Borrowing Base Verification Automation for Asset-Based Lenders
We build the checks that sit under the certificate, then place the credit and collateral people who run them.
Borrowing base verification automation checks the collateral a borrower reports against evidence from outside the borrower’s own file, so ineligibles, cross-aging and concentration limits get tested instead of retyped. KORE1 builds that layer for asset-based lenders and factors, and staffs the credit side that runs it afterward. Our twelve-month retention on placements is 92%.
Last updated: September 21, 2026
A borrowing base certificate is a form the borrower signs, weekly or monthly, telling you how much it can draw. Gross receivables at the top. Deductions down the middle. One number at the bottom that your funding desk acts on that morning. Every week, same form.
Here’s the part nobody says out loud. Almost every line on that form came out of one file the borrower sent you. The aging, the ineligible math, the concentration test, the reserve. One export, transformed eleven ways, arriving as fourteen lines of apparent precision. Fourteen lines. One source. That’s a control on paper and data entry in practice, and we handle it out of KORE1’s accounting and finance consulting group, because fixing it takes credit judgment and data plumbing in the same room.
What One Borrowing Base Certificate Actually Proves
A worked illustration on a $40 million receivables book, built from advance rates and ineligible categories the OCC publishes in its Asset-Based Lending booklet. Not a client’s file. The arithmetic ties, and the right-hand column is the whole argument.
Availability, from gross receivables down to what the borrower can actually draw
- Gross accounts receivable 40,000,000 Borrower file
- Less past due beyond terms (3,120,000) Borrower file
- Less cross-aged balances (1,860,000) Second borrower file
- Less over-concentration above the 20% cap (2,400,000) Second borrower file
- Less foreign receivables without cover (1,450,000) Second borrower file
- Less contra accounts and offsets (980,000) Second borrower file
- Less government receivables not assigned (640,000) Second borrower file
- Less intercompany and affiliate balances (550,000) Second borrower file
- Eligible accounts receivable 29,000,000 Derived
- Advance rate applied, 85% 24,650,000 Derived
- Less dilution reserve, 5% of eligible (1,450,000) Second borrower file
- Availability before outstandings 23,200,000 Derived
- Less revolver and letters of credit outstanding (19,400,000) Lender’s own system
- Net availability 3,800,000 Derived
- Borrower file, or math on it. One export, and everything computed from it.
- Needs a second borrower system. The AP ledger, the customer master, the contract file. Usually not attached.
- Confirmed outside the borrower. On most certificates this is one line, and it’s the money you already lent.
Count the deep marks. On a form built to protect $23 million of exposure, exactly one figure came from somewhere the borrower doesn’t control, and it’s the balance you were already carrying. Everything above it is the borrower’s arithmetic, checked by your analyst’s arithmetic. Verification automation moves lines down that key. That’s the entire product.
What Asset-Based Lending Software Verifies, and What It Only Records
Most asset-based lending software is very good at the part that was never the risk. It takes the certificate, applies your ineligible rules, tracks availability, posts the advance, and keeps an audit trail of who approved what. All of that is real work and worth paying for. Genuinely.
None of it is verification. A platform that recomputes the borrower’s numbers faster is still reading the borrower’s numbers. Speed up an unverified input and you get a wrong answer sooner. Faster, not safer.
The regulator is blunt about which risk matters. The OCC calls ABL “particularly susceptible to borrower fraud,” says plainly that “fraud is a frequent cause of loss in ABL,” and lists the classic patterns, among them a borrower who can “use the same receivables as collateral to obtain financing from more than one bank.” No amount of certificate processing catches that one, because the evidence that would catch it isn’t on the certificate. It never was.
First Brands is the version everyone in commercial finance read about. The automotive parts group had roughly $5 billion in annual sales and layered factoring, supply-chain finance and inventory facilities on top of a conventional funded-debt stack, and the numbers never reconciled to a single view. Judge Christopher M. Lopez converted the remaining cases to chapter 7 on September 1, 2026. Writing two days later, restructuring lawyer Shane G. Ramsey put it in one line that belongs on a wall in every collateral department. First Brands, he wrote, is “a warning about collateral no one independently verified”.
Independently. That word is doing all the work.
The field exam is the original verification layer. It is also the one your team runs twice a year, on a sample, three weeks after the numbers mattered.
Where the Thresholds Come From
- 70–85% Common advance rate on eligible receivables OCC, Asset-Based Lending booklet, version 1.1
- 10–20% Suggested ceiling for any one concentrated account OCC, Asset-Based Lending booklet, version 1.1
- 5% Dilution most lenders expect, or less OCC, Asset-Based Lending booklet, version 1.1
- 92% KORE1 twelve-month retention on placements KORE1 placement data, trailing twelve months
The thresholds aren’t the hard part. Any credit shop can recite them. The hard part is proving that a given invoice belongs on the eligible side of one, and the published guidance is silent on that by design, because it’s your job. Nobody publishes that part.
Four Places the Review Stops Being Arithmetic
Each of these is a question an aging report cannot answer about itself. They are also, in our experience, the four that eat a collateral analyst’s week.
-
Contra
The offset that isn’t in the aging
A customer who also sells to your borrower can set off what it owes. The receivable looks clean. The exposure is net. Finding it means reading the accounts payable ledger, which is a different system and rarely part of the monthly package.
-
Identity
One customer, three spellings
Concentration is tested per obligor, and obligors show up as a parent, a division and a misspelling. Names are messy. Resolve them properly and a book that passed a 20% cap sometimes doesn’t. We have watched that single fix move availability by seven figures.
-
Assignment
Government work, unassigned
Paperwork decides this one. Receivables owed by a federal agency need the Assignment of Claims Act filing actually completed before they are worth lending against, and that fact lives in a contract folder. It does not live anywhere in an aging export.
-
Liens
Collateral pledged twice
The OCC names double-pledging outright as a fraud pattern. Catching it needs UCC searches run on a schedule and reconciled to the collateral you think you have. Nothing inside the borrower’s file will ever tell you. Not once.
Exceptions need an owner with authority to hold availability. A queue with no name attached is just a slower version of approving everything.
Who Runs the Verified Certificate Once We’re Gone
A verification layer creates exceptions. That is the point, and it is also the thing that kills these builds when nobody planned for it. Somebody has to work the queue on funding mornings, decide what gets reserved against, and say no to a borrower who has a payroll run at noon. That call is nobody’s favorite.
Two seats usually carry it. A collateral or credit analyst who lives in the exceptions daily, and a portfolio manager who owns the reserve decisions and the conversation with the borrower. Sometimes that is one person at a smaller shop. It is never zero people.
We staff both. Credit analyst staffing covers the daily desk, on contract while a build stabilizes or as a direct hire once the shape of the role is settled, and our finance recruiters handle the portfolio and credit leadership above it. KORE1 has been placing finance and accounting people since 2005, and our recruiters average more than fifteen years on their desks. When the certificate is one symptom of a close that runs long, the month-end close replay diagnostic shows where the rest of the month goes.
Build it, then seat it. In that order.
How a Borrowing Base Verification Build Runs
Five steps, and the first one produces something you can argue with before you have spent anything meaningful.
-
1
Replay a certificate you already funded
Pick one borrower and one quarter that is closed. We rebuild those certificates under written rules and compare line by line against what your team actually advanced. Disagreements are the deliverable.
-
2
Write the ineligible rules down
Most shops have the rules in a credit agreement, in a spreadsheet and in one person’s head, and the three disagree in ways nobody notices until a borrower’s auditor asks which definition of past due the facility actually uses and two people in the room answer differently. We reconcile them into one tested definition per ineligible category, with the agreement section cited beside each.
-
3
Connect the second source
This is where the deep marks come from. Customer master for identity and concentration, AP ledger for contras, contract data for assignment status, UCC searches for competing liens. One source at a time, each one earning its place by changing an answer.
-
4
Score the exceptions
An exception that fires on every certificate is noise and gets tuned or retired. Alert fatigue is real. We score each rule on how often it fired, how often a human agreed, and what it would have cost you to miss, then keep the ones that earn their interruption.
-
5
Name the owners
Before go-live, two names on paper. Who clears exceptions each funding morning and who books a reserve when a line cannot be verified. If those seats are empty, we recruit them, and that is usually the honest reason a build like this lands or stalls. Empty seats stall builds.
Common Questions
What is borrowing base verification automation?
It is the layer that tests a borrowing base against evidence the borrower does not control, rather than recomputing what the borrower submitted. Ineligibles, cross-aging, concentration and lien status all get checked against a second source. Your analyst then works exceptions instead of keying totals, which is a different job with a different hire behind it.
Is this asset-based lending software, or something else?
Something else, and it sits next to the software you already run. Asset-based lending software processes, calculates and records the certificate. We build the verification and exception logic underneath it, against whatever platform you use. If your platform can host the rules, we build them there. If it cannot, they run alongside and feed it.
We already run a full ABL platform. What would KORE1 add?
Usually the second sources and the exception discipline. Platforms ship with the ineligible math and expect you to supply the evidence, so most implementations quietly run on a single borrower export. We connect the customer master, the AP ledger and the UCC searches, then prove with a replay whether any of it changes a funding decision.
How long before a lender sees anything?
About three to four weeks for the first replay on one borrower. That timing assumes we get a closed quarter of certificates and the credit agreement. Wider rollout follows whichever rules earned their keep in the replay. Messy packages go last.
Can this replace the field exam?
No, and any vendor who says otherwise is selling you a problem. The OCC treats regular monitoring and timely field audits as the best deterrents to fraud-related losses, and a field examiner does things software cannot, including physical inspection and written account verifications. What this does is stop the exam from being the only moment anything gets confirmed. Between exams, the certificate carries the weight, and right now it mostly carries it on trust. Mostly.
Our borrowers send a PDF certificate and nothing else. Does that kill it?
It slows the first borrower and changes nothing structural. A PDF is a parsing problem, and a solvable one, though the fields that go wrong are specific and worth knowing about before you start. We wrote up the three fields a machine misreads on a borrowing base certificate for exactly that reason. The bigger constraint is usually the second source, not the first document.
Can KORE1 staff the collateral analysts too?
Yes, and that is how most of these engagements end. We place credit and collateral analysts, portfolio managers and field exam staff on contract, contract-to-hire or direct hire across more than thirty U.S. metros. The build tells you which seat you need before you post the role, which is a better order than the one most lenders use.
Bring One Certificate You Already Funded
Pick a borrower and a closed quarter. We will tell you straight whether a replay would find anything worth the cost, and if the answer is no, that is a short call and a free one. Either way you learn something.
Start With One Certificate →Curious what the form looks like before you dig one out? The NCUA publishes a sample borrowing base certificate in its examiner guides.
