Case Study 2: Binding Beyond Authority — When the Pen Writes a Check the Company Never Signed

A note on sourcing. The mechanism in this case — losses that arise when someone with binding authority (an underwriter, an MGA, a coverholder, or a producer with a binding agreement) writes business outside the limits of that authority, and the carrier is left holding exposures it never agreed to take — is a real and recurring feature of the insurance industry, documented in the public record of binding agreements, coverholder audits, the Lloyd's coverholder framework, and a long history of disputes over whether a carrier is bound by an agent who exceeded their grant. Two real, public reference points anchor the pattern at a Tier-1 level: the broad nineteenth-and-twentieth-century body of agency law governing when a principal is bound by an agent who acts beyond actual authority but within apparent authority; and the modern delegated-authority / coverholder audit discipline (notably at Lloyd's of London) built precisely to catch this. To teach the underwriting lesson without misstating any particular dispute, the worked account below is a clearly-labeled composite. The doctrine and the industry control are real; the names, dollars, and dates are illustrative.

Background: authority is the whole point

Recall from Chapter 7 what underwriting authority is: the defined limits on what an underwriter (or a delegated agent) may approve — which lines, which classes, up to what limit, in which territories, within what guidelines. Authority is not red tape. It is the mechanism by which an insurance company controls the exposures it actually takes on. A carrier's appetite, its capital model, its reinsurance treaty, and its solvency all rest on the assumption that the business written is the business authorized. When that assumption breaks — when a pen writes outside its grant — the carrier can find itself liable for risks its own capital and reinsurance were never structured to hold.

This is the failure direction of §13.2's rule ("never bind a risk you do not have the authority to bind") and §13.6's discipline (when a risk exceeds your authority, you refer, you do not quietly bind). Case Study 1 showed what happens when a decision is never communicated. This case shows what happens when a decision is made without the authority to make it — and the company is bound anyway.

The composite account

[Composite — illustrative names, dates, and figures; the doctrine and control are real.]

Apex Specialty is a carrier that, like most, delegates limited binding authority to others. It grants a binding authority agreement (a "binder") to an MGA — a managing general agent (Chapter 3) we'll call Keystone Underwriting Managers — letting Keystone write a defined book on Apex's paper: certain small and mid-commercial classes, up to \$5M in property limit per risk, in specified states, strictly within Apex's published guidelines, with anything larger or out-of-appetite to be referred to Apex before binding.

For three years the arrangement works well. Then two pressures converge. Keystone's production is rewarded on volume, and a soft market (Chapter 3) makes good business hard to find. A large, attractive-looking account comes to Keystone's lead underwriter, Priya: a \$12M property schedule for a regional distribution company — well above the \$5M per-risk cap, and concentrated in a coastal county that Apex's guidelines treat as a referral trigger. The broker is pushing; the premium is large; the bind deadline is tomorrow.

The correct path is unambiguous: this exceeds Keystone's authority on two counts (the limit and the catastrophe-zone trigger), so Priya must refer it to Apex and let Apex decide. But referral feels slow, the broker is impatient, and Priya is confident the account is good. She reasons that she can write the \$12M by issuing it as a couple of separate binders to stay nominally "under" some internal interpretation of the cap, and that the coastal concentration is "not that bad." She binds it on Keystone's delegated authority, without referring it to Apex. The system shows bound coverage. Apex, relying on Keystone's compliance with the agreement, has no idea the exposure is on its book.

For the better part of a year, Apex carries a \$12M coastal property exposure it never authorized, never priced at the senior level, and — critically — never reflected in its catastrophe accumulation for that zone (Chapter 30) or its reinsurance arrangements (Chapter 27).

The underwriting / insurance issue

The issue surfaces in one of two ways, and the case teaches both. Either a loss arrives — a named storm hits the coastal county and the \$12M schedule is damaged — or a routine coverholder audit (Chapter 38's underwriting audit, applied to a delegated authority) catches the over-limit binders first. Suppose, as in the worst version, the storm comes first.

Now Apex faces a tangle of problems, each mapping onto a Chapter 13 (and beyond) concept:

  • Is Apex even bound? Priya acted outside her actual authority — Keystone's grant capped her at \$5M and required referral on the coastal trigger. But the insured and its broker dealt with Keystone in good faith, reasonably believing Keystone could bind on Apex's behalf — the classic apparent authority problem in agency law. Carriers frequently find that they are bound to the insured even where the agent exceeded the carrier's internal grant, precisely because the insured had no way to know the private limits of the agreement. So Apex may owe the \$12M claim it never agreed to insure — and its recourse runs against Keystone, not the insured.

  • The exposure was never in the capital or reinsurance picture (Chapters 27–28, 30). Because the over-limit risk was never referred or recorded as a senior-authority exposure, Apex's catastrophe accumulation for that zone understated the true concentration, and its reinsurance may not have contemplated it. A carrier prices, reserves, and cedes based on the business it believes it has written. Unauthorized business corrupts that picture — which is why a single rogue pen can do damage out of all proportion to one account: it poisons the portfolio math, not just the one file.

  • The break in the documented chain of authority (§13.5, §13.6). The file shows the binders but not a referral, not an Apex sign-off, not an authority record for a \$12M coastal risk — because none existed. The reconstruction test fails in the most damning way: a stranger reading the file learns not only what was bound but that it was bound by someone without the authority to bind it. The artificial split into "a couple of separate binders" to dodge the cap is, if anything, worse — it is documentary evidence of a deliberate workaround, the authority equivalent of Case Study 1's pretextual note.

What it shows

The case shows that authority is not a personal convenience to be optimized around; it is the carrier's control over its own solvency, and exceeding it is among the gravest errors in underwriting. §13.2 states the rule flatly — "never bind a risk you do not have the authority to bind" — and §13.6 gives the alternative — refer. This case is the cost of ignoring both. Priya's instinct that "the account is good" is beside the point in two ways: first, she did not have the authority to make that call, and the whole system of layered authority exists so that bigger, stranger, more correlated risks get more-senior judgment and a check against exactly her overconfidence; and second, even a good account written outside authority is a problem, because it enters the carrier's book invisibly, outside the capital and reinsurance structure built to hold it.

It also shows why the industry treats delegated authority with such suspicion and surrounds it with audits. When a carrier lets others bind on its paper — MGAs, coverholders, producers with binding agreements — it is handing out its pen, and every pen is a potential point of failure. The entire coverholder-audit apparatus (most visibly Lloyd's, but universal in delegated-authority business) exists to verify, after the fact, that the delegated pens stayed within their grants. That such an elaborate control exists at all is the clearest possible evidence of how real and how damaging binding-authority failures are.

Finally, the case shows the asymmetry that makes this so dangerous: the carrier can be bound to the insured even when the agent was not authorized. The insured did nothing wrong and reasonably relied on apparent authority, so the loss often must be paid; the carrier's only recourse is the long, expensive pursuit of the agent who exceeded the grant. The over-limit bind is therefore not a problem that can be "unwound" once a loss hits — it is, frequently, a loss the carrier must eat and then litigate to recover.

Outcome

In the composite — and in the real binding-authority disputes it is built from — the typical outcome is some combination of: the carrier honoring the claim to the insured (apparent authority) while pursuing the MGA or agent for breach of the binding agreement; the termination or sharp restriction of the delegated authority; a punishing coverholder/underwriting audit of the rest of that agent's book to find any other over-limit or out-of-appetite risks lurking unrecorded; reinsurance disputes over whether unauthorized business is covered by the treaty; and, often, the agent's own errors-and-omissions insurer becoming involved. Careers end. Agencies lose their carrier appointments. And the carrier's catastrophe and capital picture has to be re-baselined once the true, previously hidden exposures are surfaced.

The constructive outcome, again, is cultural and structural: the episode is why delegated authority comes with hard limits, mandatory-referral triggers, automated system caps that prevent over-limit binds rather than merely forbidding them, and regular audits. The best carriers do not rely on every pen exercising discipline; they engineer the discipline into the system, so that a Priya who wants to bind \$12M cannot, and is forced into the referral that should have happened on its own.

The lesson

Your pen is the company's pen, and it writes only as far as your authority reaches. A risk that exceeds your authority is not yours to decide — it is yours to refer. Binding past your authority does not make a big account yours; it makes a hidden liability the company's, outside the capital and reinsurance built to hold it. When in doubt, refer. The slow path that stays inside your authority beats the fast one that writes a check the company never signed.

Concretely, the case reinforces five of Chapter 13's rules:

  1. Never bind outside your authority (§13.2). The limit, the class, the territory, the referral trigger — they are the carrier's control over its solvency, not a personal hurdle to optimize around.
  2. Refer what exceeds your authority (§13.6) — and refer it as a complete recommendation, so the senior authority can decide quickly. Referral is the answer to the over-limit risk, not the obstacle.
  3. Do not game the cap. Splitting a risk to appear "under" a limit is a documented workaround that makes the error worse, not better — the authority equivalent of a pretextual file note.
  4. The file must record the authority (§13.5): whose authority bound it, the referral, the sign-off. A file that cannot show who had the authority fails the reconstruction test at its most critical point.
  5. Unauthorized business corrupts the portfolio, not just the file (Chapters 27–30 preview). It enters invisibly, outside the catastrophe accumulation, the capital model, and the reinsurance — which is why one rogue bind can do damage far beyond one account.

Discussion questions

  1. Distinguish actual authority from apparent authority. Why can a carrier end up bound to the insured even though its agent acted outside the carrier's internal grant — and why does that make an over-limit bind so hard to "unwind"?
  2. Priya believed "the account is good." Even granting that she was right about the risk, give two distinct reasons her bind was still a serious error. (Hint: one is about authority; one is about the portfolio.)
  3. The chapter says binding past your authority because referral "feels like an admission" is "one of the gravest errors in the profession." Why is referral — moving the decision up to the right authority — the discipline this case is really about, and how does a complete recommendation (§13.6) make referral fast enough that there is no excuse to skip it?
  4. Compare this case with Case Study 1. One is a failure of communication; the other a failure of authority. What do they share at the level of the file and the reconstruction test — and what does that tell you about why documentation is the spine of the defensible decision?
  5. The best carriers engineer authority limits into the system (hard caps that prevent over-limit binds) rather than relying on every underwriter's discipline. Argue for and against this: what does a prevent-it system gain over a forbid-it rule, and what does it risk losing (think about the legitimate override and the referral that should be granted)?
  6. Connect the unauthorized coastal exposure to catastrophe accumulation (Chapter 30) and reinsurance (Chapter 27). Explain why an exposure the carrier doesn't know it has is more dangerous than the same exposure written knowingly — and how it can turn a survivable storm into a solvency event.