A concentrated customer is not advanced against at a lower rate. It is struck out of the base.

How does a borrowing base actually get calculated?

In two steps, and almost every borrower we meet has modelled only the second one. Step one converts the ledger into eligible collateral by striking out everything the lender will not lend against. Step two applies an advance rate to what survives. The advance rate is the number on the term sheet. Step one is where the availability goes.

The published advance rates are not the problem. Eligible receivables carry 80% to 85%. Inventory blends to about 50% of net orderly liquidation value, with finished goods at 50% to 60%, raw materials at 40% to 55% and work in progress usually excluded or advanced at 30% to 40%. Machinery and equipment runs 50% to 75% of orderly liquidation value (ABF Journal, 1 June 2026, and PeerSense Capital Advisory, 1 July 2026).

The exclusions in step one are facility-specific and they are numerous. Receivables past a stated ageing threshold. Receivables from related parties. Receivables subject to set-off, contra accounts or disputes. Receivables from obligors outside agreed jurisdictions. Receivables from any obligor that has itself gone past due beyond a threshold, which commonly takes out that obligor's whole balance rather than the late portion. Then the reserves: dilution for credit notes and returns, rent reserves in landlord-lien states, tax and payroll reserves, and anything else the lender ranks ahead of itself.

The concentration cap sits inside step one, and it is the one that surprises people, because it is the only exclusion triggered by a customer doing well rather than badly.

What is a concentration cap, and why does it exist?

It is a ceiling on how much of the eligible receivables pool any single obligor may represent. Above that ceiling, the excess is excluded. A cap does not reduce the advance rate on the concentrated obligor. It removes the excess from the calculation before any rate applies, which is a materially worse outcome and a commonly misread one.

The lender's logic is correlation rather than credit quality. An asset-based facility is underwritten on the assumption that a diversified pool of receivables behaves statistically: some go bad, dilution is predictable, and the advance rate carries the loss. A pool where one obligor is 40% of the balance is not a pool. It is one credit exposure with administrative overhead, and the lender is not being paid an unsecured spread to take it.

No source in the surveyed market publishes typical single-obligor cap levels by facility type, or dilution reserve conventions, as a market series. Practitioner commentary describes them, and the levels move with obligor credit quality, sector, ageing profile and whether credit insurance is in place. Anyone quoting a single market-standard cap number is quoting a house convention rather than published evidence, and this article does not offer one either.

What can be said with confidence is the shape of it. The cap is negotiated once, at signing, in a schedule most borrowers do not read closely. It then binds every month for the life of the facility, recalculated against a pool that changes with every large invoice. It is the only term in the document that gets tighter when a customer relationship grows.

How much availability does this actually cost?

More than the customer's share of the ledger, because the exclusion happens before the advance rate rather than after it. Work it at a stated illustrative cap, because as above no market series exists to draw one from.

Take $10.0m of gross receivables with a single obligor at 35% of the balance, and assume a facility that caps any one obligor at 20% of eligible receivables. Assume a further $0.8m struck out for ageing, related-party and disputed items. The concentrated obligor contributes $3.5m, of which roughly $1.5m sits above the cap and is excluded. Eligible receivables land near $7.7m rather than $10.0m, and at an 85% advance rate availability is about $6.5m against the $8.5m a borrower would model off the headline rate alone.

That is roughly $2.0m of availability that exists on the ledger and not on the certificate, and the whole of it traces to one commercial relationship. The arithmetic is ours, at stated assumptions, and every input except the advance rate is illustrative. The point is the sequence, not the number.

The same effect shows up market-wide in the gap between what is committed and what is drawn. Committed asset-based lines across the surveyed lender population stood at $310.4bn in the first quarter of 2026, up 2.2% year over year, with utilization at 39.1%, up 1.8 percentage points on the quarter (Secured Finance Network, Q1 2026 ABL Survey, released 1 July 2026). Some of that headroom is deliberate liquidity. Some of it is commitment that a borrowing base certificate will not support.

Why does it bite in the month you need it?

Because the certificate is recalculated on a fixed cadence against a pool that moves, so availability is at its lowest precisely when working capital is at its highest. A large shipment to the concentrated customer raises receivables, raises the concentration percentage, and pushes more of that obligor's balance above the cap. The business has more sales, more cash tied up, and less availability, in the same month.

Collateral conditions in 2026 push the same way. Inventory cost has risen roughly 15% on average on tariff effects, turns have slowed, and liquidation values have not followed cost upward, so the divergence between cost basis and net orderly liquidation value has become the defining borrowing base issue of the second half of the year (ABF Journal, Q2 2026). Cost basis is what funds the balance sheet. Liquidation value is what funds the facility.

Credit performance in the asset class has moved with it. Non-accruals across the surveyed asset-based population ran 1.38% in the first quarter of 2026, up 42 basis points year over year and roughly triple the 0.33% to 0.56% band that held from the second quarter of 2023 through the fourth quarter of 2024 (Secured Finance Network, Q1 2026, 1 July 2026). Lenders reading that series are tightening eligibility rather than headline pricing, because eligibility is where their control sits.

The consequence for planning is simple and unwelcome. Liquidity has to be sized off trough availability under the certificate, not off the commitment on the front page and not off an average month.

What can actually be changed before signing?

Four things, in descending order of how much they move the number. First, the cap itself, and specifically whether it is a single blanket percentage or a schedule that carves out named investment-grade obligors at a higher level. That negotiation is available at signing and effectively unavailable afterwards.

Second, credit insurance on the concentrated obligor. Where a lender will recognize an insured receivable at a higher cap or a higher advance rate, the premium is being compared against availability rather than against bad debt, which is a different and usually better trade than the one the finance team modelled.

Third, the ageing and dispute definitions, and the cross-ageing mechanic that takes out an obligor's whole balance when part of it goes past due. On a concentrated pool that clause is the difference between a bad month and a covenant conversation.

Fourth, the reporting cadence. A weekly certificate on a volatile pool gives the borrower more usable availability than a monthly one, because the lender is not pricing in a month of unobserved movement. It costs finance-team time, and on a concentrated book it usually pays for itself.

None of these change the underlying position, which is that one customer is carrying a disproportionate share of the business. That is an operating problem with an operating timeline, and it shows up in the credit facility years before it shows up in a sale process.