Which Borrowers Need Continuous Secretary of State Monitoring? A Tiering Framework

Executive Summary: Continuous business monitoring is usually sold as a portfolio-wide product and is rarely worth buying that way. Most borrowers in most portfolios will never produce a Secretary of State change that matters, and monitoring all of them at the same interval spends the same on a fully amortized small balance as on a large secured facility with years left. The useful question is not whether to monitor. It is which borrowers earn which interval.

Why does uniform monitoring waste money in both directions?

Because it is simultaneously too much and too little, and the two errors are invisible to each other.

Applied uniformly, a single interval overspends on the stable majority of the book while underspending on the concentrated minority where changes actually carry consequences. Neither error announces itself. The overspend looks like a line item nobody questions. The underspend looks like nothing at all until a loss arrives, at which point it is attributed to the borrower rather than to the monitoring design.

The economics make tiering unusually easy to justify here, because the unit cost is small and known. Each completed check costs 1 credit from the same shared pool as the rest of the API suite,^[1] which means the cost of a monitoring policy is a straightforward multiplication of borrowers, frequency, and credit price. That is a number you can compute before committing, and most lenders who compute it discover the interval they were defending on cost grounds was never the expensive part.

The expensive part is analyst attention. Every alert consumes review time, and review time does not scale with credits. That is the real constraint tiering is managing.

• Credits are cheap and linear. Frequency costs what the arithmetic says it costs.

• Attention is expensive and finite. Alert volume competes for a fixed resource.

• Consequence is concentrated. A minority of the book carries most of the downside.

• Most records do not move. In any given interval, the majority of checks confirm nothing changed.

• Uniform intervals ignore all four facts. Which is why they feel simple and perform poorly.

What actually determines whether a borrower needs close monitoring?

Six variables do most of the work, and only two of them are about the borrower's creditworthiness.

Whether the exposure is secured. This is the single strongest determinant and it is structural rather than judgmental. Secured lenders are subject to a clock unsecured lenders are not: if a debtor's name changes so a filed financing statement becomes seriously misleading, the filing perfects collateral acquired "before, or within four months after" the change and stops perfecting later acquisitions unless an amendment is filed inside that window.^[2] Four months is shorter than a quarterly cycle can reliably catch, catching it is the secured party's responsibility rather than the debtor's,^[3] and the governing name is the one on the state's formation record.^[4]

Exposure size. The cost of a check is fixed. The cost of missing a change scales with the balance. That asymmetry alone justifies shorter intervals at the top of the book without any further analysis.

Remaining term. A facility with four years left has far more opportunity for something to change than one maturing next quarter. Monitoring intensity should decay as a loan amortizes toward zero.

The borrower's own filing history. The best available predictor, and the most underused. An entity that has previously lapsed, been administratively dissolved, or reinstated has demonstrated that its compliance processes fail. Entities that fail once tend to fail again, because the underlying causes are structural: a stale address the state cannot reach, no one owning the filing calendar, or a decision to defer small recurring obligations.

State of formation. Reporting cycles and reinstatement windows differ materially. Georgia caps reinstatement at five years from the effective date of dissolution,^[5] Florida permits application at any time,^[6] and Texas allows it indefinitely while limiting retroactive effect to three years.^[7] States with biennial rather than annual reports also leave officer data stale twice as long.

Whether anything has moved recently. Entity problems cluster in time. A borrower that produced a change last quarter is meaningfully more likely to produce another than one static for three years.

What does a workable tier structure look like?

Three tiers handle almost every portfolio. More than that becomes an administrative burden that nobody maintains.

Tier 1, short interval. Large secured exposures, long remaining terms, borrowers with prior lapse history, and anything that has produced a change recently. This is where the four-month clock binds and where balances justify the attention. A monthly interval gives comfortable margin against the UCC window; anything longer does not.

Tier 2, moderate interval. The general book. Secured but smaller, or unsecured but substantial. Quarterly is defensible here provided nothing in Tier 1's criteria applies, and provided the escalation rule below is running.

Tier 3, long interval. Small, unsecured, short remaining term, clean filing history. Semi-annual or annual is genuinely appropriate. This is the tier most often over-monitored out of a general sense that more is safer.

Two rules matter more than the tier definitions themselves:

• Escalation is automatic and behavior-driven. Any borrower producing a change moves up a tier until it goes quiet for a defined period. This costs nothing to implement and concentrates attention where something is already happening, which outperforms static attribute-based scoring.

• Monitoring stops at payoff. Paying to check records for closed accounts is a common and invisible waste. Removing them is the cheapest efficiency available.

The best predictor of whether a borrower will produce an entity change in the next ninety days is whether it produced one in the last ninety. A tiering model that adjusts on observed behavior beats one built purely on static attributes, and it requires no model maintenance.

Which borrowers genuinely do not need this?

Worth stating plainly, because a framework that recommends monitoring everything is not a framework.

• Fully amortized exposures approaching maturity. Limited remaining opportunity for a change to matter.

• Small unsecured balances with clean histories. The cost of the response process exceeds the exposure at risk.

• Borrowers where you hold no enforceable position. If a change would not alter what you can do, detecting it faster has no value.

• Closed and paid-off accounts. No exposure, no reason to check.

There is also a category monitoring cannot help with regardless of interval, and it is worth knowing before building a policy around it. The Secretary of State record does not capture ownership changes for corporations or LLCs. Texas states directly that "there is no filing requirement with the secretary of state when there is an ownership change" for either entity type.^[8] No cadence detects an event that is never filed. If your monitoring policy is justified primarily on catching control changes in real time, it will underperform that expectation no matter how frequently it runs.

Want to see how configurable Secretary of State monitoring fits a tiered portfolio policy? Book a demo.

What does the tiering look like applied to a real book?

An abstract framework is easy to agree with and hard to act on. Working it through on a plausible portfolio makes the shape concrete, and the shape is usually surprising in the same direction.

Take a lender with 800 active accounts. Suppose 200 are secured with more than four months remaining, of which 40 carry balances well above the portfolio median. Suppose 150 accounts are within a quarter of maturity, and roughly 60 borrowers have some prior lapse in their filing history, overlapping partially with the secured group.

Applying the criteria above, Tier 1 collects the 40 large secured exposures plus the prior-lapse borrowers not already captured, landing somewhere near 80 accounts. Tier 3 collects the near-maturity and small unsecured accounts, perhaps 250. Tier 2 takes the remainder, roughly 470.

At monthly, quarterly and annual intervals respectively, the annual check volume is about 960 for Tier 1, 1,880 for Tier 2, and 250 for Tier 3: call it 3,090 checks a year across 800 borrowers. Uniform monthly monitoring of the same book would be 9,600. Uniform annual would be 800, with the 40 largest secured exposures checked once a year against a four-month clock.

Three things fall out of that arithmetic and they generalize well beyond this example.

• Tiering costs roughly a third of uniform frequent monitoring while giving the exposures that matter a shorter interval than uniform monitoring would.

• The tier that needs the most attention is small. Around ten percent of accounts, which is what makes a monthly interval affordable there.

• Uniform annual monitoring is not cheaper in any meaningful sense. It saves a modest number of credits and leaves the largest secured positions structurally unprotected against the one clock that binds them.

The last point is the one worth carrying into a budget conversation. The choice is rarely between an expensive policy and a cheap one. It is between a policy that concentrates spend where consequence lives and one that spreads it evenly regardless of consequence, at a total cost difference small enough that it should not be the deciding factor.

A caution on the arithmetic: these are illustrative figures using plausible proportions, not benchmarks. The proportions vary considerably by lending product, and the only numbers worth planning against are the ones from your own book. The calculation itself takes an afternoon with a servicing-system export, and doing it is more valuable than adopting anyone else's ratios.

How should the tiers be built and maintained?

Start from data you already hold rather than from a scoring exercise.

Segment on attributes already in the servicing system. Secured flag, outstanding balance, maturity date, and state of formation are all present in any loan system. That produces a first-pass tiering with no new data collection and no model.

Add filing history at the next check. Whether an entity has previously lapsed is visible in its state record. Capture it once and use it as a permanent tier modifier.

Set the escalation rule before launch. It is the component that does the most work and the one most likely to be deferred and forgotten.

Review the distribution, not individual assignments. If ninety percent of the book lands in Tier 1, the criteria are too loose and the policy has quietly become uniform monitoring with extra steps. If Tier 1 is almost empty, it is too tight.

Re-tier on a schedule. Balances amortize and terms shorten, so a borrower correctly placed in Tier 1 last year may belong in Tier 2 now. An annual re-tier is usually sufficient and takes minutes if the segmentation is query-driven rather than manual.

The honest framing for a credit committee: this is a resource allocation policy, not a risk model. It decides where attention goes. It does not predict defaults, and presenting it as though it does invites the wrong questions.

What does the monitoring itself actually provide?

Precision here keeps the policy honest, because a tiering framework built on an inflated view of the tooling will have gaps exactly where it claims coverage.

Business Monitoring re-checks a borrower's Secretary of State record on a cadence you configure, from daily up to every 30 days, and reports what changed against the previous check, classified by severity.^[1] The configurable interval is what makes tiering implementable: different borrowers can carry different cadences under one policy.

The boundaries:

• Scope is the Secretary of State record only. No OFAC or other watchlist screening, no UCC filings, no court dockets, no professional licenses. Sanctions re-screening remains a customer-side workflow on a customer-side schedule, and no tier assignment changes that.

• It cannot surface unfiled events. Ownership transfers being the clearest case.

• It reports change, not cause or meaning. Tier assignment determines how fast you hear. Interpretation stays with your credit policy.

• It does not enforce your tiers. Assigning borrowers to intervals and re-tiering them over time is your system's work.

The underlying record is the same one a point-in-time lookup returns:

curl -X GET "https://apigateway.cobaltintelligence.com/v1/search?searchQuery=Acme%20Holdings%20LLC&state=TX" \
  -H "x-api-key: YOUR_API_KEY"

The difference a tiered policy makes is not access to better data. It is that the borrowers who can hurt you most are checked often enough to act inside the clocks you are subject to, and the rest are not checked more than they are worth.

What should you do first?

Three steps, in order, none requiring new tooling.

Count your secured exposures with more than four months remaining. That is the population where the UCC amendment window genuinely binds, and it is usually a much smaller number than the full portfolio. Whatever else you do, that group needs an interval shorter than four months.

Compute the actual cost. Borrowers multiplied by checks per year multiplied by credit price. Do it for a monthly Tier 1 before assuming it is unaffordable, because the assumption is usually wrong and it is the assumption driving most uniform-annual policies.

Turn on the escalation rule. Any change moves a borrower up a tier. It is a few lines of logic, it needs no maintenance, and it captures most of the benefit of a far more elaborate risk model.

For what the detected changes actually mean once they arrive, our guides to entity status transitions and the administrative dissolution reinstatement window cover the consequences each tier is designed to catch in time.^[9] Registered agent lapses are one of the documented routes into that sequence.^[10]