25% and 30% the structure is considered sound.

This ratio explains a decision that owners often take badly: a refusal despite excellent profitability. A highly profitable company that pays out all of its earnings every year does not build up equity. It is asking its banks to carry a risk it does not carry itself. Three years of retained earnings before a major financing request are often worth more than three extra points of margin.

This is also where the treatment of shareholder loans is decided. Locked in by agreement and subordinated to the repayment of bank debt, they count as quasi-equity and improve gearing. Unrestricted, they are debt. A lock-in and subordination agreement is a one-page document that moves a ratio.

Leverage measures a stock of debt. The DSCR — debt service coverage ratio — measures your ability to meet the repayments. Experienced analysts favour it because companies do not default on a balance-sheet ratio: they default on a repayment date.

The numerator is built as follows: adjusted EBITDA, minus tax actually paid, minus the change in working capital, minus maintenance capex.

A DSCR covenant is usually set between 1.05 and 1.20, depending on how predictable the cash flows are. Project finance backed by contracted revenue can go as low as 1.05; a cyclical business will be asked for 1.30.

The analyst does not stop at the DSCR in your business plan. They run stress scenarios: revenue down 10–15%, margin down 100–200 basis points, a higher cost of debt. If the DSCR falls below 1 in a plausible downside scenario, the application is reworked or declined.

Get ahead of it. Presenting a quantified downside case yourself changes the nature of the discussion: you stop being the one making promises and become the one who has measured their own risk.

The ICR — interest cover ratio — divides EBITDA by net finance costs. It measures resilience to rising interest rates or a squeeze on margins. The benchmark is above 4x. The covenant is generally set between 2.5x and 3.0x. Below 2x the position is considered fragile: one poor quarter is enough to put cash under strain.

This ratio has regained an importance it lost during the decade of low interest rates. On floating-rate debt, the rise in rates since 2022 has mechanically worsened the ICR of companies whose operations had not changed at all. It also explains why interest rate hedging requirements — a cap or swap on part of the outstanding debt — are back in loan documentation.

A growing company that funds its working capital out of its own cash is heading for strain, however profitable it is. Three items are examined.

Debtor days (DSO), compared with your sector and, above all, with their own trend. DSO slipping from 45 to 62 days over two financial years signals either weaker customer quality or a sales policy that buys revenue with credit.

Stock turnover. Inventory growing faster than revenue prompts a direct question: overproduction, obsolescence, or forward buying?

Creditor days (DPO). An unusual lengthening is the most closely read signal of all. It suggests the company is funding itself from its suppliers — in other words, it has already reached its cash limit.

The summary ratio is working capital expressed in days of revenue, and its trend over three financial years. Working capital that rises faster than activity points to growth that is not self-financed. That finding does not block an application, but it steers it towards the right instruments: invoice finance and receivables finance or a revolving credit facility rather than an amortising term loan.

Some factors weigh as much as the ratios, and business owners often discover them too late.

The Banque de France rating. Every company above a certain level of activity receives a credit rating in the FIBEN database (the French central bank's company file), on a scale from 3++ to 9. Ratings from 3++ to 4 are considered sound; from 5+ access to credit becomes noticeably harder, and a rating of 6 or below closes most doors. You can obtain your rating free of charge on request. Checking it before you approach lenders avoids discovering a problem during the credit review — an error can be challenged, but correcting it takes time.

Payment incidents and preferential liens. A registered and published lien in favour of the French tax authorities or URSSAF (the social security contributions collector) is an almost absolute stop signal. A rejected direct debit, a bounced cheque, an unauthorised overdraft that has not been regularised: every event is visible to the whole banking syndicate. Three financial years without incident are an asset that few owners consciously value.

Concentration. A customer accounting for more than 25–30% of revenue changes the risk question: the bank is no longer financing your company, it is financing your relationship with that customer. The answer is not to deny the concentration but to document it — length of the relationship, multi-year contract, switching costs for the customer.

Consistency. Three years of stable results score better than one exceptional year preceded by two mediocre ones. A credit committee values predictability more than performance.

Quality of information. Accounts filed on time, interim accounts available, a monthly cash flow forecast, the ability to produce a figure within forty-eight hours. A company with tight financial control reassures lenders regardless of its ratios, because it will spot a difficulty before its banker does.

When the ratios are borderline, the security package becomes the negotiating lever.

A Bpifrance guarantee (from France's public investment bank) covers a share of the bank's risk — in the order of 40–70% depending on the scheme and the type of transaction. On a tight application it can tip the decision, because it changes neither your leverage nor your DSCR, but the lender's net exposure.

A pledge — over shares, the business goodwill (fonds de commerce), equipment or receivables — is systematically required on major financings. The negotiation is less about the principle than about its scope and the conditions for release.

The owner's personal guarantee is the most sensitive issue. It is negotiated on three dimensions: its amount, often capped at a fraction of the outstanding debt; its duration, with possible release after several years of covenant compliance; and its ranking, with the guarantee called only after the other security has been enforced.

Vendor financing, in an acquisition, serves two purposes: it reduces the senior debt requirement and keeps the seller at risk, which reassures the banking syndicate about the reliability of the information provided during due diligence.

The ratios determine the approval; the covenants determine what happens next. Business owners rarely read this part with the attention it deserves.

Financial covenants use the ratios above, most often tested half-yearly or annually on audited accounts. Three points to examine: the testing frequency, the step-down profile, and whether there is a cure right — the equity cure, which allows the shareholder to restore a ratio by injecting equity rather than face acceleration of the loan.

Non-financial undertakings are more restrictive than they look: a cap on annual capex, restrictions on dividends until a given leverage level is reached, a ban on disposing of assets above a threshold, a change-of-control clause, and a negative pledge preventing security from being granted to another creditor.

The consequences of a covenant breach must be known before signing: a mere duty to inform, a margin increase, suspension of drawdowns, or early repayment (acceleration). The difference between these regimes is considerable.

These bank lending ratios reflect practice observed in the French market for transactions involving SMEs and mid-sized companies. They vary by sector — a recurring-revenue model can support higher leverage than a cyclical business — by the company's maturity and by prevailing market conditions. They are neither a rule, nor a commitment, nor a guarantee that financing will be obtained.