Asset-based lending fraud: 6 schemes that cost billions

This post in 30 seconds.

• The schemes: fake or nonexistent collateral, double-pledged assets, inflated valuations, fabricated paper, phantom inventory, and recycled evidence. Six patterns explain most ABL fraud losses.
• The scale: Arena Television left around £280 million owed to more than 55 lenders. Court documents say that of 8,196 financed assets, only 66 existed.
• The gap: borrowing base certificates are borrower-produced paper and field exams are periodic samples. Written for ABL credit and risk officers and field examiners who own what happens in between.

When Arena Television collapsed in late 2021, administrators went looking for the broadcast equipment that secured its loans. Court documents cited by TVBEurope say that of 8,196 assets with outstanding loans against them, only 66 existed.

More than 55 lenders had financed that equipment. Most of it was never there.

Four years later the pattern repeated on the other side of the Atlantic, twice in one month. Tricolor and First Brands both failed in September 2025 with billions in collateral that, on inspection, was pledged twice or never existed at all. This guide walks through the schemes behind those losses, the real numbers from each case, and the checks that catch each scheme, including the one check almost nobody runs between field exams.

If you run credit or risk at an asset-based lender, you already live with the uncomfortable truth behind all of this: the borrowing base certificate is the borrower's own document. We build guided visual evidence tools at VentaVid for exactly that verification gap, and if you want to see what asset-level evidence from the field looks like in practice, book a demo of Venta Capture, a product of VentaVid, and bring your ugliest collateral schedule.

Why asset-based lending keeps attracting fraud

ABL is built on a simple trade: the lender advances against collateral it rarely sees. Receivables, inventory, equipment, vehicles. The borrower reports what exists, and the advance rate is applied to that report.

The structural weakness is that the core document in the relationship, the borrowing base certificate, is produced by the borrower it is supposed to police. ABF Journal puts it plainly: asset-based lending can be particularly susceptible to fraud, especially when borrowers hit cash-flow trouble. The OCC thought the risk distinct enough to give ABL its own Comptroller's Handbook booklet, with collateral controls and monitoring systems among its core topics.

And the fraud rarely starts as a grand plan. A borrower facing a covenant breach pads one certificate "temporarily", the quarter doesn't improve, and the padding compounds. By the time Arena failed, the alleged misappropriation had run past £1.2 billion across more than 55 lenders, according to the liquidators' claim.

So the useful question isn't whether your borrowers are honest. It's which scheme your current controls would actually catch, and when.

The six schemes behind most ABL fraud losses

Every big ABL fraud case reads differently in the press. Underneath, the mechanics repeat. Six patterns cover nearly all of it.

1. Fake or nonexistent collateral

The bluntest scheme: borrow against assets that were never there. Arena borrowed against broadcast kit that largely did not exist. In a 2026 case reported by AltsWire, lenders including BlackRock's HPS allege a telecom executive secured over $500 million against receivables that were invented outright, with "many, if not all" of the pledged receivables alleged to be non-existent. The FBI opened a criminal investigation.

What makes this scheme survivable for years is that paper verifies paper. An invoice supports a receivable, a schedule supports an invoice, and nobody stands in front of the asset. The check that catches it is independent existence verification: serial-numbered, asset-by-asset evidence that the thing is physically where the schedule says it is. That is the job of equipment verification before financing, and it is the single control Arena's lenders would most wish they had run.

2. Double pledging the same collateral

One real asset, two or more lenders. The collateral exists, which is what makes this scheme harder to spot than pure fabrication: any single lender who checks will find a real machine, a real invoice, a real car.

The SEC's complaint against Tricolor's former executives includes a concrete example: the same 3,225 auto loans allegedly included as collateral in two separate securitizations. UCC and lien searches help, but they only surface interests another lender perfected and registered. An unregistered or deliberately concealed pledge never shows up.

Double pledging deserves its own playbook, and we've written a separate deep dive on double pledging collateral and how lenders detect it. For this guide, treat it as one scheme among six, not the whole story.

3. Inflated valuations and invoice inflation

Here the collateral is real but worth a fraction of its reported value. In the First Brands case, the federal indictment reported by ABL Advisor alleges invoice amounts were altered to make them appear more valuable, alongside invoices for transactions that never occurred.

Inventory versions of this scheme inflate unit counts, grade damaged stock as new, or value obsolete goods at cost. The catches are independent appraisal, margin analysis against industry norms, and visual condition evidence tied to identifiers, so the reviewer prices what is actually on the floor rather than what the spreadsheet says.

4. Two sets of books and fabricated paper

The supporting cast for every scheme above: false purchase orders, false invoices, manipulated shipping documents, and a second ledger kept for the lender. ABF Journal's fraud overview lists exactly this toolkit, two sets of books included.

Fabricated paper is a data problem, and it needs data controls: transaction-level reconciliation of collections, confirmation of receivables with the actual account debtors, and metadata checks on submitted documents. No camera fixes a forged invoice. What visual evidence adds here is a cross-check: when the paper says 40 pallets shipped and the warehouse capture from the same week shows 12, the paper has a problem.

5. Phantom inventory and empty-box warehouses

An old scheme that keeps working: a warehouse full of sealed boxes holding nothing, staged for exam day. The ABF Journal overview names it without irony, warehouses filled with empty boxes purportedly holding inventory.

Field exams are supposed to catch this, and sometimes they do. One ABF Journal case study describes a seafood borrower whose 2007 field exam could verify roughly $85,000 of more than $8.2 million in listed collateral. The honest reading cuts both ways: the exam worked, and it worked years into the fraud. Between exams, the boxes sit unexamined. Unannounced spot captures of specific racks, with the locations picked by the lender rather than the borrower, change that economics.

6. Recycled or stale evidence

The quiet scheme, and the one aimed directly at lenders who already ask for photos. The borrower resubmits last year's photos of the equipment, photographs an asset borrowed for the day, or sends images pulled from the internet. Ordinary photo requests cannot tell the difference, which is why "send us some pictures" is a control that mostly reassures the person who designed it.

Catching recycled evidence requires provenance: when the image was created, on which device, whether it was recorded in the session or uploaded from a gallery, and whether the same image or the same scene has been submitted before. That is a solvable problem (our guide to verifying customer-submitted photos covers the mechanics), but only if the capture process is built for it from the start.

The case wave: Arena, Tricolor, First Brands

Three cases now anchor every ABL fraud conversation, and each one maps to the schemes above.

A note on where these figures come from, because precision matters in a fraud post: they are drawn from court documents and liquidators' claims reported by TVBEurope and Broadcast, the SEC's complaint, the bankruptcy examiner's report as covered by GTR, and the federal indictment as covered by ABL Advisor. Where a figure is an allegation rather than a finding, it says so. Nothing here is our estimate.

Arena Television: the equipment that wasn't there

Arena, an outside-broadcast firm, collapsed in November 2021 with debts of around £280 million owed to more than 55 asset-based lenders. The fraud surfaced when an auditor tried to verify serial numbers on financed equipment. Kroll, the administrators, allege the directors misappropriated proceeds of asset-backed lending in excess of £1.2 billion, and a High Court default judgment of £100 million was entered against them.

The detail that should bother every credit officer: dozens of sophisticated lenders financed the same fictional kit for years, because each one relied on documentation instead of independently sighting serialized assets.

Tricolor: the loans pledged twice

Tricolor, a subprime auto lender, raised over $1.9 billion through asset-backed securities before its September 2025 bankruptcy. In August 2026 the SEC charged three former executives with a multi-year scheme to defraud investors by double pledging subprime auto loans, describing a collateral hole that grew to approximately $800 million. Two of the three entered plea deals in the parallel criminal case, as reported by Investment Executive.

Vehicles are serialized assets with VINs. The double pledging allegedly survived because collateral data moved between parties as spreadsheets, and nobody matched VINs across facilities against the physical cars.

First Brands: the invoices that matched nothing

First Brands, an auto-parts group, filed for Chapter 11 in September 2025. Forensic work by Alvarez & Marsal, described in a debtor filing and reported by GTR, found that roughly $2.5 billion of about $3.0 billion in invoices sold to factors could not be matched to actual inventory sales. The court-appointed examiner put lender losses at an estimated $2.3 billion. The former CEO disputes the allegations.

First Brands is the receivables-side cautionary tale: the collateral was paper from the start, so the fraud was a paper exercise too.

The controls stack: what catches what

No single control catches all six schemes, and every layer of the standard stack has a blind spot the recent cases exploited. It helps to be honest about each layer before adding a new one.

Paper controls: certificates, covenants, lien searches

Borrowing base certificates, compliance certificates, UCC and encumbrance searches. Cheap, scalable, and the foundation of every facility. Their blind spot is structural: certificates are produced by the borrower, and lien searches reveal only interests another lender perfected. First Brands' factors held mountains of compliant paper.

Periodic human controls: field exams and appraisals

Field exams test receivables against shipping documents and purchase orders, and inventory against receipt documents and test counts, with AR roll-forwards and margin analysis on top, as SFNet describes. A good examiner with the right sample finds real problems.

Three limits, none of them the examiner's fault. Exams are periodic, often quarterly or annual. They sample rather than count everything. And they are usually announced, which gives a motivated borrower weeks to stage the warehouse. The Ocean Fresh exam that found $85,000 of $8.2 million proves the method works and arrives late.

Continuous asset-level evidence: the layer almost nobody runs

Between the certificate and the exam sits a gap of months in which the lender sees nothing. Closing it does not require more site visits. It requires the borrower's own staff to show the collateral, in a way the lender can trust.

That is what Venta Capture is built for. The lender sends a secure, personal capture link. Someone already standing in the warehouse opens it on their phone, no app and no account, and follows the lender's own steps: scan the serial plate, photograph the asset from set angles, walk the rack with a short video, answer the conditional questions. The submission comes back as a structured evidence package, not a folder of loose photos.

The integrity layer is what separates it from "send us some pictures":

• Sealed on receipt. Every submission gets a SHA-256 fingerprint and a digital seal, with a downloadable manifest for external verification. Media recorded in the session carries a visible stamp with date and time.
• Independent timestamps. Each case receives a timestamp from an external timestamping authority, so "when did this evidence exist" is answered by a party outside both lender and borrower.
• Session-bound recording. A session code sits in the frame during recording, so recycled or pre-recorded material stands out. Gallery uploads are accepted where configured, but labelled with their provenance unverified and checked for signs they existed before.
• Signals, not verdicts. Device reuse, virtual-camera software, file-instead-of-recording and similar observations are scored against a versioned rule set and surfaced with their caveats. A flag is a reason for a human to look, never an accusation.
• A full audit trail. Who created the link, when it was opened, what arrived when, who reviewed it. Useful when a workout turns into litigation.

Here's the mapping that matters, scheme by scheme:

While writing this post I read the Arena, Tricolor and First Brands coverage back to back, and the same sentence kept surfacing in different words: nobody independently looked at the assets. The middle column is the reason the looking that did happen came late, since the borrower usually knows the exam date weeks in advance.

One practical note on cost: a capture link costs the borrower's warehouse manager a few minutes, not a site visit. Lenders can start with the free plan and build a first collateral-verification flow the same day.

What continuous visual evidence does and does not prove

A fraud post owes you the limits as much as the pitch, so here they are.

• It complements lien searches and field exams. It does not replace them. A sealed video proves what was in front of a camera at a point of receipt. It does not search registries, confirm receivables with account debtors, or reconcile collections.
• The timestamp proves receipt, not the event. Server-side receipt time says when the system received the submission, not when a machine was installed or damaged.
• Signals are observations, not findings. Multiple recording attempts are normal behaviour in themselves, and a device seen before can be a shared work phone. The product raises a flag for a human. A person decides.
• A genuine recording can still show a staged scene. Harder to manipulate does not mean impossible to deceive. Guided capture raises the cost and skill required to stage evidence; it does not make fraud impossible.
• Receivables fraud stays a data problem. First Brands would not have been caught by cameras. Invoice-level reconciliation and debtor confirmation are the controls there; visual evidence covers the hard-asset side of the book.

If your fraud exposure is concentrated in receivables, start with reconciliation. If it includes equipment, vehicles, or inventory, the between-exams visual gap is probably your cheapest unclosed hole.

How lenders put a capture step into the credit cycle

Lenders who add guided capture tend to use it at four moments, in rising order of suspicion:

1. Before funding. An existence and condition capture of serialized equipment, matched to the schedule, before the advance. The equipment verification workflow: identifiers, condition, surroundings.
2. Periodic condition captures. The lender sends capture requests on its own calendar, monthly or quarterly per asset class, so the portfolio produces a stream of sealed, dated evidence between field exams.
3. Exception-triggered. A missed payment, a covenant wobble, a dilution spike, or an anomaly in the borrowing base triggers a same-day capture request for the specific collateral in question.
4. Pre-renewal or pre-waiver. Before extending, amending, or waiving, the file gets refreshed evidence rather than a twelve-month-old exam.

Each submission lands as a case the credit team can review, assign, and escalate, with the full evidence package a workout lawyer can later rely on. Two captures of the same asset can be linked, so condition at funding and condition at default sit in one file.

Frequently asked questions

How common is fraud in asset-based lending?

Common enough that the OCC gives ABL its own handbook with collateral controls as a core topic, and 2025 alone produced two multi-billion-dollar US cases. Industry commentary has long noted that ABL can be particularly susceptible to fraud when borrowers hit cash-flow stress.

What are the biggest red flags of ABL collateral fraud?

Borrowing base growth that outruns revenue, dilution spikes, resistance to unannounced inspections, round-number invoices, one customer dominating receivables, and documents that cannot be traced to source systems. Many mirror the fraud red flags claims handlers watch for: the evidence arrives too neat, too fast, or too recycled.

Do field exams catch fraud?

Sometimes, and they remain essential. But exams are periodic, sample-based, and usually announced, so a motivated borrower can stage around them. The practical fix is shortening the time between independent looks at the collateral, not abandoning exams.

What is double pledging in lending?

Pledging the same asset or receivable as collateral to two or more lenders at once. The SEC's Tricolor complaint alleges the same 3,225 auto loans sat in two securitizations simultaneously. Lien searches catch only perfected interests, so identifier-level matching matters.

How can lenders verify collateral remotely?

With guided capture: the borrower's staff open a secure link on a phone and follow the lender's steps, scanning serial plates and filming the assets, with the submission sealed and independently timestamped on receipt. No app, no account, and no site visit for routine checks.

Can borrowers fake photo or video evidence?

Ordinary emailed photos are easy to recycle or stage, which is why they prove little. Purpose-built capture binds media to the session with in-frame codes, checks provenance, and flags anomalies as signals for a human reviewer. Harder to manipulate, not impossible, and no vendor should tell you otherwise.

What happened in the Arena Television fraud?

Arena collapsed in November 2021 owing around £280 million to more than 55 asset-based lenders, after an auditor tried to verify serial numbers on financed broadcast equipment. Court documents cited in trade coverage say only 66 of 8,196 financed assets existed. Liquidators allege misappropriation in excess of £1.2 billion.

Talk to us

If you run credit, risk, or field examination at an asset-based lender, the question this wave left behind is blunt: how long could a borrower in your book pledge assets nobody looks at? Between the certificate and the next exam, the honest answer is usually "months".

What it takes to try it.

• Free plan, no credit card. Build a collateral-verification flow and send your first capture link the same day.
• Nothing for the borrower to install. Their warehouse or yard staff open a secure, personal link on any phone.
• Stuck? Book a free setup call and we build your first flow together.

A real person on the VentaVid team sets up your first capture flow with you.

Start for free or book a 15-minute demo and bring one real collateral schedule. We'll build the capture flow for it live.