Key Takeaways
- MCA stacking fraud is accelerating as the industry scales, with platforms like Stripe Capital originating over 81,000 advances in 2025 alone, creating more opportunities for borrowers to stack undisclosed positions.
- Traditional bank statement review misses stacking signals because it relies on manual spotting of ACH debits, which is error-prone at scale.
- AI-powered bank verification can automatically flag multiple daily ACH debits, unusual merchant payment patterns, and inconsistencies that indicate undisclosed positions.
- A layered verification approach combining document integrity checks, transaction pattern analysis, and real-time application tracking is the most effective defense against stacking.
- Lenders that automate bank statement analysis reduce stacking losses and accelerate legitimate deals simultaneously.
MCA Stacking Fraud Is Growing Alongside Record Origination Volume
Understanding how to prevent MCA stacking fraud has never been more critical for funders and ISOs. The merchant cash advance market is expanding at an extraordinary pace. Stripe Capital alone originated 81,000 MCAs and business loans in 2025, while Square Loans pushed $7 billion in merchant funding through its platform last year. That kind of volume creates fertile ground for stacking, the practice where a merchant takes on multiple cash advances from different funders without disclosing existing positions.
Stacking is not a new problem, but the scale of it is changing. More funders in the market means more places for a borrower to hide open positions. More speed-driven underwriting means less time spent scrutinizing each bank statement page. And more competition for deals means pressure to fund fast, sometimes at the expense of thorough verification.
For lenders processing hundreds or thousands of applications per month, manual bank statement review simply cannot keep up. The signals are there in the transaction data, but finding them requires a systematic approach that goes beyond eyeballing PDFs. This article breaks down exactly how stacking works, why traditional review processes miss it, and what a modern, AI-assisted bank verification workflow looks like in 2026.
How MCA Stacking Actually Works
The Mechanics Behind Multiple Positions
Stacking happens when a merchant secures a second, third, or even fourth cash advance from different funders before paying off existing ones. The borrower may not disclose these positions on applications. In some cases, brokers actively shop deals across multiple funders simultaneously, knowing that the approval windows overlap just enough to fund before anyone catches the duplication.
The financial damage is straightforward. If a merchant's daily revenue can realistically support one ACH debit of $500, but three funders are each pulling $400, that merchant's cash flow collapses within weeks. Default rates on stacked positions are dramatically higher than on single-position advances because the merchant was never underwritten for the total debt load.
Why Manual Bank Statement Review Misses Stacking
A skilled underwriter reviewing three months of bank statements can spot ACH debits from known funders. But this approach has serious limitations at scale. First, not every funder uses a recognizable company name in ACH descriptions. Second, reviewing 90 days of transactions across multiple accounts for every application takes time that most teams simply do not have when they are processing high volumes. Third, merchants sometimes provide statements from secondary accounts that do not show existing MCA debits, while the primary operating account with visible debits is conveniently left out.
The problem compounds when documents arrive in fragments. An applicant sends pages one through three of a six-page statement. The underwriter reviews what they have, flags the file as incomplete, and waits. Meanwhile, the deal sits in limbo or, worse, gets pushed through with an incomplete picture. As we explored in our analysis of why MCA lenders lose deals to slow application intake, these bottlenecks do not just cost time; they cost revenue and create risk blind spots.
Using AI-Powered Bank Verification to Catch Stacking
Automated Transaction Classification
The first layer of defense against stacking is automated bank statement parsing that classifies every transaction, not just the ones that happen to catch a reviewer's eye. AI-powered extraction tools can ingest a full bank statement PDF and categorize each line item by type: ACH debits, wire transfers, point-of-sale deposits, NSF fees, and more. When that classification runs automatically, the system can flag patterns that matter for stacking detection.
Specifically, look for multiple recurring ACH debits in the $200 to $2,000 daily range. These are the hallmark of active MCA positions. Machine learning models trained on MCA transaction data can recognize funder-related debits even when the ACH description is vague or abbreviated. This is fundamentally different from a human reviewer scanning pages and hoping to notice something unusual.
Document Integrity and Completeness Checks
Stacking fraud often involves document manipulation. A merchant might alter statement pages to remove evidence of existing positions, or submit partial statements that skip the months where debits are most visible. AI document analysis can detect signs of tampering, such as inconsistent fonts within a statement, mismatched page numbering, gaps in date sequences, or metadata anomalies in the PDF itself.
Let's Submit addresses the completeness problem at the intake stage. When an applicant receives a secure upload link, they submit their full document set in one place. The platform's AI extraction automatically parses business information, financials, and bank statement data, making it immediately apparent when pages are missing or when date ranges do not cover the required period. This front-end completeness check means underwriters spend less time chasing documents and more time evaluating the data that actually matters.
Cross-Application Pattern Analysis
A sophisticated stacking operation does not just involve one merchant. Industry fraud cases, including the recently publicized prosecution of Saul Shalev for alleged systematic fraud in the small business finance space, remind us that organized fraud rings target multiple funders with coordinated applications. When your document intake and extraction pipeline feeds into a centralized dashboard, you gain the ability to spot patterns across applications: similar addresses, overlapping bank accounts, identical contact details, or the same broker submitting suspiciously similar deals in rapid succession.
This kind of cross-referencing is nearly impossible when applications live in scattered email threads and spreadsheets. It requires a structured data layer, which is exactly what building a scalable MCA application pipeline is designed to deliver.
Building a Layered Verification Workflow That Actually Stops Stacking
No single technology or process eliminates stacking fraud entirely. The most effective approach layers multiple checks throughout the application lifecycle.
Layer one: structured document collection. Eliminate the chaos of email attachments and partial submissions. Use a dedicated upload portal where applicants submit complete bank statements, applications, and supporting documents in one session. This ensures your team starts with a full picture rather than assembling fragments over days.
Layer two: automated extraction and flagging. AI parses every document upon upload. Bank statements get scanned for recurring ACH patterns, NSF frequency, average daily balance trends, and deposit consistency. Any application with three or more recurring ACH debits in the MCA-typical range gets automatically flagged for enhanced review. This does not replace human judgment; it directs human attention where it matters most.
Layer three: verification against application data. Cross-reference what the merchant disclosed on their application against what the bank statements actually show. If the application states "no existing positions" but the statements reveal multiple daily ACH debits to entities with funder-like names, that discrepancy triggers an escalation. Automated extraction makes this comparison instant rather than manual.
Layer four: ongoing portfolio monitoring. Stacking does not only happen at origination. A merchant may take on additional positions after you fund. While pre-funding verification is the primary defense, lenders increasingly use periodic bank statement re-verification to monitor active positions. This is where having an efficient, repeatable document intake process pays dividends beyond the initial deal.
The accounting mistakes that MCA companies commonly make early on, as recently outlined by industry accounting expert David Roitblat, often stem from poor visibility into portfolio risk. Stacking losses that go undetected at origination show up later as unexpected defaults that blow through reserve assumptions. Prevention at the verification stage is orders of magnitude cheaper than recovery after the fact.
What This Looks Like in Practice
Consider a mid-size funder processing 400 applications per month. Their underwriting team of four people manually reviews bank statements, spending an average of 20 minutes per file just on transaction scanning. That is over 130 hours monthly dedicated to a task that still misses stacking signals roughly 15 to 20 percent of the time, based on industry estimates of undetected stacking rates.
Now layer in an AI-powered workflow. Applications arrive through a secure upload link or forwarded email. Bank statements are parsed automatically, with transaction data extracted and categorized within minutes. The system flags any application showing three or more recurring ACH debits consistent with MCA repayments. Those flagged files go into an enhanced review queue, while clean applications proceed to standard underwriting.
The result is not about removing humans from the process. It is about ensuring that your team's expertise is applied to the files that actually need it. As we discussed in our piece on reducing manual data entry in MCA lending, every minute saved on routine extraction is a minute your team can spend on judgment calls that protect the portfolio.
Speed matters here too. The pressure to compete with platforms like Stripe Capital and Square Loans, which fund in hours rather than days, tempts lenders to cut corners on verification. But the goal should not be choosing between speed and safety. With the right intake and extraction tools, you achieve both. Let's Submit processes documents and extracts data in minutes, giving underwriters the structured information they need to make fast, informed decisions without skipping the checks that catch stacking.
Frequently Asked Questions
What is MCA stacking and why is it dangerous for lenders?
MCA stacking occurs when a merchant holds multiple active cash advances from different funders at the same time, often without disclosing existing positions. It is dangerous because the merchant's cash flow was never underwritten to support the combined debt load. Default rates on stacked positions are significantly higher than on single-position advances, and recovery is difficult because multiple funders are competing for the same daily revenue stream.
How can lenders detect stacking from bank statements?
The primary indicator of stacking in bank statements is multiple recurring ACH debits in amounts consistent with MCA repayments, typically between $200 and $2,000 per day. Lenders should also look for frequent NSF or overdraft fees, declining average daily balances over the statement period, and deposits that appear to come from other financing sources rather than business operations. AI-powered extraction tools can automatically flag these patterns across hundreds of statements, catching signals that manual review often misses.
Can AI detect altered or tampered bank statements?
Yes. AI document analysis can identify several signs of bank statement tampering, including inconsistent fonts or formatting within a document, irregular spacing, mismatched page numbers, gaps in transaction dates, and PDF metadata anomalies such as unexpected creation tools or modification timestamps. While no system catches 100 percent of sophisticated alterations, automated integrity checks significantly raise the bar compared to visual inspection alone.
What is the best way to prevent MCA stacking fraud?
The most effective approach is a layered verification workflow. Start with structured document collection through a secure portal so you receive complete, unaltered statements. Apply automated AI extraction to classify transactions and flag stacking indicators. Cross-reference disclosed positions against actual bank data. Finally, maintain the ability to re-verify bank statements on active positions periodically. Platforms like Let's Submit automate the collection and extraction layers, allowing your underwriting team to focus on analysis and decision-making.
Conclusion
Stacking fraud thrives in the gaps between speed and diligence. As MCA origination volume continues to surge across the industry, those gaps are widening for any lender still relying on manual bank statement review and fragmented document collection. The good news is that closing those gaps does not require slowing down. It requires smarter infrastructure.
AI-powered bank verification catches the transaction patterns, document anomalies, and data inconsistencies that stacking depends on. Combined with a structured intake process that ensures complete submissions from the start, lenders can protect their portfolios without sacrificing the speed their sales teams need to compete.
Let's Submit gives MCA lenders both sides of that equation: a streamlined applicant upload experience and AI-driven document extraction that surfaces the data underwriters need. Visit letssubmit.ca to see how async verification and intelligent extraction fit into your workflow.