Key Takeaways
- Square Loans has held loss rates below its 4% target for over a decade, proving that disciplined cash flow verification at scale is achievable.
- Platform lenders like Square succeed because they own first-party transaction data, a structural advantage independent MCA funders must compensate for with stronger bank verification workflows.
- Independent funders who adopt asynchronous, AI-guided bank verification can close the data gap without needing embedded platform access.
- The growing gulf between platform loss rates and independent funder loss rates is a verification problem, not a credit model problem.
A Decade of Sub-4% Losses Demands Attention
When a lender beats its own loss target for a single year, it could be luck. When it does so for more than ten consecutive years, it represents something structural. Square Loans, now the largest originator of online small business loans tracked by deBanked, has consistently held losses below 4% since launching in 2014. For independent MCA funders watching from the sidelines, this track record raises an uncomfortable question: what MCA underwriting best practices are platform lenders executing that the rest of the industry is not?
The answer is not a secret algorithm or a proprietary credit score. It is access to real-time, first-party cash flow data. Square sees every card swipe, every deposit, every refund flowing through its payment ecosystem before a merchant ever applies for capital. That level of visibility makes fraud nearly impossible and credit decisions remarkably precise.
Independent MCA funders do not have this luxury. They rely on bank statements, broker-submitted documents, and scheduled verification calls to approximate the same picture. The result is a persistent gap in loss performance that widens as platform lenders scale. This article breaks down how that gap forms, why traditional verification methods fail to close it, and what independent funders can do right now to bring their underwriting closer to platform-level discipline.
Why Platform Lenders Win on Loss Rates
First-Party Transaction Data Is the Moat
Square does not ask merchants to upload bank statements. It does not schedule verification calls. It does not rely on third-party aggregators to pull transaction histories. Instead, every dollar a merchant processes through Square's point-of-sale system feeds directly into the underwriting model. This creates a closed loop where the same data used to approve a loan is the same data used to collect repayment.
The implications for loss rates are profound. When a lender can see six months of daily transaction volumes in real time, it can detect revenue declines before a merchant even recognizes the trend. Seasonal dips, customer concentration, and spending anomalies all surface naturally in the data. There is no opportunity for a merchant to present a curated version of their finances, because the lender already has the unfiltered version.
According to deBanked's November 2026 reporting, Square's leadership has emphasized that this consistency is not accidental. It is the direct result of underwriting against verified, real-time cash flow rather than static documents.
The Independent Funder's Structural Disadvantage
Independent MCA funders operate in a fundamentally different environment. They receive applications through brokers, collect bank statements as PDF uploads, and attempt to verify transaction authenticity through manual review or scheduled phone calls. Each step introduces friction, delay, and the possibility of manipulation.
Consider the typical workflow. A broker submits an application with three months of bank statements. An underwriter reviews the PDFs, checks for obvious signs of tampering, and schedules a live verification call with the merchant. The merchant walks through their banking portal while the underwriter watches. If the call gets rescheduled twice due to time zone conflicts, three to five business days evaporate before the funder even confirms whether the statements are real.
During that window, the merchant may have taken a second advance from a competing funder. The bank statements may have been generated using increasingly sophisticated document fabrication tools. As one industry analysis noted, generative AI tools can now produce convincing fake bank statements with a single prompt, raising the stakes for any funder still relying on static document review as a primary verification method.
This is the gap. Platform lenders verify continuously and passively. Independent funders verify intermittently and actively. The loss rate difference is the cost of that asymmetry.
How Independent Funders Close the Verification Gap
Asynchronous Visual Verification as a Bridge
Independent funders cannot replicate Square's embedded data access. But they can dramatically improve the quality and speed of their verification by shifting from scheduled calls to asynchronous, recorded banking sessions. Instead of coordinating a live call where an underwriter watches a merchant scroll through their bank portal, the merchant records the session on their own time. The underwriter reviews it when ready.
This approach solves three problems simultaneously. First, it eliminates scheduling delays. A merchant in Vancouver can record their RBC portal at 9 PM Pacific without waiting for an underwriter in Toronto to be available. Second, it creates a permanent, timestamped record of exactly what the merchant's banking portal showed at the time of application. Third, it allows AI-guided coaching to ensure the merchant shows the right accounts, date ranges, and transaction details without an underwriter directing them live.
Exact Balance was built specifically for this workflow. Applicants receive a secure link, record their banking session directly in the browser with no software installation required, and submit the recording for review. The platform's AI-guided recording coach walks merchants through each step, verifying completion in real time. Underwriters then review the recording, check the activity log, and mark the verification as complete, all from a single dashboard.
Catching What Static Documents Miss
The real power of visual verification is not just convenience. It is fraud resistance. A fabricated PDF bank statement can look perfect on paper. But a live screen recording of a merchant navigating their actual banking portal is extraordinarily difficult to fake. The underwriter sees the URL bar, the session behavior, the loading times, the navigation patterns. Any attempt to use a spoofed or cloned banking portal introduces visual artifacts that trained reviewers, and increasingly AI models, can detect.
We explored this dynamic in depth in our analysis of how MCA lenders use AI to detect fake banking sessions in screen recordings. The key insight is that visual verification captures contextual signals that no document analysis tool can replicate. Font rendering, DOM element behavior, SSL certificate indicators, and real-time balance updates all serve as implicit authentication factors.
For funders processing hundreds of verifications per month, this approach scales without adding headcount. The merchant does the work of recording. The AI does the work of guiding. The underwriter does the work of reviewing. Nobody sits on hold waiting for a call to connect.
Building the Audit Trail That Regulators Expect
Sub-4% loss rates are not just a credit quality metric. They are also a signal to capital markets partners, warehouse line investors, and regulators that a funder's underwriting process is rigorous and auditable. As the MCA industry faces increasing regulatory attention in 2026, from state-level disclosure laws to federal scrutiny of lending practices, the ability to produce a complete verification record for every funded deal is becoming a competitive requirement.
Every async recording creates a timestamped, encrypted artifact stored in the cloud. Paired with activity tracking that logs when links were opened, recordings started, and submissions completed, this produces the kind of compliance documentation that satisfies both internal audit teams and external stakeholders. We covered the growing importance of this capability in our piece on how MCA audit season exposes bank verification documentation gaps.
What This Looks Like in Practice
Imagine a mid-size Canadian MCA funder processing 150 deals per month. Under the old model, each deal requires a scheduled verification call averaging 20 minutes, plus 10 minutes of scheduling coordination. That is 75 hours per month of underwriter time devoted exclusively to verification calls, not counting no-shows, reschedules, and time zone conflicts. Realistically, the actual time cost is closer to 100 hours.
Under an async model, the same funder sends 150 verification requests via email. Merchants record at their convenience. Underwriters batch-review recordings during focused work blocks, spending an average of 8 to 12 minutes per recording. Total review time drops to roughly 25 to 30 hours per month. The remaining 70 hours can be redirected to deeper credit analysis, portfolio monitoring, or processing additional deal flow.
The loss rate impact compounds over time. Faster verification means shorter time-to-fund, which reduces the window for stacking. Visual verification means higher fraud detection rates, which eliminates losses from fabricated documents. Complete audit trails mean better data for refining credit models, which improves decision accuracy with every funding cycle.
Square achieves sub-4% losses because it never has to ask whether the data is real. Independent funders can approach that same confidence by making verification visual, asynchronous, and AI-assisted. The technology exists. The question is whether funders will adopt it before the performance gap becomes insurmountable.
Frequently Asked Questions
How do platform lenders like Square achieve such low loss rates on business loans?
Platform lenders achieve low loss rates primarily because they underwrite against first-party transaction data. Square sees every payment a merchant processes through its ecosystem, giving it a real-time, unfiltered view of revenue and cash flow. This eliminates the need to rely on self-reported financials or uploaded bank statements, which can be manipulated. The closed-loop model also allows automatic repayment through transaction withholding, further reducing default risk.
Can independent MCA funders match platform-level loss rates?
Independent funders are unlikely to match platform loss rates exactly, because they lack embedded transaction data access. However, they can significantly narrow the gap by upgrading their verification workflows. Shifting from scheduled live calls to asynchronous screen recordings of merchant banking sessions provides visual proof of cash flow that static documents cannot match. Combined with AI-guided recording workflows and comprehensive audit trails, this approach gives independent funders a verification quality level that compensates for the absence of first-party data.
What are the best MCA underwriting practices for bank verification?
The most effective MCA underwriting practices for bank verification in 2026 include requiring visual verification of live banking sessions rather than relying solely on uploaded PDFs, using asynchronous workflows that eliminate scheduling delays, deploying AI-guided recording tools that ensure merchants show the correct accounts and date ranges, and maintaining encrypted, timestamped recordings as permanent audit artifacts. These practices reduce fraud exposure, accelerate funding timelines, and create the compliance documentation that capital markets partners increasingly require.
How does async bank verification prevent MCA fraud?
Async bank verification prevents fraud by capturing a live screen recording of the merchant navigating their actual banking portal. Unlike static PDF bank statements, which can be fabricated with generative AI tools, a recorded browser session shows the real URL, real-time page loading behavior, SSL indicators, and dynamic balance updates. These contextual signals are extremely difficult to spoof. The recording also creates an immutable record that can be reviewed by multiple team members and retained for compliance purposes.
Conclusion
Square Loans' decade-long track record of sub-4% losses is not magic. It is the result of underwriting against verified, real-time cash flow data. Independent MCA funders cannot replicate that embedded data advantage, but they can close the verification gap by adopting asynchronous, AI-guided bank verification workflows that capture visual proof of merchant finances before funding.
The funders who treat verification as a strategic investment, rather than a scheduling chore, will be the ones whose loss rates improve and whose capital partners stay engaged. Exact Balance provides the infrastructure to make this shift: browser-based recordings, AI-guided coaching, encrypted storage, and a unified underwriter dashboard. Visit exactbalance.ca to see how async verification fits into your workflow.