Chapter 12: Terms, Conditions, and Coverage Structuring: Designing the Policy That Fits the Risk

"You can decline a risk, or you can structure it. The underwriters who only know how to do the first spend their careers saying no to accounts the second kind are quietly making money on." — A line I have repeated to every trainee I have ever mentored [constructed teaching example]

Overview

You have done the hard work of the last four chapters. You gathered the information (Chapter 8), assessed the hazards and controls (Chapter 9), put the loss history through the math (Chapter 10), and built an indicated rate (Chapter 11). Now the account is in front of you with a number attached — and the number, honestly, is too high. The risk as submitted, priced for everything that could go wrong, would cost the insured more than the market will bear, or would carry an exposure your own catastrophe budget cannot absorb at any price. A junior underwriter looks at that and declines. You have one more move, and it is the move this chapter teaches: you restructure the deal. You raise the deductible so the insured keeps the small, frequent losses you don't want to administer anyway. You put a percentage wind deductible on the catastrophe peril so the insured shares the tail you can't fully diversify. You sublimit the one exposure that is dragging the whole price up, and you write an endorsement that carves out the hazard you cannot get comfortable with. The same risk, the same insured — but a different policy, one that fits.

That is what coverage structuring is: the art of building a policy whose terms make a risk acceptable that the base form, at the base limit, with a flat deductible, was not. Price is only half the deal. The other half is the shape of the coverage — where the insured's money stops and yours begins, how high the policy goes, what it carves out, and what behavior it engineers. Two policies at the same premium can be wildly different bets for the insurer depending entirely on how they are structured. A senior underwriter who has internalized this chapter rarely declines a risk outright; instead they ask, at what terms would I write this? — and that question, asked well, is the difference between a book that grows profitably and one that only ever shrinks.

We will work through the structuring toolkit in the order you would actually reach for it: deductibles first (the most common lever), then the retentions and large-deductible programs that hand whole layers back to a sophisticated insured, then limits and sublimits (how high the policy goes and where you cap it), then the coinsurance clause that polices under-insurance, then the endorsements and manuscript language that restrict or broaden the form, and finally the binder and the coverage letter — the documents that make the deal real. Throughout, hold one idea: every structural feature is doing double duty, slicing the cost of a loss and shaping the insured's incentives to prevent it.

In this chapter, you will learn to:

• Use the deductible as both a price lever and a behavior lever, and choose among per-occurrence, aggregate, and percentage forms.
• Define the self-insured retention (SIR) and the large-deductible program, and explain how each transfers risk back to the insured.
• Set per-occurrence and aggregate limits and carve out sublimits so the policy responds as you intend.
• Apply the coinsurance clause and compute the penalty that under-insurance triggers.
• Restrict or broaden coverage with endorsements, and manuscript language for a risk no standard form fits.
• Issue a binder and a coverage letter, and state precisely what each obligates.

Learning Paths

Structuring is where a risk you priced in Chapter 11 becomes a policy you can actually bind. Every reader needs the deductible and limit material; the four paths weight the rest differently.

🏠 Personal Lines: Deductibles (§12.1) and the catastrophe/wind deductible (§12.1) are central to how homeowners is written in coastal and wildfire states — you will see the named-storm deductible again in Chapter 15. Coinsurance (§12.4) matters less in personal lines, but insurance-to-value (Chapter 15) is the same idea wearing a different name. Skim the large-deductible and SIR sections (§12.2); they are a commercial world. 🏢 Commercial Lines: All of it. This is the chapter where commercial coverage gets designed. The SIR and large-deductible programs (§12.2), sublimits (§12.3), and manuscripting (§12.6) are everyday tools on middle-market and large accounts, and the Harbor Steel terms you draft here carry straight into the decision in Chapter 13 and the property deep-dive in Chapter 19. 📊 Analytics: Structure changes the shape of the loss distribution you priced in Chapter 10 — a deductible truncates the small losses, a limit caps the tail, a sublimit reshapes one peril. Watch §12.1 and §12.3 for how each feature moves the expected loss the rate has to cover; mis-modeling a deductible credit is a common pricing error. 📜 Certification: §12.1–§12.4 map to the policy-provisions and coverage-structuring content tested in AINS and across the CPCU commercial lines; deductibles, limits, sublimits, and coinsurance recur on every exam, and the binder/coverage-letter mechanics (§12.7) are core insurance-operations material.

12.1 Deductibles: per-occurrence, aggregate, and percentage (catastrophe) deductibles

Start with the lever you will reach for most. A deductible is the portion of a covered loss the insured agrees to pay before the insurer pays anything — the first dollars of every claim, retained by the insured. On a property policy with a \$25,000 deductible, a \$100,000 fire pays \$75,000; a \$20,000 fire pays nothing, because it never crosses the threshold. That is the simple mechanic. The reason deductibles are a structuring tool, and not merely a discount, is everything that mechanic does beyond reducing the insurer's payout.

A deductible does three jobs at once, and you should be able to name all three before you set one. First, it sheds the losses you don't want — the small, frequent, high-friction claims that cost almost as much to adjust as they cost to pay. An insurer that pays first-dollar on a thousand \$800 claims spends a fortune on claims handling for losses the insured could easily have absorbed. Second, it lowers the price, because the insured is now retaining expected losses the rate no longer has to cover (more on the math in a moment). Third — and this is the job the junior underwriter forgets — it preserves the insured's skin in the game, the incentive to prevent and minimize loss that we first met in Chapter 1 as the answer to moral and morale hazard. An insured who eats the first \$25,000 of every loss thinks hard about its housekeeping, its maintenance, its safety program. Remove the deductible and you have quietly removed that incentive from every risk in your book.

📋 At the Desk Here is the deductible-credit math, because underwriters get it wrong in both directions. Raising a deductible from \$10,000 to \$50,000 does not save the insured the full \$40,000 of additional retention as premium — it saves them the expected losses that fall in that \$10K–\$50K layer, which is a much smaller number, because most losses are small but only some losses land in that band each year. Suppose a class of risk produces, per \$1,000 of value, an expected \$1.20 of losses in the \$10K–\$50K per-claim layer. Then moving the deductible from \$10K to \$50K should reduce the pure premium by about that \$1.20 per \$1,000 — plus a credit for the claims-handling expense you no longer incur on losses that now fall below the deductible. The trap cuts both ways: give too much credit and you have underpriced the risk for the losses you still carry above \$50K; give too little and the broker takes the account to a competitor who did the math right. The deductible credit is a pricing decision, not a courtesy — build it from the expected loss in the retained layer, the way Chapter 11 built every other piece of the rate.

There are three deductible structures you must be able to tell apart, because they shed completely different losses.

A per-occurrence deductible applies to each loss event separately. A \$25,000 per-occurrence deductible means the insured absorbs the first \$25,000 of every claim — three fires in a year means three deductibles. This is the default on most property and liability policies, and it is the one that does the most behavioral work, because it bites on every single loss.

An aggregate deductible applies to the insured's total retained losses across the policy period, after which the insurer pays from the first dollar. An insured with a \$250,000 aggregate deductible pays its own losses until they sum to \$250,000 for the year, then the policy responds in full. This structure is common in large-deductible and loss-sensitive programs (§12.2): it caps the insured's total retention, which a sophisticated buyer with a predictable frequency of small losses often prefers, because it converts an open-ended stream of small retained losses into a known maximum. The trade-off for the underwriter is real — an aggregate deductible weakens the per-loss incentive once the insured nears the cap, because the marginal loss is the insurer's, not theirs. You often see a per-occurrence deductible and an aggregate stop combined, so the insured retains each loss up to a point but never more than a ceiling in total.

The third structure is the one that matters most for the account on your desk. A percentage deductible — almost always a catastrophe deductible — is expressed not as a flat dollar amount but as a percentage of the insured value, and it applies only to a named catastrophe peril, usually wind or earthquake. A "5% named-windstorm deductible" on a \$20,000,000 building means that when a named storm causes the loss, the insured absorbs the first \$1,000,000 — 5% of \$20M — before the policy pays. (For an ordinary, non-catastrophe loss, a separate, much smaller flat "all-other-perils" deductible applies; we will set both on Harbor Steel.) The percentage deductible exists for one reason: catastrophe is the correlated peril that breaks the law of large numbers (Chapter 1), and the insurer cannot afford to pay first-dollar on a loss that strikes ten thousand insureds in the same county on the same afternoon. The percentage deductible forces the insured to retain a meaningful slice of the catastrophe loss, which simultaneously reduces the insurer's correlated exposure, prices the tail the insurer does keep, and — not incidentally — gives the insured a reason to invest in mitigation (the storm shutters, the roof straps, the elevation) that reduces the loss for both of them.

⚠️ Underwriting Trap The most expensive deductible mistake is a flat dollar deductible on a catastrophe-exposed property. A \$50,000 flat deductible looks like real risk-sharing on a routine fire. On a hurricane that does \$8,000,000 of damage to that same building, a \$50,000 deductible is a rounding error — the insurer eats \$7,950,000 of a correlated loss it is taking simultaneously across its entire coastal book. Coastal carriers learned this the hard way through the great Gulf and Atlantic storms; the percentage wind deductible exists precisely because flat deductibles fail on catastrophe. When you see a coastal property submitted with a flat deductible on the wind peril, that is not a competitive feature to match — it is a mispriced policy you should be glad a competitor wrote.

🔍 Check Your Understanding 1. A coastal warehouse has a \$25,000 flat deductible and a 5% named-windstorm deductible. A kitchen fire does \$60,000 of damage; a hurricane does \$3,000,000. The building is insured for \$10M. How much does the insured retain in each case, and which deductible applies to which loss? 2. Name the three jobs a deductible does. Which one does a junior underwriter most often forget, and which chapter first taught you why it matters?

12.2 Self-insured retentions and large-deductible programs

The deductible hands the insured the first dollars of a loss. Two related structures hand them whole layers — and, with them, a degree of control and responsibility that changes the underwriting entirely. These are the tools of the sophisticated commercial buyer: the company large enough, financially strong enough, and risk-aware enough to keep a meaningful chunk of its own losses rather than pay an insurer to carry them.

A self-insured retention (SIR) is an amount of loss the insured retains and pays directly before the insurer's coverage attaches at all — and the crucial distinction from a deductible is who handles the claim within the retained layer. Under a deductible, the insurer typically adjusts and pays the whole claim, then bills the insured back for the deductible portion; the insurer is in the loss from the first dollar, administratively. Under an SIR, the insured is responsible for the loss within the retention — the insured (often through a third-party administrator) investigates, defends, and pays claims up to the SIR, and the insurer's policy does not respond, and often does not even know about a claim, until the loss pierces the retention. The insurer is genuinely off the risk below the SIR. This matters for three reasons: the insured controls (and pays for) claims handling in the retained layer; the insurer's limits sit excess of the retention rather than eroding from the first dollar; and the insured's loss data, below the SIR, may never reach the insurer at all — which is something you must account for when you underwrite the renewal, because the loss runs you see only show the claims that pierced the retention.

⚖️ Compliance Corner SIRs and large deductibles run into a regulatory wall that flat deductibles do not, and you must know where it is. In most lines the insured may retain as much as it wishes — but workers' compensation is statutory (Chapter 22 owns this fully), and an injured worker is entitled to the full statutory benefit regardless of any retention the employer agreed to. The worker is not a party to the deductible. So in a workers'-comp large-deductible program, the insurer must pay the statutory benefit from the first dollar and then collect the deductible back from the employer — which means the insurer is exposed to the employer's credit risk if the employer fails before reimbursing. That is why large-deductible WC programs require collateral (letters of credit, trusts, cash) and why a credit deterioration in the insured is an underwriting event, not just a finance problem. The structure that looks like "the insured keeps its own losses" is, in WC, "the insurer fronts the losses and lends the insured the money, secured." Never structure a large-deductible WC program without confirming the collateral and the admitted-paper and filing requirements in the state.

A large-deductible program is the high-volume cousin of the SIR, used mostly in workers' compensation, commercial auto, and general liability for mid-to-large accounts. The insured takes a deductible far above the routine — commonly in the hundreds of thousands of dollars per occurrence, sometimes higher — and in exchange pays a dramatically lower premium, because it is now retaining the bulk of its own expected losses and buying the insurer's paper and capacity mostly for the catastrophic layer above the deductible and for the services (claims handling, regulatory compliance, the "admitted" policy that satisfies a certificate-of-insurance requirement). The economics are compelling for a financially strong insured with a large, predictable loss frequency: it stops paying the insurer's expense load and profit margin on the working-layer losses it can comfortably fund itself, and it keeps the cash-flow benefit of paying those losses over time rather than as upfront premium.

The structure is not free of risk for either side, and an underwriter must see both. For the insured, the large deductible converts a fixed premium into a variable cost that can spike in a bad year — a string of serious claims, each retained up to the deductible, can cost far more than the guaranteed-cost premium would have. For the insurer, the exposure is the insured's credit: the policy pays statutory and third-party claims from the first dollar and relies on the insured to reimburse the deductible, so if the insured becomes insolvent mid-program, the insurer is left holding losses it priced as someone else's. Hence the collateral. Hence, too, the underwriting question that defines these programs: is this insured financially strong enough, and is its loss frequency predictable enough, that handing it a large retention makes the risk better rather than worse?

🤖 Model vs. Judgment A pricing model scoring a large-deductible submission sees a truncated loss history — only the claims that pierced the deductible ever appear in the data — and will happily report a low loss ratio and recommend an aggressive price. The model is reading a censored distribution and does not know it. The underwriter's judgment is to ask what the model cannot: how many claims below the deductible is this insured actually generating (which the broker can supply from the TPA), is its frequency trending up, and is its balance sheet strong enough to fund a bad year? A clean loss run on a large-deductible account can mean a genuinely good risk — or it can mean the insured is absorbing a worsening frequency of losses you simply cannot see. Reading the censored history correctly is exactly the kind of judgment the model cannot supply, and it is why these accounts are underwritten by people, not algorithms.

For Harbor Steel, the SIR and large-deductible structures are mostly not the right tool, and knowing why is itself instructive: at roughly \$45M in revenue with a single plant and a loss history that includes a \$1.2M fire, the company is neither large enough nor financially deep enough to want to retain a serious layer of its own catastrophe or property risk, and a coastal property exposure is exactly the kind of correlated, high-severity risk you do not want a mid-size insured self-funding. We will hand Harbor Steel risk-sharing through deductibles and the percentage wind deductible, not through an SIR. The large deductible enters the file only on the workers'-comp line, lightly, and only as a question for Chapter 22.

12.3 Limits: per-occurrence, aggregate, and sublimits

If the deductible decides where the insured's money stops and yours begins, the limit decides where yours stops. The limit is the maximum the insurer will pay, and getting its structure right is as much a part of designing the policy as setting the price. Two policies with identical premiums and deductibles can expose the insurer to wildly different maximum losses depending entirely on how the limits are stacked.

The foundational distinction is between two ways of capping the insurer's payout. A per-occurrence limit caps what the insurer pays for any single loss event. A general-liability policy with a \$1,000,000 per-occurrence limit pays at most \$1,000,000 for any one occurrence, no matter how many claimants or how large the total demand. An aggregate limit caps the total the insurer will pay for all covered losses across the policy period. A policy with a \$1,000,000 per-occurrence limit and a \$2,000,000 aggregate will pay up to \$1M for any one event but no more than \$2M for the year combined — so two large occurrences could exhaust the aggregate and leave the insured uncovered for a third. The distinction is the difference between "how bad can one loss be?" and "how bad can the whole year be?", and you must underwrite both, because a risk with acceptable per-occurrence severity but a high frequency of serious losses can blow through an aggregate in a way the per-occurrence limit never reveals.

📋 At the Desk The per-occurrence-versus-aggregate structure is where liability underwriting is quietly won or lost. Property is usually written per-occurrence with no annual aggregate (each fire is its own event, and the building can only burn so many times), but liability is where aggregates bite, because the same ongoing operation — a product in the market, a contractor's completed work, a polluting process — can generate many claims in one year. When you see a products-liability exposure with a single product line in wide distribution, the aggregate, not the per-occurrence limit, is the number that should worry you: one defective production run can produce dozens of occurrences. The discipline is to size the aggregate to the plausible bad year, not the typical year, and to know that an exhausted aggregate is a coverage gap the insured discovers at the worst possible moment.

A sublimit is a limit inside a limit: a cap on a particular coverage, peril, or category of loss that is lower than the policy's overall limit. A commercial property policy with a \$20,000,000 limit might carry a \$250,000 sublimit for "accounts receivable," a \$100,000 sublimit for "valuable papers and records," a \$500,000 sublimit for "off-premises property in transit," and — critically for catastrophe management — a sublimit on a peril like flood or earth movement that the insurer is willing to cover but not to the full policy limit. The sublimit is one of the most precise tools in the structuring kit, because it lets the underwriter say yes, but only this far to an exposure that would otherwise force a decline or a much higher price. The exposure that is dragging the whole account's price up — the one peril or coverage that carries disproportionate severity — can frequently be sublimited down to a level the insurer is comfortable with, leaving the rest of the policy at full limit and a reasonable price.

POLICY LIMIT STRUCTURE — a commercial property policy        [constructed teaching example]

  POLICY LIMIT (Building + BPP)                              ████████████████████  $20,000,000
    ├─ Flood (sublimit, named peril)                         ███                     $2,000,000
    ├─ Earth movement (sublimit)                             ██                      $1,000,000
    ├─ Off-premises property in transit (sublimit)           █                         $500,000
    ├─ Accounts receivable (sublimit)                        ▌                         $250,000
    └─ Valuable papers & records (sublimit)                  ▌                         $100,000

  AOP deductible (all other perils): $25,000 flat
  Named-windstorm deductible: 5% of insured value = $1,000,000

  Read it top-down: the full limit answers a total fire loss; each sublimit caps a coverage the
  underwriter will write but not to the full $20M. The wind peril is governed by a PERCENTAGE
  deductible, not a sublimit — the insurer keeps the full limit above a large retained slice.

Read that structure the way an underwriter reads it. The \$20M at the top is what the policy pays for the exposure it is really written for — a total fire loss of the building and contents. Each sublimit beneath it is a coverage the underwriter is willing to include for convenience or competitiveness but is not willing to expose to the full limit, either because the data is thin (transit), the severity is unbounded (flood), or the coverage is a courtesy rather than the core deal (valuable papers). The catastrophe peril, note, is handled differently — not by a sublimit but by the percentage deductible from §12.1 — because the underwriter wants the insured to keep the full \$20M limit above the storm but to retain a large first slice of any storm loss. Sublimit versus percentage deductible is a real design choice, and which one you reach for depends on whether you want to cap the insurer's top exposure to a peril (sublimit) or force the insured to retain its bottom slice (deductible).

🔍 Check Your Understanding 1. A liability policy is written at \$1M per occurrence / \$2M aggregate. The insured suffers three covered occurrences in one year, of \$800K, \$1.4M, and \$600K. What does the insurer pay on each, and is the insured fully covered for all three? Where is the gap? 2. Explain the difference between handling a flood exposure with a \$2M sublimit and handling a wind exposure with a 5% deductible. What does each one protect the insurer from?

12.4 Coinsurance: the clause that punishes under-insurance

Now we come to a clause that is not a limit, not a deductible, and not an endorsement, but a condition — one that quietly reshapes every property claim, and that more brokers and insureds misunderstand than any other provision in the policy. The coinsurance clause is a property-policy condition requiring the insured to carry insurance equal to a stated percentage (commonly 80%, 90%, or 100%) of the property's full value; if the insured carries less than that percentage, the insurer pays only a proportional share of any loss — even a partial loss far below the policy limit. The clause exists to combat a specific adverse-selection problem in property insurance: left alone, a rational insured would buy a limit only big enough to cover the partial losses it actually expects (a kitchen fire, a damaged loading dock), not the total loss it will probably never suffer — and would pay far too little premium for the limit it carries. Coinsurance forces insurance-to-value, and it does so with a penalty that bites at claim time.

The mechanic is a single formula, and it is worth committing to memory because you will explain it to an upset insured at least once a year:

$$\text{Loss payment} = \text{Loss} \times \frac{\text{Insurance carried}}{\text{Insurance required}} - \text{deductible}$$

where "insurance required" is the coinsurance percentage times the property's full value at the time of loss. If the insured carries at least the required amount, the fraction is 1 (or capped at 1), and the loss is paid in full up to the limit, less the deductible. If the insured carries less, the fraction is below 1, and the insured becomes a "coinsurer" of its own loss — bearing a share proportional to its under-insurance. Work the number, because it surprises people:

THE COINSURANCE PENALTY — an 80% coinsurance clause        [constructed teaching example]

  Building's full value at time of loss .............. $10,000,000
  Coinsurance requirement (80%) ...................... $8,000,000   ← insurance REQUIRED
  Insurance the insured actually carried ............. $6,000,000   ← insurance CARRIED
  A partial fire loss ................................ $2,000,000
  Deductible ......................................... $50,000

  Penalty ratio = carried / required = 6,000,000 / 8,000,000 = 0.75

  Loss payment = $2,000,000 × 0.75 − $50,000 = $1,500,000 − $50,000 = $1,450,000

  The insured carried $6M of limit and suffered a $2M loss — well within the limit — yet collects
  only $1,450,000. The missing $550,000 is the coinsurance penalty: the price of insuring to 60% of
  value when the policy required 80%. The limit was never the problem. The VALUE was.

Look hard at that result, because it is the part that traps insureds: the loss was \$2M, the policy limit was \$6M, the loss was nowhere near the limit — and the insured still took a \$550,000 haircut. Coinsurance does not penalize you for hitting your limit; it penalizes you for carrying too small a limit relative to value, on every partial loss, all year long. That is why an insured who shaves its declared values to save premium is making a far worse bet than it realizes: it saves a little premium today and silently under-collects on every claim until it corrects the valuation.

⚠️ Underwriting Trap The coinsurance trap runs in both directions, and the underwriter owns both sides of it. The insured's trap is under-declaring values to cut premium and then discovering the penalty at claim time. The underwriter's trap is subtler and more dangerous: accepting under-stated values at binding means you collected too little premium all year (you rated a \$10M building as a \$6M building), and the coinsurance penalty does not make you whole — it just transfers the shortfall to the insured at claim time, where it becomes a bad-faith complaint, a furious broker, and a lost account. Coinsurance is not a substitute for getting the values right at underwriting. The disciplined move is to verify values up front — an appraisal, a cost-estimator, a square-footage check — and, where you want to take valuation risk off the table entirely, to write the property on an agreed-value basis, which suspends the coinsurance clause in exchange for an agreed statement of value (agreed value is owned by Chapter 19; we will use it on Harbor Steel there). Coinsurance polices under-insurance; agreed value removes the dispute. A good property underwriter reaches for the second on any account where a coinsurance fight at claim time would poison the relationship.

The coinsurance clause connects directly to the recurring theme that pricing follows risk (Chapter 11): the premium is built on the declared value, so a wrong value is a wrong price before any rating factor is applied. It also connects to adverse selection (Chapter 1): coinsurance is one more device by which the policy defends itself against an insured's incentive to reveal less than the full exposure. And it sets up the agreed-value option that Chapter 19 will use to take valuation risk off the Harbor Steel property entirely.

12.5 Endorsements that restrict or broaden coverage

The deductibles, limits, sublimits, and coinsurance terms reshape how much the policy pays. Endorsements reshape what it covers. An endorsement, as you met it in Chapter 5, is a document attached to the policy that adds to, deletes from, or modifies the standard form — and it is the single most flexible tool the underwriter has for fitting coverage to a risk. Endorsements come in two flavors, and a structuring underwriter uses both deliberately: those that restrict coverage (to make a marginal risk acceptable without raising the price into uncompetitiveness) and those that broaden it (to win and keep a desirable account, or to fill a genuine coverage gap the standard form leaves open).

Restrictive endorsements are the underwriter's scalpel. When a risk carries one hazard you cannot get comfortable with — but is otherwise a good account — you do not have to decline the whole thing or price the whole thing as if that hazard were certain. You carve the hazard out. A property with an aging roof gets an endorsement settling roof losses on an actual cash value basis (depreciated value, not replacement cost — ACV is owned by Chapter 15) rather than replacement cost, so the insurer is not buying the insured a new roof through a storm claim on a roof already at the end of its life. A contractor with a worrisome past gets a particular operation excluded. A manufacturer with a discontinued product line that generated claims gets that product's completed-operations exposure excluded going forward. Each restrictive endorsement does the same job: it removes a slice of exposure the underwriter is unwilling to price into the base deal, turning a decline into a qualified yes.

📋 At the Desk The discipline with restrictive endorsements is to restrict the specific hazard, not the whole coverage, and to be able to say in plain words what the insured is giving up and why. "We'll write the property, but the thirty-year-old roof is settled at actual cash value until you replace it — because we're not willing to fund a full roof replacement through a wind claim on a roof that's already at the end of its life, and once you replace it we'll endorse it back to replacement cost." That sentence is fair, specific, and defensible; the broker can take it to the insured and the insured can decide. The bad version of this tool is the blanket exclusion stapled on without explanation, the one the insured never understood and discovers at claim time. A restriction the insured knowingly accepted is a structured deal; a restriction the insured never registered is a coverage dispute and a bad-faith exposure waiting to happen. Document what you restricted, why, and that the broker was told.

Broadening endorsements run the other way, and they are how you win accounts and close real gaps. The standard form is a compromise written for the average risk; a particular insured frequently needs more. Additional insured endorsements (owned by Chapter 21) extend the insured's liability coverage to a landlord, a customer, or a general contractor who requires it by contract — table stakes on most commercial accounts. A "blanket" version of a coverage the form grants per-named-party can broaden it to whole categories. Replacement-cost endorsements, ordinance-or-law coverage (which pays the extra cost of rebuilding to current code after a loss), spoilage coverage, and dozens of others each fill a gap the bare form leaves open. The underwriter's job is to know which broadenings are free risk (a coverage the insured needs that adds little exposure) and which are real exposure that must be priced or sublimited. Ordinance-or-law on an old building in a jurisdiction with strict modern codes, for instance, is not a free add — it can add substantially to a total loss, and it should be priced and often sublimited, not thrown in.

The structuring mindset holds both directions in one hand. On a single account you will frequently restrict one hazard and broaden another in the same policy — ACV on the roof (restrict) and additional-insured status for the insured's major customers (broaden) — because the goal is never "more coverage" or "less coverage" in the abstract. The goal is the right coverage: a policy whose shape matches the risk, priced for what it actually carries.

12.6 Manuscripting for complex risks

Sometimes no standard form fits the risk, and no combination of off-the-shelf endorsements gets you there. A novel exposure, an unusual operation, a contract with bespoke insurance requirements, a risk that spans coverage lines in a way the bureau forms never anticipated — for these, the underwriter (with legal and sometimes actuarial support) writes manuscript coverage: policy language drafted for this specific risk rather than pulled from a bureau form. We met manuscript versus bureau forms in Chapter 5; here we care about when and how an underwriter reaches for manuscripting as a structuring tool, and about the discipline it demands.

Manuscripting is the most powerful and the most dangerous tool in this chapter. Powerful, because it lets you write a risk that would otherwise be uninsurable for lack of a form — you can define the covered exposure, the triggers, the exclusions, and the limits exactly to fit a one-of-a-kind situation. Dangerous, because every word you write is a word a court will later interpret, often against the drafter, and you are working without the decades of litigation that have already tested and clarified the standard forms. A bureau form has been through the courts; its ambiguities have been ruled on; you know, more or less, what it means. A manuscript form is new language with no track record, and an ambiguity in it is construed against the insurer who drafted it — the doctrine of contra proferentem, which you met in the contract-law frame of Chapter 4. The standard forms are not standard by accident; they are standard because they have been litigated into clarity, and abandoning them means abandoning that clarity.

⚖️ Compliance Corner Manuscripting also runs into the filing regime from Chapter 4. In the admitted market, policy forms generally must be filed with — and in many states approved by — the regulator before use, and a manuscript form is a non-standard form that triggers form-filing requirements; you cannot simply draft and bind it in every state. This is one of the structural reasons the surplus-lines (non-admitted) market exists (owned by Chapter 4): surplus-lines carriers have freedom of form and rate, which is precisely why the genuinely novel, manuscript-heavy risks — the new cyber wordings, the bespoke specialty placements — so often land in the excess-and-surplus market rather than the admitted one. When a risk needs a manuscript form, part of the structuring decision is which market can legally write that form, and the answer often moves the placement out of the admitted market entirely.

The discipline of manuscripting is conservatism in drafting. You define terms precisely, you state exclusions explicitly rather than relying on implication, you resolve ambiguities in the drafting room rather than leaving them for a court, and — the cardinal rule — you do not manuscript what a standard form already covers well. Reach for manuscript language only where the standard forms genuinely fail the risk, keep the manuscripted portion as narrow as the situation allows, and lean on bureau language for everything the bureau already handles. For Harbor Steel, the answer is that you do not need to manuscript: a custom metal-fabrication account, however cat-exposed, is a well-understood commercial risk that the standard commercial-package forms, with the right endorsements and the percentage wind deductible, cover perfectly well. Knowing that — knowing when not to reach for the most powerful tool — is itself a mark of the experienced underwriter.

12.7 The binder and the coverage letter

The structuring is done. You have a deductible scheme, a percentage wind deductible, a limit and sublimit structure, a coinsurance approach (or agreed value), the endorsements that restrict and broaden, and a decision about whether any of it needs manuscripting. Now the deal has to become real — and between "I've decided the terms" and "the formal policy is issued weeks later" sits a document that legally binds the insurer immediately: the binder.

A binder is a temporary contract of insurance that provides coverage immediately, pending the issuance of the formal policy. It is short — the named insured, the coverage, the limits, the deductibles, the effective date and time, the premium or a basis for it, and any conditions — but it is real insurance: if a covered loss occurs after the binder takes effect and before the policy is issued, the binder responds. This is why a binder is not a casual document. The moment you bind, your insurer is on the risk for the full agreed coverage, whether or not anyone has yet typed the policy. Binders exist because commerce cannot wait for paperwork — a business closing on a building Monday morning needs coverage Monday morning, not when the policy prints three weeks later — and the binder bridges that gap.

📋 At the Desk A binder binds exactly what it says, so the cardinal rule is that the binder must reflect the terms you actually agreed — including every subjectivity and restriction. If you intend to write Harbor Steel's property on an ACV-roof endorsement with a 5% wind deductible subject to a roof-replacement warranty and a hot-work program, the binder must say so. A binder that omits a restriction you intended grants the broader coverage — the binder is the contract, and you cannot retroactively narrow it after a loss by pointing to terms you "meant" to include. The most expensive errors in a binder are the conditions left off it. Conversely, a binder issued with conditions the insured has not yet met (a "subject-to" binder) is a structuring choice with teeth: it binds coverage but conditions its continuation on the insured delivering the subjectivity by a deadline, and Chapter 13 is where we turn those subjectivities into the formal conditions of the quote.

The coverage letter (or quotation letter) is the binder's calmer cousin and serves a different moment. Where the binder grants coverage, the coverage letter offers it: it sets out, in writing, the terms on which the insurer is willing to write the risk — the coverages, limits, deductibles, premium, and, crucially, the subjectivities (the conditions the insured must satisfy before or shortly after binding; subjectivity is owned by Chapter 13). The coverage letter is the document the broker takes back to the insured to say "here is the deal on the table." It does not bind anything; it is an offer, open for acceptance, that becomes a binder when the broker accepts and the underwriter binds. The discipline with the coverage letter is completeness and precision — every term that matters, every subjectivity, every restriction, stated plainly enough that there can be no dispute later about what was offered. A vague coverage letter is the seed of a coverage dispute; a precise one is the structured deal written down.

THE PATH FROM DECISION TO COVERAGE        [constructed teaching example]

  SUBMISSION ──► UNDERWRITING DECISION ──► COVERAGE LETTER ──► BINDER ──► POLICY ISSUED
   (the ask)      (accept w/ terms &        (the offer:        (coverage   (the formal
                   subjectivities, Ch.13)    terms + subj.)     starts NOW)  contract)
                                                  │                │
                                            broker takes      insurer on
                                            it to insured      the risk for
                                            for acceptance     exactly what
                                                               the binder says

That path — submission to decision to coverage letter to binder to policy — is the spine of how a structured deal actually gets placed, and you will walk it on every account. The terms you designed in §12.1 through §12.6 are only as good as the binder and coverage letter that record them. Structure the policy brilliantly and then leave a subjectivity off the binder, and you have given away exactly the protection you so carefully built.

🔍 Check Your Understanding 1. What is the difference between a binder and a coverage letter, in terms of what each legally obligates the insurer to do? 2. You intend to write a property risk with an ACV-roof restriction, but the binder you issue does not mention the roof at all. A storm destroys the roof the next week. What coverage applies, and why?

🗂️ The Underwriting File

Structuring Harbor Steel's terms. You have the indicated, debit-rated price from Chapter 11. Now you design the policy that makes this catastrophe-exposed account acceptable at that price — because the price alone, without the right structure, would not be enough to write it. This is the chapter where the deal takes shape.

Start with the catastrophe peril, because it is the one that decides whether Harbor Steel is writable at all. The plant sits in a named-windstorm zone on the Gulf Coast, and the building is insured for \$20M. A flat deductible on the wind peril would be the mistake from §12.1 — a rounding error against a hurricane that takes the whole roof and floods the floor. So you set a 5% named-windstorm deductible: on a named storm, Harbor Steel retains the first \$1,000,000 (5% of the \$20M building value) before the policy responds. That single term does three things at once — it sheds the correlated first slice of any storm loss off your coastal book, it prices the tail you do keep more honestly, and it gives the owner a direct financial reason to invest in the storm mitigation that protects you both. For everything that is not a named storm — the fire risk that is the account's real frequency driver — you set a separate, much smaller all-other-perils (AOP) deductible, illustratively in the \$25,000–\$50,000 range, enough to shed the small claims and keep the insured's housekeeping incentive sharp without making routine coverage meaningless.

Next, the roof. The building still wears its original 1994 built-up roof, roughly thirty years old and at the end of its service life (the COPE read from Chapter 9). You are not willing to fund a brand-new roof through a storm claim on a roof that is already worn out — that would be paying for maintenance the insured deferred. So you attach an actual cash value (ACV) roof endorsement: roof losses settle on a depreciated basis until the insured completes a warranted roof replacement, at which point you endorse the roof back to replacement cost. This is the restrictive-endorsement logic of §12.5 doing exactly its job — carving out the one hazard you cannot price into the base deal, turning a near-decline into a qualified yes, with a clear path for the insured to earn back full coverage.

Then the income exposure. Harbor Steel carries \$10M of business income coverage, and the fire history tells you a serious fire could take the plant offline for many months. You note here a 12-month period of indemnity on the business-income coverage — long enough that a rebuild after a major loss does not outrun the coverage — and you flag that the full business-income valuation, the coinsurance-versus-agreed- value decision, and the period-of-indemnity analysis get their deep treatment in the property chapter (Chapter 19 owns business income, the period of indemnity, and agreed value). For now the terms are drafted: you will recommend writing the property on an agreed-value basis to take the coinsurance dispute off the table entirely (the §12.4 logic), so that a partial loss is never reduced by a valuation penalty on this account.

Finally, sublimits and the structural housekeeping. You add the modest sublimits a fabrication account needs — property in transit (steel moving to job sites), valuable papers, accounts receivable — each capped well below the \$20M limit, and you note that the inland-marine exposure (steel in transit, contractors' equipment) will be picked up properly in Chapter 26. You do not manuscript anything: the standard commercial-package forms, with these endorsements and deductibles, fit this risk well, and §12.6 is clear that reaching for manuscript language a well-understood risk does not need is a mistake.

What this chapter settles, and what it doesn't. The terms are now drafted: a 5% named-windstorm deductible, an AOP deductible, an ACV-roof endorsement until a warranted replacement, an agreed-value property basis with a 12-month business-income period of indemnity, and a set of sublimits. What is not yet settled is the decision itself — whether to issue this quote, with what subjectivities, and at whose authority. That is Chapter 13: the roof-replacement warranty, the hot-work permit program, the sprinkler certification, and the infrared electrical scan become formal subjectivities on the quote, and the account, exceeding the line underwriter's authority, goes to referral. The structure is built. The decision to put it on the table — and to defend it — comes next.

Conclusion

Price is only half the deal; structure is the other half, and it is the half that turns a marginal risk into an account you can write profitably. This chapter gave you the toolkit, in the order you reach for it. Deductibles shed the small frequent losses, lower the price by the expected loss in the retained layer, and — never forget the third job — preserve the insured's incentive to prevent loss; the percentage catastrophe deductible is the indispensable tool for the correlated peril a flat deductible cannot handle. Self-insured retentions and large-deductible programs hand whole layers back to a sophisticated insured, trading premium for the insured's credit risk and a censored loss history you must underwrite around. Limits — per-occurrence, aggregate, and sublimits — decide where the insurer's payout stops and let you say yes, but only this far to the one exposure dragging up the price. The coinsurance clause polices under-insurance with a penalty that bites on every partial loss, and agreed value (Chapter 19) takes that dispute off the table where you want valuation certainty. Endorsements restrict the hazard you can't price and broaden the coverage that wins the account, and manuscripting writes the risk no standard form fits — powerfully and dangerously, because every word is construed against the drafter. And the binder and coverage letter make the deal real, binding exactly what they say, which is why the most expensive structuring error is a subjectivity left off the binder.

Two themes ran through all of it. Pricing follows risk (Chapter 11): every structural feature changes the expected loss the rate must cover, so a deductible credit, a sublimit, or an under-stated value is a pricing decision, not a courtesy. And underwriting is judgment (Chapter 1): the choice to restructure a risk rather than decline it, to reach for a sublimit instead of a higher price, to write agreed value instead of trusting coinsurance, is judgment the rating plan cannot make for you. Harbor Steel's terms are drafted — the wind deductible, the ACV roof, the agreed-value property, the period of indemnity, the sublimits. In the next chapter, those terms meet the moment of truth: the decision to accept, decline, or modify, the subjectivities that condition the quote, and the discipline of defending the call to your manager, the broker, and the auditor. The structure is built. Now we decide.

Key Terms

• Deductible — the portion of a covered loss the insured pays before the insurer pays; a tool that sheds small losses, lowers price by the retained expected loss, and preserves the insured's incentive to prevent loss. May be per-occurrence, aggregate, or a percentage (catastrophe) deductible.
• Self-insured retention (SIR) — an amount of loss the insured retains and handles directly (often through a third-party administrator) before the insurer's coverage attaches at all; unlike a deductible, the insurer is genuinely off the risk, and claims handling, within the retention belongs to the insured.
• Per-occurrence vs. aggregate limit — a per-occurrence limit caps the insurer's payout for any single loss event; an aggregate limit caps the total payout for all covered losses across the policy period.
• Sublimit — a limit inside the policy limit: a cap on a particular coverage, peril, or category of loss set lower than the overall limit, used to include an exposure without exposing the full limit to it.
• Coinsurance clause — a property-policy condition requiring the insured to carry insurance equal to a stated percentage of full value; if the insured carries less, the insurer pays only a proportional share of any loss, even a partial one — the penalty that punishes under-insurance.
• Binder — a temporary contract of insurance that grants coverage immediately, pending issuance of the formal policy; it binds exactly the terms it states, so omitted conditions grant the broader coverage.
• Large-deductible program — a structure (common in workers' comp, auto, and GL) in which the insured takes a deductible far above the routine in exchange for a much lower premium, retaining its working-layer losses while the insurer fronts statutory/third-party payments and relies on collateral against the insured's credit.

Spaced Review

1. A coastal property is submitted to you with a flat \$50,000 deductible on all perils, including windstorm. What is the structuring mistake, what would you replace it with, and which chapter taught you why catastrophe breaks the diversification a flat deductible assumes? (§12.1; Ch. 1)
2. From Chapter 11: you raised an account's deductible from \$10,000 to \$50,000. Explain why the premium credit is not \$40,000, and what number it actually should be built from. (§12.1; Ch. 11)
3. From Chapter 6 and Chapter 9: a deductible and a loss-control requirement (a hot-work permit program) both reduce a fabrication risk's losses, but through different mechanisms. Distinguish the behavioral work a deductible does from the physical-hazard work a loss control does. (§12.1; Ch. 6, Ch. 9)
4. (The recurring pricing-discipline question.) An insured asks you to drop the coinsurance clause and write the property on whatever limit it chooses, with no agreed value. Would accepting that help or hurt your combined ratio over time, and why? Name the specific mechanism by which it would erode the rate. (§12.4; Ch. 1, Ch. 3)
5. You bind a risk on a "subject-to" binder with four subjectivities, but you forget to list one of them on the binder document. A covered loss occurs the next day. What coverage applies, and what does this tell you about the relationship between the binder and the terms you "intended"? (§12.7)