Executive Summary: A UCC filing tells you someone else has a claim on the collateral. A court judgment tells you someone else already won an argument about getting paid. Read separately, each one is a fact. Read together, in order, with attention to how recently each landed, they become an argument about where a business is in its financing life cycle. This post is about interpretation, not search mechanics: what the combination of lien data and litigation data implies, and where that inference stops being reliable. Before going further, the coverage boundary needs stating plainly, because every reading below depends on it. Cobalt's UCC data covers roughly 10 to 11 states. Cobalt's court records cover New York State and Miami-Dade County, Florida. That is the entire footprint. A clean combined result inside those jurisdictions is a clean result inside those jurisdictions and nothing more. If you need the search-and-integration side rather than the interpretation side, the Court Records API integration quickstart covers that ground and this post will not repeat it.
Why Do UCC Filings and Court Records Mean Something Different Together Than Apart?
What does a UCC filing actually establish on its own?
A UCC-1 financing statement is a notice filing. It records that a secured party claims an interest in a debtor's collateral, and it is filed in a state office rather than a court.[1] The filing does not assert that the debtor is in trouble. It asserts that a creditor wanted priority, and priority in Article 9 generally runs to whoever filed or perfected first.[2]
That is why a UCC count in isolation is one of the most over-read numbers in small-business underwriting. Equipment finance leaves UCC filings. Bank revolvers leave UCC filings. Ordinary working-capital lines leave UCC filings. A business with four active filings and no other adverse signal is frequently just a normally financed operating company.
What does a court record establish that a UCC filing cannot?
A civil judgment records an outcome. Somebody sued, and the court resolved it in a way that created an obligation. That is a categorically different kind of fact than a notice filing. A judgment implies that an earlier informal remedy already failed, which is why judgment data reads as a lagging but high-confidence distress marker rather than a routine financing artifact.
Timing matters here in a way that is easy to miss. A judgment lien is generally subordinate to a security interest that was perfected before the judicial lien arose.[2] So the practical question is never only "does a judgment exist." It is "where does this judgment sit relative to every perfected filing on the same collateral," which is a question you can only answer with both data sets in front of you.
Why does the combination change the conclusion?
Because the two sources answer different halves of the same question. UCC data tells you how much of the collateral is already claimed. Court data tells you whether anyone has stopped negotiating and started collecting. A business can be heavily encumbered and completely current. A business can be unencumbered and losing a lawsuit that will wipe out a year of margin. Only the pair distinguishes those cases.
What Does Each Combination of Lien and Litigation Signal Actually Imply?
The useful mental model is a two-by-two, read as a set of hypotheses rather than as scores. None of these are verdicts. Each one tells an underwriter what question to ask next.
• Heavy UCC, no litigation. The most common profile for a normally leveraged, performing business. Multiple secured creditors extended credit and none of them has needed a court. Verify who the secured parties are and what collateral they claim, then move on.[1]
• Heavy UCC, recent judgments. The stacking-and-distress profile. Layered financing plus a creditor who has already escalated to judgment suggests cash flow is being split across positions faster than it can service them. This is the pattern most worth a manual review before funding.
• No UCC, active litigation. Frequently an asset-light services business in a dispute. Consultancies, agencies, staffing firms, and professional practices often carry no financeable collateral at all, so the absence of UCC filings says little, and the litigation may be commercial, employment-related, or contractual rather than credit-driven.
• No UCC, no litigation. The genuinely quiet file, with one caveat. Inside a limited coverage footprint, quiet can mean quiet or it can mean out-of-scope, and those look identical in a response payload.
• Light UCC, repeat litigation across years. A business that avoids secured borrowing but is repeatedly in court is telling you something about how it manages counterparty relationships, not about its balance sheet.
Two of these combinations are frequently misread in the same direction. Underwriters tend to over-penalize the first row and under-penalize the third. Heavy UCC with a clean court record is usually a functioning business with a normal capital structure. No UCC with an active suit is often the file that actually deserves a second look.
How should the heavy-UCC-plus-judgment case be handled differently?
Treat it as a sequencing question rather than a volume question. Pull the filing dates on every UCC record and the filing and disposition dates on every case, put them on a single timeline, and look at what happened in what order. Layered positions that all predate a quiet period read differently from layered positions that cluster in the six months before a judgment. Our companion analysis on how UCC data reveals loan stacking in MCA portfolios covers the stacking side of that read in more depth.
Why is the asset-light case the one teams miss?
Because the underwriting reflex is collateral-first. A firm with no equipment, no inventory, and no receivables pledged produces an empty UCC response that feels like a clean result. It is not a clean result. It is a null result from a data source that had nothing to describe. The court record is doing all the informational work in that file, and if the business sits outside New York or Miami-Dade, you have no court record either.
Why Does the Sequence of Filings Carry More Signal Than the Counts?
What does filing order reveal that a total cannot?
Counts compress time out of the picture, and time is where the signal lives. Three UCC filings spread across six years describe a business that has renewed financing on a normal cadence. Three UCC filings inside ninety days describe a business that went looking for money three times in one quarter. The count is identical. The implication is not remotely the same.
The same logic applies to litigation. A single judgment from 2019 that was satisfied is a historical fact. A judgment entered last month, still open, sitting behind two filings made this year, is an active situation.
Which timing patterns are worth flagging?
• Clustered UCC filings inside a single quarter. Compressed borrowing activity, which in the merchant cash advance context often indicates positions being layered rather than refinanced.
• A UCC filing dated immediately before a large draw request. Someone else may have gotten to the collateral first, and priority runs to the earlier filing or perfection.[2]
• A judgment entered while earlier UCC filings remain unterminated. The judgment creditor is standing behind perfected secured parties, which affects both recovery expectations and the borrower's practical cash position.
• Repeated litigation involving the same plaintiff. A relationship that has broken down more than once, which is a governance signal rather than a credit signal.
• A long quiet period following heavy prior activity. Usually benign, occasionally a business that has stopped being able to raise money at all.
• New filings after a dormant stretch of several years. Worth a call, because something changed and the filing record will not tell you what.
How far back should the window run?
Far enough to see a cycle, which in practice means several years rather than several months. Business bankruptcy filings have been climbing: filings rose 11 percent in the twelve months ending December 2025, with business filings up 7.1 percent over the same period, and the Administrative Office of the U.S. Courts reported a further 11.9 percent increase in the following reporting period.[3][4] A twelve-month lookback in a rising-distress environment will systematically understate how long a file has been deteriorating.
What Does a New UCC Filing Shortly After a Judgment Tell You?
Why is this the most actionable pattern in the set?
Because it is the one combination where the ordering itself is the finding. A judgment lands. Weeks later, a new UCC-1 appears. The most economical explanation is that the business raised emergency financing to satisfy or survive the judgment, pledging collateral it had previously kept unencumbered. That is a different situation from a business that was already leveraged and then lost a case.
The reason this reads so strongly is that it inverts the normal order. In ordinary financing, borrowing precedes trouble. Here, trouble precedes borrowing, and the borrowing is reactive. Neither data source shows that on its own. UCC data shows a new filing with no explanation. Court data shows a judgment with no follow-up. The sequence is only visible when both are on the same timeline.
What should an underwriter do with that finding?
Ask about it directly, and ask about pricing. Emergency financing taken under judgment pressure is rarely cheap, and its cost shows up in the merchant's daily cash position rather than in any filing. The public record gives you the question. Only the merchant, the bank statements, and the payoff letters give you the answer.
"If courts are cheap enough, then it's worth it to run on every application automatically." Yehudah Aron, Cucumber Capital
That instinct is the right one for this pattern specifically. Sequence-based signals are only detectable if you have both data sets on every file rather than pulling court records selectively after something else already raised a flag. Selective pulls find the cases you already suspected. Systematic pulls find the reversals.
Does this pattern generalize outside merchant cash advance?
Partially. In construction and equipment finance, the same sequence often reflects a mechanics lien dispute followed by a bridge facility, which is a normal industry rhythm rather than distress. Peter Chong at Highwire framed the construction pre-qualification question as "are there any open liens against them? Any litigations associated to this specific contractor?" In that context the two questions are asked together because the answers routinely interact, not because either is decisive.
Where Does This Reading Break Down, and What Are the Coverage Limits?
What exactly is covered, and what is not?
This is the part that has to be stated without softening. Cobalt's court records cover New York State and Miami-Dade County, Florida. Not federal courts. Not PACER. Not the other 49 states or the other 66 Florida counties. UCC coverage runs across roughly 10 to 11 states. Secretary of State entity data covers all 50 states, TIN and EIN verification covers all states, and OFAC screening is global, but the two data sets this post is about are the narrowest in the stack.
The coverage is demand-driven rather than aspirational. Roughly 80 percent of Cobalt's funder customers file judgments in exactly those two venues. New York is the center of alternative lending and hosts the state's electronic filing system for civil matters,[5][6] and Miami-Dade became the second hub after a large migration of funders to South Florida.[7] Broader jurisdictions are on the roadmap, including a planned human-assisted queue for unsupported venues, which would be slower and has not shipped.
Customers say this back to us plainly. Lara Hodgson at RoxWrite put the objection in one sentence: "Most of our clients are not in New York." Cameron Kelliher at Elementix was equally direct: "We'll stay away from the court stuff then. I wish the court stuff was rounded." Both reactions are reasonable. A two-jurisdiction court product is the wrong fit for a national portfolio with no geographic concentration.
What does a clean result actually mean?
It means no matching records were found in New York State or Miami-Dade County. It does not mean the business has no judgments. It does not mean the business has no liens outside the covered UCC states. Writing "clean" into a credit memo without the jurisdictional qualifier is how a coverage limit turns into a documented misrepresentation during an audit.
There is one place where the footprint is unusually strong. New York is the only state where SOS status, UCC filings, contractor licensing, and court records all overlap in Cobalt's coverage. For a New York-concentrated book, that single-state depth is the most complete picture available through one integration.
Why does the regulatory backdrop make public-record reading more important, not less?
Because the reporting-based visibility is shrinking. The CFPB's revised Section 1071 rule, finalized in May 2026, excludes merchant cash advances from the definition of a covered credit transaction, so MCA providers will not be collecting or reporting that data.[8] Where reporting regimes do not reach, the public record is what remains. UCC filings and court dockets are not a substitute for a reporting regime, but they are the durable, primary-source layer underneath it.
The enforcement record supports the same point. The New York Attorney General's Yellowstone Capital settlement produced a judgment in excess of $1 billion, cancelled more than $534 million in outstanding obligations for over 18,000 small businesses, and barred the operator from the industry.[9][10] The Federal Trade Commission separately obtained a $20.3 million judgment and a permanent industry ban against an MCA operator for deceptive practices and unlawful asset seizure.[11][12] Those outcomes are public court records. They are exactly the class of fact this data surfaces.
Ready to see what a combined lien and litigation read looks like on your own files? Book a demo and bring three merchants you already know the answer on.
How Do Underwriting Teams Get This Data Without Building Two Integrations?
What are the existing options?
Before naming any product, the honest landscape. Unicourt aggregates dockets across a wide footprint and is what many funders already use for litigation. LexisNexis and Wolters Kluwer sell deep public-record research with correspondingly deep pricing. PACER covers federal filings only and charges per page. CSC handles UCC filing and search at scale for secured lenders. And a large number of shops still run manual courthouse and portal searches, which is slow but authoritative.
Each solves part of the problem, through a different interface, on a different billing model. Max Weisz described the stitched-together reality of running roughly 500 files a day: "we run New York court separately and then we run Unicourt, and then we run UCC searches." Three tools, three result formats, three places for a timeline to fall apart. Anthony at CorFinGroup framed the same problem from the buying side: "I'm looking for one cost effective solution for KYC and KYB."
Where does Cobalt fit?
The differentiator is integration surface, not breadth. UCC data returns in the same call as Secretary of State data, billed as a separate credit. Court records run through a dedicated endpoint. Both come back as structured JSON through one API key, which means the timeline reconstruction described earlier is a join in your own system rather than a manual reconciliation across three vendors. The trade is explicit: narrower coverage, far less integration work.
Pricing is one credit per lookup, the same as an SOS lookup. Gate Rock Capital named the threshold where this stops being a specialty tool and starts being a default check: "where I would pay $4 a pull is when you have the state index on court search."
Court lookups are asynchronous only. A `callbackUrl` is required, there is no synchronous mode, and typical completion runs 30 to 120 seconds because the data is pulled live from the court site rather than served from a cache.
curl -G "https://apigateway.cobaltintelligence.com/courtCases" \
-H "x-api-key: $COBALT_API_KEY" \
--data-urlencode "businessName=Example Trucking LLC" \
--data-urlencode "jurisdiction=testNewYork" \
--data-urlencode "callbackUrl=https://your-app.example.com/webhooks/court-cases"The `testNewYork` and `testMiamiDade` jurisdictions run without consuming credits, which makes them the right way to build and test the callback handler before spending anything. Responses include judgment details, case number, case type and division, filing dates, and parties. Amounts appear where the underlying record includes them, which is not every record.
How Should an Underwriting Team Operationalize the Combined Read?
What belongs in the automated layer versus the manual layer?
Automate the retrieval and the timeline construction. Keep the interpretation manual, at least until you have enough labeled outcomes to know whether your reading of each quadrant is actually predictive on your book.
• Pull both sources on every file in covered jurisdictions. Selective pulling defeats sequence detection, because you cannot detect an unexpected order if you only look after something else already surprised you.
• Store filing dates as first-class fields, not as display strings. Every pattern in this post is a date comparison.
• Record the jurisdiction on every result, including null results. A null from an uncovered venue and a null from a covered venue are different facts and must not collapse into the same field.
• Route the heavy-UCC-plus-recent-judgment quadrant to manual review. That is the quadrant where the cost of a wrong automated call is highest.
• Flag any new UCC filing dated within 90 days after a judgment. This is the reversal pattern, and it is cheap to detect once both date sets are in the same table.
• Re-pull on renewal rather than trusting the original file. Delinquency and default rates in small business lending have been rising, and a file that read clean at origination may not read clean at renewal.[13]
How does this fit the wider verification stack?
As one layer, in sequence. Entity status first, then liens, then litigation, then tax identity: SOS, then UCC, then court records, then TIN. Each layer narrows the question the next one answers. Our writeup on UCC filings plus SOS status as a two-source lien verification covers the first two layers, and the pre-funding litigation check workflow covers the third.
The macro context argues for keeping all four. The Federal Reserve's 2026 Report on Employer Firms found that 77 percent of small employer firms reported rising costs, tariff-related cost increases, or both in the prior twelve months, and that expectations for revenue and employment growth fell to their lowest levels since 2020.[14][15] Merchant cash advance obligations also surface in bankruptcy proceedings in ways that turn on how the underlying agreement and UCC filing were structured.[16] None of that changes how to read a quadrant. It changes how often the harder quadrants show up.
What Should You Not Ask This Data To Do?
Why is there no risk score?
Because Cobalt does not return one. There is no fraud score, no risk score, and no decisioning output of any kind. What comes back is structured public-record data: filings, dates, parties, case numbers, and amounts where available. The interpretation described throughout this post is your model, applied to that data, and it should be validated against your own outcomes rather than adopted as received wisdom.
That distinction matters commercially and legally. A data source that hands you a number invites you to stop asking why. A data source that hands you records keeps the underwriting judgment where it belongs, with the underwriter and documented in the file.
What are the honest failure modes of this reading?
Name matching is the first one. Business names collide, DBAs proliferate, and a judgment against a similarly named entity in the same county is a false positive waiting to be written into a decline. Verify the entity through SOS data before attributing any case to it.
The second is availability: not every court record includes an amount, and a judgment without a stated amount cannot be sized from the record alone. The third is scope, stated three times in this post because it is the limit that gets forgotten. Two jurisdictions for courts, roughly 10 to 11 states for UCC.
.jpg){kind=small}
.png)