RESEARCH DOSSIER
Commercial Insurance Policy Tower and Attachment Optimization
Abstract. A commercial insurance policy tower distributes an enterprise's covered losses among retained amounts, primary insurance and one or more excess layers, subject to policy-specific conditions governing attachment, limits, aggregation, defense costs, exhaustion and payment. Its economic performance depends simultaneously on the underlying loss distribution, premium and financing terms, claims handling, contractual wording and the solvency and performance of participating insurers. This monograph develops a source-based comparative account of those interacting elements, combining original analysis of primary policy material, judicial decisions, actuarial professional standards and risk-manager survey evidence with transparent hypothetical calculations. It distinguishes the arithmetic crossing of an attachment threshold from the legal satisfaction of a contractual condition for excess insurance to respond. It also separates per-occurrence tower capacity from annual aggregate capacity, the policyholder's retained financial burden from the insurer's insured indemnity, and nominal limits from recoverable protection. The mathematical analysis derives layer expected payments from the severity survival function and demonstrates the consequences of retention, aggregate depletion, correlated claims, alternative exhaustion language and incomplete underlying limits. The comparative discussion reviews selected U.S. judicial authorities, United Kingdom insurance legislation, European solvency rules and internationally recognized supervisory and actuarial guidance while identifying their respective scopes. The central finding is that attachment optimization is properly evaluated as a joint contract-and-finance decision: a low expected premium can coexist with elevated uncovered loss, liquidity strain or delayed recovery, and a formally tall tower can contain substantial gaps at precisely the states of loss for which it is purchased. The original contribution is an integrated, explicitly reproducible analytical synthesis of established contractual and actuarial concepts, not a new proprietary theory or a claim of empirical causal identification.
Keywords: Commercial insurance; policy tower; primary insurance; excess liability; commercial umbrella; attachment point; self-insured retention; deductible; occurrence limit; aggregate limit; exhaustion; follow-form policy; vertical exhaustion; horizontal exhaustion; actuarial pricing; stop-loss transform; conditional tail expectation; coverage allocation; corporate risk financing; insurance contract interpretation.
Introduction
A commercial insurance program is a set of contingent payment promises issued by one or more insurers against specified categories of enterprise loss. Its structure responds to a basic corporate finance problem: uncertain obligations can exceed available liquid resources at the time payment is required, even where their long-run expected value is manageable. The corporation therefore purchases contractual rights to third-party indemnification, retains selected losses and arranges primary and additional insurance layers to sustain the operational consequences of severe events. The economic problem is familiar in the literature on the allocation of uncertain losses; the contractual architecture of an actual corporate program adds distinct complications that cannot be represented by a single aggregate coverage number.
The California commercial insurance guide describes the separate general liability aggregate and products-completed operations aggregate within commercial general liability coverage and discusses commercial umbrella insurance above underlying liability. The NAIC insurance glossary identifies commercial umbrella and excess coverage as responding above basic insurance, self-insurance or stated retentions, while recognizing possible gaps in basic coverages. These institutional descriptions establish the basic distinction between an underlying contract, the level at which another insurer may become responsible and the amount potentially available from that insurer. They do not settle how any particular occurrence is covered.
The empirical background is changing. The FERMA/PwC 2026 survey documents respondents' reported adjustments in insurance purchasing and concern about exclusions and wording changes. The associated survey methodology describes 1,329 respondents across 79 countries, with 78% from Europe, recruited through professional associations and the PwC network. It is a cross-sectional sample of responding practitioners, rather than a probability sample of all corporations or all insurance purchases. Accordingly, its descriptive figures identify current reported practices and perceptions but cannot be treated as universal adoption rates or estimates of the causal effects of particular retention strategies.
This dossier investigates six related questions. First, which conceptual elements determine an insurer's layer response? Second, which contractual words determine whether, and when, attachment has occurred? Third, how should a corporation compare higher retention with higher premium and additional limits? Fourth, how do multiple claims, shared annual limits and different coverage lines change the actual capacity available at the time of loss? Fifth, what can comparative law and insurer supervision explain, and where must the contract itself govern? Sixth, how can a decision maker distinguish measurable risk improvement from the appearance of improved protection?
The inquiry combines primary legal authority, examples of disclosed insurance wording, regulator guidance, original mathematical derivations and established theoretical research. The numerical calculations below are stylized, with probabilities and premiums chosen solely for reproducibility. They are not estimates of any real insured's risk, an insurer quotation or jurisdiction-specific legal advice. The relevant actuarial discipline is expressed in ASOP No. 53, which addresses prospective costs of property/casualty risk transfer and retention, including losses, expenses, reinsurance and cost of capital. Its jurisdictional reach is professional actuarial practice governed by the U.S. Actuarial Standards Board, rather than an international statute.
Foundations, Definitions, and Institutional Development
The policy tower as a contractual capital structure
A policy tower is an ordered or partly ordered collection of insurance layers arranged around a particular insured exposure. In an ordinary severity tower, the insured or captive retains a first amount, the primary insurer accepts the next interval and excess insurers assume higher intervals. A layer commonly described as “$10 million excess of $5 million” has a nominal attachment point at $5 million and a maximum indemnity of $10 million for a covered event, subject to its governing wording. Where the tower contains multiple excess policies, a high layer may be excess over scheduled primary and lower excess insurance.
An analytically useful representation identifies each layer by an attachment threshold, an occurrence or claim limit, an annual aggregate limit where applicable, a coverage trigger, an underlying-insurance schedule, a list of covered and excluded loss components, payment and exhaustion requirements, and the applicable legal entity, period and territory. The words “attachment” and “exhaustion” concern different logical operations: attachment can refer to the level in the loss distribution at which a layer is positioned, whereas contractual exhaustion concerns satisfaction of the conditions under which that layer becomes payable.
In the SEC-disclosed excess policy specimen, the follow-form grant incorporates terms of a scheduled followed policy except where modified and conditions attachment on exhaustion of aggregate underlying limits through payments of covered loss by or on behalf of the relevant insurers or insureds. This documented wording is particularly informative because it identifies both the scheduled underlying policies and who may make the payment that satisfies the contractual condition. It is a specimen from a specific filing, not a universal standard clause for every commercial line.
The tower resembles a financing structure in the limited sense that different parties bear different portions of a contingent obligation. That analogy has boundaries. An insured loss must meet the insuring agreement and survive applicable exclusions and conditions; insurer liability does not arise solely from an economic loss crossing a numerical threshold. A policy's erosion by defense expenses, treatment of settlements, other-insurance conditions and disputes about common causes can materially alter the nominal sequence.
Primary, excess, umbrella and self-insured structures
Primary insurance ordinarily attaches at the first contractually covered level after applicable deductibles or retentions. A deductible is an amount identified in the policy that the insured economically bears, commonly with the insurer retaining defined claims-handling and payment responsibilities followed by reimbursement, although wording and line of business vary. A self-insured retention ordinarily requires the insured to bear or fund specified covered loss before the insurer's obligation attaches; it can also change defense obligations and the identities of parties who may satisfy the retention. Their economic resemblance does not make their claims mechanics interchangeable.
The California Department of Insurance explanation expressly introduces commercial umbrella coverage that can extend over commercial automobile, commercial general liability and other liability lines; where the umbrella covers a gap without a corresponding underlying contract, a self-insured retention may apply. This illustrates why “excess” and “umbrella” must be defined by the contract, rather than inferred exclusively from their marketing labels. An excess follow-form layer generally follows identified underlying coverage subject to its modifications; umbrella wording may also include broader grants or exceptions. Neither label supplies an automatic universal answer on drop-down, defense, exclusions or exhaustion.
Reinsurance is a separate contractual relationship in which an insurer transfers part of its risk to a reinsurer. Although reinsurers may employ layers and attachment points similar to those found in corporate programs, the insured corporation's direct rights under the original policy and the reinsurer's liability to the ceding insurer are distinct. The Lloyd's market overview describes specialist syndicates, broker distribution and subscription underwriting as mechanisms for distributing and pricing risk. This market structure can supply capacity across layers or participants, while the allocation of legal obligation remains governed by the issued policies and relevant regulatory arrangements.
The corporation's objective and institutional constraints
An insurance tower is selected among feasible contracts, not from an unconstrained mathematical continuum. Underwriters choose available terms, upper layers can have minimum premiums, different carriers may demand different primary wording, and counterparties can impose minimum limits under loan, lease, construction and supply arrangements. The corporation also confronts maximum tolerable retained loss, available cash and credit, claims-continuity needs and insurer credit exposure. The decision consequently includes expected cost and the shape of extreme outcomes, duration of payment disputes, contractual uncertainty and operational requirements.
The economic foundation of deductible insurance is developed in Arrow's insurance discussion, whose treatment of risk aversion and premium loading explains why full protection against small claims need not maximize a risk-averse buyer's utility when insurance has a positive loading. Raviv's optimal-policy article extends the theoretical analysis of contract design, including indemnity and claims cost considerations. These are theoretical results under specified assumptions. Neither provides an off-the-shelf deductible for a complex corporation with heterogeneous exposures, layered legal contracts and capital constraints.
Recent theory also illustrates uncertainty about the distribution itself. In Gollier's ambiguity analysis, the form and comparative attractiveness of an insurance contract depend on the structure of ambiguity and preferences toward it. Thus a single fitted loss distribution cannot mechanically determine an optimal upper-layer purchase. The appropriate response to model ambiguity is to examine materially different plausible distributions and the contract outcomes they imply.
Historical development as a sequence of contract refinements
Layered commercial insurance grew through the division of risks into limited participations, the separation of primary and excess obligations, greater specification of aggregate limits and the development of specialist markets. Rather than assigning a speculative single point of origin to the modern tower, the verifiable institutional history consists of several streams: co-insurance and market subscription; the economics of deductible and risk sharing; and judicial interpretation of words in excess policies.
The older U.S. case of Zeig, discussed and distinguished by later courts, addressed whether a policyholder's compromise settlement below underlying limits necessarily barred access to excess insurance. Later decisions increasingly focused on precise exhaustion wording. In Ali v. Federal Insurance, the U.S. Court of Appeals for the Second Circuit distinguished mere liability sufficient to cross the attachment amount from exhaustion through payment of covered losses required by the particular contracts. In Fireman's Fund v. OneBeacon, the same circuit analyzed different excess-policy wording in a dispute between an insurer and its reinsurer, holding that a below-limits settlement could satisfy exhaustion where the operative terms did not specify an exclusive payment mechanism. Its separate reinsurance holding applied a follow-the-settlements clause to a reasonable allocation. The juxtaposition demonstrates that doctrinal evolution and different facts cannot be collapsed into a blanket jurisdiction-wide rule. Actual clauses and the court's interpretive task matter.
Attachment and Aggregate Limits
Layer arithmetic and insured indemnity
For a single covered loss amount X, a hypothetical layer with numerical attachment a and limit l can be represented by the capped excess function
The expression is an arithmetic idealization. It assumes an eligible indemnifiable loss on a consistent monetary basis, a single occurrence, no aggregate depletion, no defense-cost erosion, no co-insurance, no exclusion, and satisfaction of every relevant contractual condition. It describes a prospective amount within a layer before interpreting whether the insurer is legally obligated to pay it.
Suppose the corporation retains the first $1 million of a covered event, a primary layer covers $4 million excess of $1 million, the first excess layer covers $10 million excess of $5 million, and a higher excess layer covers $15 million excess of $15 million. For an otherwise eligible $30 million occurrence, the stylized allocation is $1 million retained and $29 million across the three insurance layers. For a $35 million occurrence, the same tower leaves an additional $5 million above the $30 million ceiling. These dollar amounts follow exclusively from the assumed numerical architecture. The calculation makes no claim about the actual availability of any issued policy.
The relevant event must first be translated from a gross economic loss into each contract's “loss,” “claim,” “occurrence,” “damages,” defense expense, insured person and covered property definitions. A corporate incident can generate property destruction, third-party bodily injury, contractual penalties, business interruption and notification costs. The components may invoke different policies, periods, exclusions and retentions. Counting the same economic loss as available under every listed line overstates aggregate indemnification.
Occurrence ceilings, annual aggregates and reinstatements
A per-occurrence limit constrains payment arising from a contractually defined occurrence or claim. An annual aggregate constrains the cumulative payment made for designated claims within the policy's relevant aggregate period. The California CGL guidance identifies both general and products-completed operations aggregate mechanisms; it also explains that exhaustion may end further payment for the affected coverage during the policy term. The separate existence of aggregate categories is significant: exhaustion of one bucket cannot simply be assumed to exhaust every bucket of the policy.
For a simplified annual sequence of eligible claims x_1,...,x_n subject to the same retention r, occurrence cap l and annual cap G, total insurer payment may be represented as
This mathematical structure is suitable only when the policy applies those limits in the assumed sequence. Real claim allocation can depend on paid-versus-incurred accounting, defense expenses, claims-made reporting, batch clauses, sublimits, aggregate corridors and reinstatement premiums. The order of occurrence allocation may matter operationally where competing claims and disputed coverage draw on the same remaining aggregate.
For example, suppose a primary insurer covers up to $5 million per event but only $6 million in the annual aggregate. Two separate otherwise covered events of $4 million each imply at most $6 million of primary indemnification over the two events, leaving $2 million without further payment from that primary aggregate. Whether a specific excess policy responds to the residual $2 million, requires insured payment, remains above a fixed retained amount or denies a drop-down is a contractual question. The primary aggregate arithmetic by itself does not trigger the upper policy. The distinction becomes especially important when the excess policy's attachment is written in terms of an underlying aggregate rather than a per-occurrence loss.
A reinstatement restores a defined amount of exhausted capacity, often subject to conditions and additional premium; it is not the same as a fresh attachment point or a new policy year. A thorough tower schedule records available reinstatements, when they arise, whether they apply per event or in the aggregate, and whether an annual aggregate loss can create a temporary unprotected interval before reinstatement becomes effective.
Two independent attachment gates
Actual excess recovery commonly requires two linked demonstrations. The loss gate asks whether an insured loss is of a type covered under the excess policy, within its term and territorial scope, and at a magnitude that reaches the layer's numerical band. The exhaustion gate asks whether the scheduled underlying insurance has been reduced or exhausted by the precise mechanism required by the excess policy. The two gates can produce different results for the same gross loss.
Consider $12 million of potentially covered liability, an underlying $5 million limit and $10 million of excess capacity above it. If the underlying insurer settles for $3 million and the corporation pays $2 million toward the remaining underlying band, arithmetic loss above $5 million equals $7 million. In a contract accepting covered-loss payments by or on behalf of the insured, these payments could satisfy an appropriately worded exhaustion clause. In a contract demanding payment of the full underlying limits by specified underlying insurers, identical financial flows may fail the exhaustion gate. No general conclusion is possible without the wording, applicable law and characterization of the insured's payment.
The SEC specimen's attachment clause expressly admits covered-loss payments by or on behalf of insurers or insureds, and therefore documents one feasible drafting solution to a recurring settlement problem. By contrast, the policies discussed in Quellos v. Federal Insurance required actual payment of the underlying limits by the specified insurers. The Washington Court of Appeals applied that wording to reject attachment following settlements that had not met those contractual requirements. This outcome concerns those particular policies and applicable judicial interpretation.
Defense expenses, sublimits and the effective top of the tower
The amount nominally described as a “limit” can correspond to different financial resources depending on defense treatment. Where defense expenses erode liability limits, high defense spending reduces capacity available for settlement or judgment. Where specified defense is outside the limit, the nominal indemnity cap can remain intact, but coverage and payment of defense may still depend on other conditions. Any scenario model that treats legal expense as automatically paid above indemnity without reference to the contract can materially understate residual losses.
Sublimits can produce a local ceiling lower than the advertised occurrence or aggregate limit, including restrictions for identified perils, activities, territories or loss components. Likewise, a participant's share on a subscription placement can differ from the total layer amount. Program valuation should use the maximum contractually payable amount for the specific event and legal entity, rather than simply summing the face values of certificates or marketing summaries.
The optimal upper limit is therefore event-dependent. For one covered severity pathway, successive contractual limits may sum into an economically meaningful tower. For another loss, a sublimit, exclusion, claims-made trigger, aggregate exhaustion or underlying insurer payment clause can interrupt that sequence. The existence of several layers is evidence of several contractual promises, not conclusive evidence that the promises operate as one continuous unbroken indemnity.
Cross-Policy Exhaustion Scenarios
Exhaustion by underlying payment versus exhaustion by economic loss
When an event is above the stated attachment, the corporation may expect that the excess carrier will pay the difference between gross loss and attachment. Whether this occurs depends on the exact policy. The Second Circuit's Ali decision construed clauses requiring exhaustion by payment of underlying losses and rejected the insureds' position that unpaid liability reaching the attachment amount was sufficient. The decision therefore supplies authoritative evidence for a contractual-payment gate in the particular policies before that court.
The California appellate court in Qualcomm v. Certain Underwriters also enforced the excess policy's actual-payment exhaustion requirements, distinguishing the more permissive reasoning associated with Zeig. The Washington court's reasoning in Quellos likewise relied on precise text demanding payment by underlying insurers. These decisions show a recurring litigation risk: agreement on total economic damages does not establish that the contractual path to excess recovery has been completed.
The conclusion changes with different wording. In Fireman's Fund v. OneBeacon, a dispute about the insurer's excess-policy allocation and its reinsurance recovery, the Second Circuit found that an underlying below-limits settlement could suffice under an exhaustion provision that did not mandate a particular payment mechanism; it differentiated the wording interpreted in Ali and separately applied the reinsurance follow-the-settlements provision. This source comparison supports a practical editorial conclusion: a research or purchasing analysis must quote the operative exhaustion sentence and identify the permitted payer, payment type and relevant scheduled insurance. References to “vertical attachment” without those qualifiers are incomplete.
An explicit settlement comparison
Consider a scheduled primary limit of $5 million and an excess policy nominally providing $10 million above it. Assume a covered loss ultimately valued at $9 million. The primary carrier compromises its claim for $3.5 million; the policyholder absorbs $1.5 million of the original primary band; the remaining $4 million of economically eligible loss lies above the numerical attachment. Three illustrative clauses could produce three different results.
Under wording requiring the underlying insurer itself to pay its full $5 million policy limit, the $3.5 million settlement would not by itself satisfy the exhaustion condition. Under wording accepting covered-loss payments by or on behalf of the insurer or insured until the scheduled $5 million level is exhausted, the combined $5 million payment might satisfy the clause, subject to classification of the insured payment and any other conditions. Under a clause interpreted as satisfied by proof of loss above the underlying limit without an exclusive actual-payment requirement, the $4 million excess band could be reached despite the settlement discount. These are conditional contract interpretations, not predictions of outcomes in particular lawsuits.
The disclosed follow-form excess wording supplies a concrete example of the second kind of specified payment language. Its scope cannot be assumed to extend to an unrelated commercial property or liability policy. In a real claims file the dates, release terms, claim eligibility, allocated expenses and source of each payment must be established before any payout model can be finalized.
Horizontal and vertical exhaustion across policy years
A corporation with long-tail bodily injury or environmental liability may have multiple triggered policy years. The resulting insurance structure is a matrix of time and attachment: primary and excess contracts for several years, with distinct insurers, limits, prior-insurance provisions, allocation rules and defense obligations. A numerical annual tower diagram does not determine whether the insured can ascend one year's layers before using lower layers in other years.
The California Supreme Court in Montrose Chemical v. Superior Court addressed whether the insured could seek excess indemnification through vertical exhaustion of insurance for a particular policy period, rather than first exhausting all lower-level coverage across every triggered period. The court's analysis focused on the language of the excess policies in a continuous-injury context; a later modification clarified that primary exhaustion was not an issue presented in the dispute. The case supports careful attention to the temporal matrix and the precise question adjudicated.
The Second Circuit's Olin v. OneBeacon opinion applied New York's all-sums and vertical-exhaustion analysis in a context of relevant prior-insurance language, while addressing the prevention of duplicate recovery across policies. That reasoning reinforces a basic accounting principle: an insured may have several potential indemnitors for a loss, but may not count the same dollars more than once as a net economic recovery.
These authorities cannot be transported wholesale to unrelated claims-made financial lines, occurrence-based property coverage or non-U.S. jurisdictions. They show how the meaning of a tower changes when loss occurrence, injury progression, trigger and allocation span time. A cross-year model should include a loss-to-year mapping and separate insurance remaining by period and layer, followed by the governing legal allocation and exhaustion rules.
Different coverage lines after a single enterprise event
A major corporate incident may produce liability claims under commercial general liability, property damage and business interruption claims under property forms, a cyber claim for information loss and financial-line claims against directors. One physical or operational event does not create a common attachment or common aggregate unless contracts expressly connect them. Separate causes of action and different insured persons may trigger different dates or definitions, and one policy's payments may affect another only where a contract provision or governing law establishes that result.
A useful event ledger records each loss component once, identifies potential applicable policies, tags the applicable limits and claims period, and then allocates expected recoveries subject to anti-duplication requirements. The analyst should separately track contested coverage and the probability, duration and amount of recoveries. Even if multiple policies are implicated, the corporation's unindemnified net obligation is determined after legal allocation and overlapping payment credits.
For example, a hypothetical facility failure could yield $12 million of property reconstruction cost, $6 million of business income loss and $8 million of third-party liability. Summing the enterprise gross exposure gives $26 million. Without reading the property and liability contracts, it is invalid to place the full $26 million atop a single “corporate tower.” A property policy might exclude part of the business income, a CGL policy might exclude damage to the insured's own property, and separate retention and aggregate limits may govern the liability. The correct model treats three exposure accounts and then documents legally valid interaction.
Annual aggregate exhaustion and a high-severity second claim
Suppose one policy has a $5 million per-occurrence cap and a $6 million annual general aggregate, with no reinstatement. An initial eligible occurrence is paid at $4 million, leaving $2 million of aggregate. A later eligible $7 million occurrence could receive at most $2 million from this primary policy despite its $5 million per-occurrence headline. If the umbrella attaches after scheduled underlying exhaustion, a later payment may depend on whether the upper insurer recognizes aggregate exhaustion, whether a specified retention applies to the second event, and whether the umbrella is required to drop down to take the exhausted primary's place.
This example reveals a difference between exhaustion of the primary insurer's annual aggregate and the point at which a single high-severity event passes a per-event attachment. A limit schedule showing only “primary $5 million plus excess $10 million” omits a variable critical to second-event protection. A robust annual simulation should process the claims chronologically, reduce only the applicable aggregate buckets and re-evaluate the subsequent contractual gate after each paid claim.
Credit impairment, missing underlying insurance and timing
A lower insurer's insolvency can affect cash recovery without changing the underlying loss severity. Some excess forms state that failure to maintain underlying insurance leaves the upper insurer liable only as if the scheduled insurance remained in force; others use different language. The corporation can thus experience a funding gap at the nominal lower layer even while the upper insurer's attachment position remains contractually unchanged.
Payment timing is a separate risk. A claim can be finally valued in an amount above a layer's attachment while being unpaid due to settlement negotiations, litigation or contested allocation. The liquidity consequences belong in corporate risk-finance analysis even when the insurer ultimately pays. A funding stress test should therefore distinguish (i) covered loss recognized, (ii) loss agreed or judicially established, (iii) underlying amount actually paid and (iv) excess indemnity actually collected.
The New York DFS filing instructions state, under their filing-guidance heading for excess policies, that excess payment should not be contingent upon payment of the underlying policy. This regulatory filing instruction must be identified by its administrative context and jurisdiction; it is not a universal judicial holding about all U.S. excess contracts. Read alongside the cases and SEC-disclosed specimen, it illustrates why regulatory acceptability, policy drafting and case-specific litigation are distinct sources of contractual risk.
Comparative Validation and Source-Critical Analysis
United States: wording, state regulation and case law
In the United States, commercial policy architecture is shaped by state insurance regulation, issued contracts, governing law and the case law interpreting the wording. The national NAIC glossary and state guidance offer useful classifications, but they neither unify all policy forms nor replace state-specific decisions. The contrast between the New York filing guidance and judicial cases is a source-critical warning against treating a regulator's summary instruction as though it resolved every private dispute under a different policy and applicable law.
The reviewed judicial materials show at least three distinct analytical problems. The first is the content of the exhaustion requirement, where Ali, Qualcomm and Quellos enforced explicit payment language, while Fireman's Fund analyzed less restrictive wording. The second is allocation across multiple policy years, discussed in Montrose and Olin. The third is the validity of the underlying loss itself and its consistency with follow-form exclusions, which remains a separate condition even where all financial thresholds have been met.
The comparison deliberately avoids claiming that a single state or circuit has adopted an unchanging universal rule for all policies. The reported outcomes depend on the claims, forms, insurance years and questions before the courts. For empirical work, a judicial case is evidence of a decided dispute under specified law, not a random sample of how frequently insureds experience exhaustion disputes.
United Kingdom: contract formation and claims-related terms
In the United Kingdom, the Insurance Act 2015 addresses fair presentation and remedies for breach in non-consumer insurance, as well as warranty-related rules. Section 10 changes the effects of warranty breaches, and section 11 restricts insurer reliance on certain risk-reduction terms where noncompliance could not have increased the risk of the loss that occurred. These rules may influence whether a loss is covered before any attachment arithmetic is applied. They do not supply a universal excess-insurance attachment formula or displace every individually negotiated exhaustion condition.
The Third Parties (Rights against Insurers) Act 2010 creates mechanisms by which qualifying third parties obtain and enforce certain rights against insurers when the insured becomes insolvent or falls within defined categories. It provides a useful reminder that corporate insolvency can affect the holders and enforcement of insurance rights. Its operation depends on the statutory circumstances and applicable territorial rules; it should not be recast as a blanket right for a corporate insured to bypass underlying payment requirements.
These statutory sources have greater authority for the legal propositions they address than trade commentary. Their relevance to a specific tower still depends on governing law, contract date, the nature of the insured and whether the statutory provision applies. A comparative dossier should record the exact text version and avoid inferring that British non-consumer contract rules automatically govern a policy written for an overseas subsidiary.
European solvency and international supervision
European insurance solvency regulation addresses the undertakings' valuation and capital requirements rather than prescribing a unique insured's optimal attachment. The Solvency II technical provisions rule provides general requirements for insurance obligations to policyholders and beneficiaries; Article 77's valuation formulation uses a best estimate and risk margin framework for technical provisions. These are insurer-side requirements, and direct application to a corporate buyer's optimal retention would confuse the insurer's liability valuation with the buyer's purchase decision.
At the international supervisory level, the IAIS Insurance Core Principles provide a framework for insurance supervision, with ComFrame addressing internationally active insurance groups. They inform the institutional environment in which insurers manage underwriting, solvency, governance and risk transfer. Their generality makes them appropriate for comparative supervisory context, not for attributing an insurer's particular exhaustion requirement to an IAIS rule.
The ISO 31000:2018 guidelines frame risk management as a process spanning identification, evaluation, treatment and monitoring. Used in the corporation, that perspective supports combining exposure mapping with financing and later claims reviews. ISO 31000 is a general risk management guideline rather than a standard insurance tower wording or a certifiable insurance policy form. It therefore contributes process logic without determining contract indemnification.
Financial statements and the limits of accounting analogy
The IFRS 17 standard overview explains the measurement, presentation and disclosure of insurance contracts within its scope, including fulfillment cash flows and contractual service margin. It became effective for annual periods beginning on or after January 1, 2023. For a commercial policyholder, accounting for purchased insurance, possible claims recoveries and other corporate obligations requires analysis under the standards applicable to that reporting entity; IFRS 17's insurer-focused model is not a generic formula for recognizing every buyer's insured asset. Contractual collectibility and economic stress calculations likewise remain distinct from financial statement recognition.
The important comparison is between three registers: an exposure register describing possible gross loss; a contractual register describing possible insurer obligations; and an accounting register describing recognized assets, liabilities and expenses under applicable reporting rules. An enterprise may suffer a $20 million loss, have a litigated $15 million claim against insurers and recognize a different amount or at a different time in its financial statements. Calling all three figures “the insured amount” obscures the decision problem.
Evidence hierarchy and the absence of a global uniform dataset
The highest-confidence inputs for a particular tower are executed policies, endorsements, declarations, loss reports and relevant legal orders. Institutional guidance and judicial opinions supply authoritative context within their stated scope. Survey data reveal reported practice but require scrutiny of sample design and denominators. Economic theory supplies conditional derivations, while numerical illustrations reveal model behavior only under declared inputs.
The FERMA sample is particularly instructive. Its reach to 79 countries creates a useful cross-border perspective, but its 78% European respondent share and voluntary professional-network recruitment limit geographic representativeness. Respondents' selection of risk retention among strategy options does not establish actual changes in balance-sheet funding or future insured losses. An internationally applicable purchasing model should therefore draw parameter estimates from the target enterprise's own credible experience and from relevant sector and location-specific data rather than adopting reported survey frequencies as direct probabilities of claims.
The primary policy specimen in an SEC filing establishes that the words printed in that filing existed in a disclosed contract. It does not prove that every buyer could purchase the same wording, that the clause was ultimately enforceable in every state or that the premium was competitive. The legal cases establish court interpretations of particular provisions and facts. The actuarial standards identify professional analysis and disclosure expectations in their stated scope. The distinction among these kinds of evidence is itself a core result of the comparison.
Internal validation of the hypothetical calculations
Three identities provide basic controls over the analytical examples. First, for nonnegative X and fixed a and l, layer payment lies between zero and l. Second, under a genuinely contiguous single-occurrence tower with full coverage and no duplicative benefit, gross eligible loss equals the sum of the amounts retained and paid by all layers, up to any uncovered amount above the top. Third, expected values of the component payments sum to expected gross loss when the decomposition is exhaustive and disjoint.
The discrete scenario obeys all three controls. The distribution sums to one; E[X]=1.58 million; expected retained and insured amounts sum to 0.22+0.40+0.51+0.45=1.58 million. At X=30 million, the payments are 1, 4, 10 and 15 million; at X=8 million, they are 1, 4 and 3 million, with zero upper excess. All figures are evaluated before premium, legal delay, adjustment expense, credit loss or coverage denial. The arithmetic is therefore reproducible, but its economic completeness is intentionally limited.
A separate aggregate example obeys the cap 4+2=6 million when two underlying claims of 4 million each are subjected to a $6 million annual aggregate. It deliberately leaves the excess insurer's response unspecified because that outcome cannot be derived from the primary limits alone. This is a substantive validation choice: manufacturing an exact excess-payment amount without a governing contractual clause would make the example precise at the cost of accuracy.
Integrated Synthesis and Unresolved Research
What the evidence establishes jointly
The reviewed material supports a coherent interpretation of commercial tower design. The numerical architecture divides contingent financial severity into bands; the issued contracts specify conditions on which those bands become collectible; the buyer's financing system determines which residual losses can be sustained; and the insurer's capacity and claims performance determine the practical value of purchased promises. A corporate program is consequently best evaluated through a mapping from an event to its covered loss components, applicable policies, legally collectible payments and resulting net cash requirements.
The implications follow from separate evidence strands. The California regulatory guide and disclosed excess policy demonstrate that occurrence caps, aggregates, retentions, follow-form conditions and underlying schedules are concrete features of insurance documents. The U.S. appellate decisions show that payment and exhaustion language can change access to the same nominal layer. Actuarial standards explain why estimates require coherent exposure bases, credible claims data and expense and capital treatment. FERMA's survey documents demand for adaptive insurance purchasing but does not establish which design is optimal for an individual company. The integrated inference is that a limit chart must be paired with a clause-level coverage map and a cash-flow stress test before it can serve as a decision tool.
A practical sequence of analytical verification
The research process begins with a source-locked list of legal entities, locations, business activities and loss categories. It then reconstructs a policy schedule for every line, including policy period, trigger, insuring clauses, exclusions, underlying forms, attachment, limits, aggregates, defense treatment, reinstatement and amendments. Next, historical claims and exposure records are reconciled to the relevant definitions, including inflation, reporting development and claims settlement conventions.
A scenario engine can then evaluate each candidate tower under consistent event sets. The event set should include ordinary attritional losses, one large claim, two aggregate-consuming claims, a common-cause multi-line event, long-tail losses across years, a discounted underlying settlement, an insolvent underlying insurer and delayed excess payment. Outputs should report gross loss, eligible loss, paid underlying amount, nominal excess amount, contractually collectible excess, unresolved disputed amount and corporate liquidity requirement separately. Comparative results should be assessed across alternative loss models, not one preferred point estimate.
Finally, the commercial decision must evaluate premiums, capital cost, financing arrangements and counterparty risk using the same underlying scenarios. A lower-premium tower that raises the probability of an intolerable retained cash demand may have a worse total financing outcome than a higher-premium structure. Conversely, purchasing very low attachment insurance for predictable modest losses may use funds less effectively than a credible retained-loss arrangement when administrative loading and corporate liquidity are considered.
Claims governance is part of tower performance
The tower continues to evolve after placement. Notifications, defense management, settlement authority, allocation agreements and the timing of underlying payments can affect when excess insurers become liable. Thus decisions about settlements should include the wording of all potentially implicated excess policies at the outset. A settlement that is economically attractive when evaluated against the primary insurer alone may have a materially different total cost if it changes the ability to satisfy an upper-layer exhaustion condition.
An evidence-preserving claims file should contain the scheduled policies and endorsements, applicable claim definitions, underlying payment history, written positions of carriers, settled and reserved amounts by coverage, allocation methodology and operative legal decisions. It should also specify whether defense costs reduce limits and how disputed claims affect aggregate monitoring. This documentary infrastructure supports both scientific analysis and reliable corporate reporting.
Unresolved empirical questions
The literature and primary materials examined do not establish a universal empirical relationship between attachment level and total corporate risk-financing cost. Several questions require appropriately designed studies. How much of the observed difference in premiums between towers is attributable to expected loss, insurer expenses, competition or risk selection? How frequently do disagreements over exhaustion delay commercial recovery, and how does that frequency vary by clause and jurisdiction? What share of apparent upper-layer capacity becomes unavailable following aggregate erosion, exclusions or credit deterioration? How do firms' liquidity policies change the benefit of high-layer insurance under correlated catastrophe losses? Which disclosure and data-sharing practices materially reduce claim-allocation friction?
The questions call for confidentiality-protected claims and policy records, credible exposure denominators, contract-text classification, event dates and observation periods long enough to study tail claims. Without those inputs, quantitative answers would overstate current evidence. The present monograph defines the relevant distinctions and offers reproductions under declared assumptions; it does not claim original industry-wide causal estimates or a new contract doctrine.
Implications for international comparability
A cross-border program may contain a centrally arranged master policy, locally issued policies, different limits and distinct regulatory requirements. A quoted $100 million aggregate corporate limit is not necessarily transferable across every entity and country on identical terms. The research consequence is to use a matrix of insured entity, insured peril, territorial trigger, underlying requirement, limit currency, payment jurisdiction and period, with explicit aggregation and allocation rules. Any loss model that pools claims without those attributes can produce an internally consistent numerical answer to the wrong contract question.
Supervision and solvency information are necessary for counterparty evaluation, and the IAIS and EIOPA materials provide relevant background. They remain analytically separate from the policyholder's actual entitlement in a specific claim. Theoretical insurance design and standards-based actuarial methods can be transported between countries with appropriate adaptation, while local statutes, policy wordings and judgments require jurisdiction-sensitive verification. The result is a shared scientific vocabulary capable of supporting comparative research without making unlike legal systems artificially identical.
Conclusions
A commercial insurance policy tower is an integrated arrangement of retained loss, primary indemnity and excess contractual promises, with capacities constrained by occurrence limits, aggregate limits, coverage provisions and payment conditions. Its economic design requires evaluating the expected and extreme losses that remain on the corporation's balance sheet together with total premium, liquidity needs and claims uncertainty.
The arithmetic of a layer is transparent: under specified simplifying assumptions, a layer pays the loss above its attachment, capped by its limit, and its expected payment equals the integral of the severity survival function across that interval. The arithmetic does not establish contractual attachment. The reviewed U.S. court decisions show that precise exhaustion language, settlement mechanics and interyear allocation can determine whether excess payment is available. A high nominal limit can therefore coexist with material unrecovered loss.
Annual aggregates and multiple claims create a second dimension of design. A per-occurrence diagram cannot measure the remaining annual protection following earlier payments. Corporate events engaging more than one line create an additional coverage-allocation dimension. Valid analysis models contract-defined insured events and accounts for each underlying payment and remaining aggregate separately.
Retention and premium decisions are inseparable from capital and loss-model uncertainty. The reproduced severity scenario demonstrates that a small high-loss probability can dominate the financial value of an upper layer even when the layer's expected payment is modest. Actuarial standards support disciplined data selection and pricing validation; the risk-manager survey demonstrates reported interest in changing retentions and limits but cannot establish universal success rates.
The comprehensive conclusion is that attachment optimization must be performed on a source-verified, event-specific and jurisdiction-aware representation of the insurance program. A decision-ready tower analysis states exactly which insurer is legally responsible for which covered loss, at which point, under which conditions, and with which effect on the buyer's residual cash requirement. That is the basis for scientifically defensible comparison among alternative commercial insurance structures.
References
Actuarial Standards Board. 2013. Actuarial Standard of Practice No. 25: Credibility Procedures, revised December 2013, effective May 1, 2014. https://www.actuarialstandardsboard.org/asops/credibility-procedures/
Actuarial Standards Board. 2016. Actuarial Standard of Practice No. 23: Data Quality, revised December 2016, effective April 30, 2017. https://www.actuarialstandardsboard.org/asops/data-quality/
Actuarial Standards Board. 2017. Actuarial Standard of Practice No. 53: Estimating Future Costs for Prospective Property/Casualty Risk Transfer and Risk Retention, adopted December 2017, effective August 1, 2018. https://www.actuarialstandardsboard.org/asops/estimating-future-costs-prospective-propertycasualty-risk-transfer-risk-retention/
Actuarial Standards Board. 2019. Actuarial Standard of Practice No. 56: Modeling, revised December 2019, effective October 1, 2020. https://www.actuarialstandardsboard.org/asops/modeling-3/
Ali v. Federal Insurance Co., 719 F.3d 83 (2d Cir. 2013), decided June 4, 2013. https://law.justia.com/cases/federal/appellate-courts/ca2/11-5000/11-5000-2013-06-04.html
Arrow, Kenneth J. 1963. “Uncertainty and the Welfare Economics of Medical Care.” American Economic Review 53 (5): 941–973. Reproduced in Bulletin of the World Health Organization 82 (2) (2004): 141–149. https://pmc.ncbi.nlm.nih.gov/articles/PMC2585909
California Department of Insurance. n.d. Commercial Insurance Guide, sections on Limits of Insurance and Commercial Umbrella. Accessed November 15, 2026. https://www.insurance.ca.gov/01-consumers/105-type/95-guides/09-comm/commercialguide.cfm?itcat=partner_guides&itterm=grow-vol-2&page=3&source=post_page-----2e4e23bbe47b----------------------
European Insurance and Occupational Pensions Authority. n.d. Solvency II Single Rulebook: Directive 2009/138/EC, Article 76, General Provisions. Accessed November 15, 2026. https://www.eiopa.europa.eu/rulebook/solvency-ii-single-rulebook/article-2159_en?prefLang=ga
European Insurance and Occupational Pensions Authority. n.d. Solvency II Single Rulebook: Directive 2009/138/EC, Article 77, Calculation of Technical Provisions. Accessed November 15, 2026. https://www.eiopa.europa.eu/rulebook/solvency-ii-single-rulebook/article-2160_en
Federation of European Risk Management Associations and PwC. 2026. Global Risk Manager Survey Report 2026, published October 5, 2026, including methodology and insurance-management findings. https://ferma.eu/global-risk-manager-survey-report-2026/
Fireman's Fund Insurance Co. v. OneBeacon Insurance Co., No. 20-4282 (2d Cir. Sept. 15, 2022). https://law.justia.com/cases/federal/appellate-courts/ca2/20-4282/20-4282-2022-09-15.html
Gollier, Christian. 2013. “Optimal Insurance Design of Ambiguous Risks.” Toulouse School of Economics Working Paper 12-303, revised January 2013; subsequent article in Economic Theory 57 (2014): 555–576. https://www.tse-fr.eu/publications/working-paper/optimal-insurance-design-ambiguous-risks-1
International Accounting Standards Board. 2017, as amended. IFRS 17 Insurance Contracts, effective for annual reporting periods beginning on or after January 1, 2023. IFRS Foundation standard overview. https://www.ifrs.org/issued-standards/list-of-standards/ifrs-17-insurance-contracts/
International Association of Insurance Supervisors. 2024. Insurance Core Principles and ComFrame, including December 2024 updates. https://www.iais.org/activities-topics/standard-setting/icps-and-comframe/
International Organization for Standardization. 2018. ISO 31000:2018, Risk Management — Guidelines, second edition, confirmed 2023. https://www.iso.org/standard/31000
Lloyd's. n.d. Understanding the Lloyd's Market. Accessed November 15, 2026. https://www.lloyds.com/about-lloyds/our-market
Montrose Chemical Corp. of California v. Superior Court, 9 Cal. 5th 215 (Cal. 2020), opinion April 6, 2020, modified May 27, 2020. https://law.justia.com/cases/california/supreme-court/2020/s244737.html
National Association of Insurance Commissioners. n.d. Glossary of Insurance Terms, entry “Umbrella and Excess (Commercial).” Accessed November 15, 2026. https://content.naic.org/glossary-insurance-terms?id=37
New York State Department of Financial Services. n.d. Property Insurers: Rate and Form Filing Instructions, headings on other insurance, excess policies and deductibles. Accessed November 15, 2026. https://www.dfs.ny.gov/apps_and_licensing/property_insurers/rate_form_filings
Olin Corp. v. OneBeacon American Insurance Co., No. 15-2047 (2d Cir. July 18, 2017). https://law.justia.com/cases/federal/appellate-courts/ca2/15-2047/15-2047-2017-07-18.html
Qualcomm Inc. v. Certain Underwriters at Lloyd's, London, 161 Cal. App. 4th 184 (Cal. Ct. App. 2008). https://law.justia.com/cases/california/court-of-appeal/2008/d050433/
Quellos Group LLC v. Federal Insurance Co., No. 68478-7-I (Wash. Ct. App. Nov. 12, 2013). https://law.justia.com/cases/washington/court-of-appeals-division-i/2013/68478-7.html
Raviv, Artur. 1979. “The Design of an Optimal Insurance Policy.” American Economic Review 69 (1): 84–96. https://ideas.repec.org/a/aea/aecrev/v69y1979i1p84-96.html
U.S. Securities and Exchange Commission, EDGAR. 2025. Excess Insurance Policy, Follow Form Excess Coverage, filed as part of a Form 40-17G disclosure. https://www.sec.gov/Archives/edgar/data/1851077/000119312525217116/d933694d4017g.htm
United Kingdom. 2010. Third Parties (Rights against Insurers) Act 2010, chapter 10, as enacted. https://www.legislation.gov.uk/ukpga/2010/10/enacted
United Kingdom. 2015. Insurance Act 2015, chapter 4, as enacted, including sections 10 and 11. https://www.legislation.gov.uk/ukpga/2015/4/enacted?timeline=false&view=extent