RESEARCH DOSSIER
Binding Authority Agreement Design and Risk Limits
Abstract. A binding authority agreement allocates decision rights through which an intermediary can commit insurance capacity within a defined contractual mandate. Its design determines how an insurer's risk appetite is translated into permitted business, policy forms, pricing discretion, territorial permissions, portfolio limits, reporting obligations and exit arrangements. This monograph examines the binding agreement as a mechanism of insurance governance, using comparative analysis of Lloyd's requirements and guidance, a historical model wording, the United States managing general agent model law and a North Carolina statutory example, European Union outsourcing rules and international supervisory principles. Agency economics and transaction-cost scholarship provide the disciplinary background. The analysis distinguishes underlying policy liability, individual underwriting authority, premium throughput, cumulative exposure and claims settlement authority. Reproducible hypothetical cases demonstrate the consequences of delayed reporting, overlapping assignments and concentration. The central finding is that effective delegation depends on the coherence of the contract, the data that measure compliance and the mechanisms that intervene before authority is exhausted. Termination must preserve the lawful servicing of obligations already undertaken. No proprietary transactions or causal estimates are presented.
Keywords: Binding authority; delegated underwriting; managing general agent; coverholder; underwriting authority; aggregate exposure; insurance contracts; authority limits; bordereaux; agency risk; runoff; insurance governance.
Introduction and disciplinary foundation
The object of delegated underwriting
An insurer commits financial capacity to a promise of contingent performance. Its underwriting decisions determine the contractual conditions under which that commitment arises. Delegation separates the location of the underwriting decision from the institution whose capital supports the promise. The binding authority agreement is the contract through which that separation becomes operational: it identifies the intermediary, assigns a field of authorized action and prescribes the conditions under which an insurance commitment may be made.
A specialized intermediary may possess better knowledge of a local trade, a particular industrial process or a distinctive population of risks. The carrier may possess superior access to diversified capital, actuarial analysis, reinsurance and claims infrastructure. An arrangement combining these capabilities can improve access to insurance and reduce the cost of evaluating risks. It also creates a dependence on information held outside the carrier and introduces a boundary between business decisions the delegate can take and decisions reserved to the principal.
The Lloyd's Europe wording guidance distinguishes the delegation contract from the policy issued to the insured. The first regulates the relationships through which binding takes place; the second defines the insured event and the carrier's obligation. The distinction is fundamental to an analysis of excess authority. A contractual breach between intermediary and insurer and the enforceability of a policy against an insurer are different legal questions. Their answers may diverge under governing law and the facts of the transaction.
Economic foundations
Jensen and Meckling's (1976) analysis of agency costs provides a general language for studying decisions made by one party with financial consequences borne partly by another. In their original article (pp. 308–309), the authors distinguish monitoring expenditure, bonding expenditure, and residual loss. These components separate the cost of controlling an agent from the economic consequences of decisions that remain misaligned after controls. In a binder, underwriting rules, compensation arrangements and reporting obligations can act as monitoring devices; their cost must be considered alongside residual underwriting and conduct risk. Its subject is the firm and its ownership structure; application to insurance binding authority is the present monograph's analytical interpretation.
A delegate's interests and a carrier's interests can coincide on sustainable underwriting but diverge on the timing of income and loss. A commission received when premium is written has a different timing from a liability loss paid years later. A carrier may depend on business volume to cover fixed costs, while the intermediary may have greater opportunities to expand its own revenue by accepting additional risks. A narrow mandate controls some of that divergence; monitoring, remuneration and credible intervention address the remainder.
The choice between narrow rules and discretionary delegation has a transaction-cost dimension. Regan and Tennyson (1996), in Agent Discretion and the Choice of Insurance Marketing System, examine how distribution structures relate to risk matching and compensation. Their reported abstract findings concern independent and exclusive agency arrangements. Regan (1997) relates vertical integration in insurance to product complexity, uncertainty and relationship-specific investment. Regan's empirical design uses a cross-section of 149 property-liability insurance groups operating in 1990, including 112 independent-agency groups and 37 exclusive dealers. The analysis uses logistic regression, with distribution organization as the dependent variable. Product composition and historical accounting measures serve as proxies; the observations do not identify the effect of particular contract clauses. These results provide conceptual context for specialization. They do not establish optimal numerical binder limits or the causal superiority of delegated underwriting.
The economic issue is the total cost of allocating decisions. A referral of every unusual exposure may reduce unauthorized acceptance while increasing delay and leaving valuable information unused. Expanding discretion can improve the use of specialist information but magnifies the consequences of weak monitoring. The contractual problem is to identify the decisions that are inexpensive and reliable to devolve, the decisions that require carrier judgment and the conditions that change this allocation over time.
Method and evidence boundary
The research combines functional clause analysis, comparative legal and supervisory interpretation, and deterministic illustrative arithmetic. The primary-source population consists of publicly accessible market requirements, official guidance, a model binding agreement, one US model law and one enacted state example, EU prudential provisions and IAIS standards. The academic literature is purposively selected for its relevance to distribution organization and delegated decision incentives. This is a structured interpretive inquiry rather than an exhaustive systematic literature review.
The unit of institutional comparison is a contractual function: authority, information, exposure, compensation, intervention or continuing service. The unit of quantitative illustration is a hypothetical insurance commitment and the portfolio into which it enters. No population of actual contracts is sampled; the historical FCA inspection provides a separate observed example, with its own limitations. All money in the worked examples is assumed to be US dollars, and all thresholds are invented for transparent arithmetic. They have no regulatory or commercial prescription.
Binding authority legal architecture
Parties and the source of decision rights
A binding arrangement requires a precise identification of who can act and whose liability is engaged. At Lloyd's the managing agent acts in relation to underwriting members organized through a syndicate; the coverholder implements the agreed mandate. Other markets can use an incorporated carrier, an MGA or another authorized intermediary. The designation is meaningful only in conjunction with the permissions, contract and governing law that apply to the legal entities involved.
The carrier's authorization must be traced through the complete chain, including participating capacity and persons authorized to exercise a particular function. A signature from a broker that merely arranges placement cannot substitute for consent from a principal empowered to grant the mandate. A person holding separate underwriting and broking responsibilities must be identified by the capacity in which the person acts at each stage.
Lloyd's contracts of delegated authority requirements specify a range of mandatory agreement particulars, including the relevant participants, business perimeter, financial limits and continuing duties. These are market requirements made under its Intermediaries Byelaw. Their existence supplies evidence of the functions considered essential in that market; their detailed legal force must be understood in that institutional setting.
Contractual authority can be distinguished from administrative permission. The individual permitted to issue a document may be different from the underwriter who accepts the risk. Similarly, claims settlement is a separate delegation of discretion over the carrier's obligation after the loss. A system that grants all permissions to a user merely because the user can quote a policy erases these distinctions.
The agreement and its schedule
A binder has an operative body, a schedule containing selected parameters and, potentially, amendments and incorporated underwriting guidelines. The body establishes legal relationships; the schedule makes their operating boundaries specific. Incorporation of attachments should be controlled through titles, dated versions and an order of precedence, so that a subsequent email cannot silently displace a signed restriction.
A reference to a policy wording repository needs enough information to identify the approved form and endorsements. A reference to an underwriting manual needs a specific approved version and a lawful mechanism for change. A rate table is incomplete if the delegate cannot determine which expenses or adjustments it incorporates. A declaration of a maximum risk limit requires an identified unit: an insured, a premises, a policy, an event, an occurrence or another measurable exposure.
The historical LMA 3113 model wording provides a concrete example of how a scheduled agreement connects classes, territory, individual liability, gross premium and policy duration to its operative clauses. The version examined excludes USA and Canada-domiciled coverholders; subsequent variants must be checked for actual use. The text is used as historical evidence of contractual structure and is not presented as the currently required form.
Actual authority and insurance enforceability
A contract of delegation allocates actual authority between the parties. Whether an intermediary's representation also binds an insurer in a third-party dispute depends on applicable agency and insurance law, evidence of the principal's conduct, the policy formation process and any relevant statutory protections. The rights of an innocent policyholder cannot be inferred from an internal schedule alone.
Consider an authorized delegate with a maximum per-risk commitment of $1 million that purports to grant $1.2 million. Internally, a discrepancy of $200,000 is apparent if the schedule and policy use the same definition. Externally, an adjudicator may need to consider who communicated the apparent grant of authority, whether the insurer confirmed coverage, whether ratification occurred and what law governs. The possible outcomes are jurisdiction-specific. Contract design should anticipate preservation of records and prompt referral to the carrier rather than assume automatic cancellation or automatic coverage.
The evidentiary chain should record who initiated and accepted a transaction, the schedule version then in force, the terms conveyed to the customer and any carrier approval. This permits reconstruction of the authority question without confusing a database field with a legal determination. The carrier should preserve the distinction between remediation for breach of mandate and adjudication of the policyholder's claim.
Version-specific authority and lead identification
The historical form must be distinguished from later public documentation. The LMA's model agreement and endorsement register identifies LMA3113A and separate financial-crime and data-protection endorsements. The listing establishes the existence and identification of the later form; the linked editable document did not render substantive text in the retrieval environment. No uninspected clause in that document is treated as verified.
The LMA's April 1, 2025 lead security endorsement bulletin identifies LMA3184 for several model binding authority agreement families and LMA3185 for service-company underwriting agreements. The bulletin states that delegation agreements incepting or renewing from July 1, 2025 must require identification of the Lloyd's lead in policy documentation. This requirement is market-specific; it demonstrates that agreement version control can affect the information delivered to policyholders, as well as the internal allocation of decision rights. Operational deployment should therefore map a verified contractual amendment to the document-generation rules, effective-date tests and the actual output supplied to insureds.
Contract certainty and amendments
Lloyd's wording expectations connect execution of delegation contracts to the time authority can be exercised. This provides a practical boundary for version control: a proposed agreement, a quotation and a duly authorized agreement are separate states.
An amendment requires specification of its effective time and treatment of transactions initiated before that time. A reduction in authority at noon presents at least three relevant populations: policies already bound, quotations issued but not accepted and requests received afterward. The agreement should define who may honor quotations, on what conditions, and whether existing endorsements continue under the earlier rules. Without temporal clarity, both the policyholder and the portfolio record can be left with contradictory commitments.
A sound administrative record associates every policy transaction with the version of the mandate that governed it. Electronic controls should update at the effective time specified by the contractual amendment. A later change may not rewrite the historic audit trail.
Class, territory and limit schedules
Class definitions and eligible business
The business perimeter identifies risks the delegate may evaluate without individual permission. Class labels are useful only when supported by operative descriptions. Commercial property can include or exclude substantial business interruption, machinery breakdown, catastrophe perils and contingent loss. A policy sold under a familiar marketing name may contain a novel exposure through an endorsement.
The schedule should identify named products, eligible occupancies or activities, insured types, coverage modules, risk characteristics that require referral, exclusions and maximum term. Policy wording and underwriting appetite should use consistent language. An eligibility exception should describe whether an intermediary has permission to quote, may seek approval or must decline. It should also identify who has authority to give the approval.
A schedule that specifies a product but omits endorsements may permit a material change in coverage through a seemingly routine administration operation. For this reason authority over an endorsement should be explicit, especially where it adds a peril, extends a territory, raises a sublimit, alters a deductible or changes a retroactive date. Renewals should be assessed as new decisions unless the mandate and applicable law prescribe another treatment.
Consider an industrial property binder limited to buildings and contents. A request for a six-month business interruption extension might materially increase event severity despite leaving the building sum insured unchanged. A rule based solely on its marketing class or original limit would miss that additional commitment. The exposure definition therefore needs to cover all relevant policy benefits on an analytically consistent basis.
The three territorial dimensions
A location restriction can govern the regulatory location of the insurance risk, the domicile of the insured or the geographical scope of the coverage. Lloyd's wording expectations, cited above, explicitly separate these dimensions. Regulatory permissions and the risk-location analysis are also distinct from the law chosen to govern the binder dispute.
Take a fictional enterprise domiciled in Country A, operating a factory in Country B and purchasing liability cover for activities in Countries B and C. The binder may permit domicile in A but risk only in B, with coverage restricted to B. The domicile restriction is satisfied while the extension to C fails the assumed coverage boundary. A single field labeled territory cannot reliably represent the whole result.
A further distinction concerns the geographical location of the claim event and the place where legal proceedings may be brought. Liability policies may define these on different bases. A geographically precise authority schedule should specify the relevant field for each test and require referral when the necessary information is absent. Regulatory classifications should be determined by the applicable law and regulator's criteria, not inferred mechanically from the address of the broker.
Cross-border contracts create interactions among licensing, tax, distribution requirements, customer documentation and data protection. The carrier and delegate need agreement on who identifies those obligations, how updates are implemented and how ambiguities are escalated. The binder's permission to accept a territory is an internal mandate; it cannot itself create a missing legal permission.
Rates, deductibles and conditions
Pricing authority has several dimensions. A delegate may be restricted to an approved rate matrix, allowed a defined adjustment band, given a discretionary but monitored technical pricing mandate, or required to obtain a specific referral for nonstandard terms. These choices depend on the structure of the business and the capacity provider's appetite.
Assume a manual premium of $10,000 for a particular risk, with an authorized adjustment of plus or minus 10 percent. The permissible range is $9,000–$11,000 if all other inputs remain unchanged. A quoted $8,800 premium requires referral; a $9,200 premium with a modified deductible may also require referral, because the pricing permission is not a general authority to change coverage terms.
The agreement must distinguish premium before and after commissions, taxes and fees. Any deviation calculation needs to start from a consistent price basis and a known model version. Changes to minimum deductibles or exclusions can shift severity without visibly changing the headline premium. An audit of rates therefore needs the complete approved input set and the quoted terms.
An exception should carry the approved decision's identity, conditions, expiry and financial impact. Reusing an approval for other risks introduces a different form of delegation and requires the appropriate consent. The number of exceptions and their directional bias are relevant monitoring variables: frequent downward price deviations may mean the technical rate no longer fits the market or the referral controls are too permissive. Neither interpretation should be assumed without further analysis.
Portfolio risk limits and quantitative tests
Distinct financial quantities
The contractual maximum liability of an underlying policy specifies the carrier's potential legal payment subject to the policy terms. An individual binding limit constrains the size of a commitment a delegate may make. A premium volume limit constrains written production. An accumulation limit constrains the amount of some defined measure of exposure across a group of risks. A claims authority limit constrains the delegate's settlement or decision power after loss. These quantities answer different questions.
For example, an authority to underwrite policies up to $2 million and write $10 million of annual premium does not determine the portfolio's maximum covered loss from a single catastrophe. Premium contains information about price and exposure but neither uniquely identifies severity nor encodes dependence. A short premium cycle and a long claims horizon also create differences between the periods to which each limit refers.
The LMA model discussed earlier defines its gross premium income calculation in section 12 through written premium movements. A real agreement should specify its own accounting basis, including currency treatment, cancellations, additional premium and whether return premium restores authority. An aggregate risk limit needs a similarly explicit definition: sum insured, gross signed liability, expected scenario loss, modeled event loss, net retained loss or some other measure.
Premium limit and reporting latency
Suppose a fictional binder has a $10 million premium limit and an early-warning threshold of 85 percent. The received cumulative bordereau shows $8.4 million. Unreported binding events total $0.7 million, and the delegate proposes another $0.35 million. All figures are on the same assumed gross written premium basis.
| Component | Assumed premium |
|---|---|
| Received and reconciled | $8,400,000 |
| Bound but not received | $700,000 |
| Proposed transaction | $350,000 |
| Total after proposed binding | $9,450,000 |
| Remaining headroom | $550,000 |
The observable warning threshold is $8.5 million. On the report alone, headroom appears to be $1.6 million and the warning threshold is not crossed. Including the unreported transactions, current headroom is $0.9 million and the warning threshold has already been exceeded. The proposed transaction would still be inside the hard premium limit in this specific example.
If a further $0.8 million were bound, the result would be $10.25 million, an excess of $250,000. A periodic report can establish historic compliance but cannot guarantee the headroom available at the moment a binding decision is made. Pre-bind checks require information on already committed transactions; where information is delayed, a conservative reservation, reduced limit or mandatory referral can manage that uncertainty.
For illustration, an allowance for new production during a reporting delay can be estimated using daily binding volume. At $60,000 per day over seven days, an assumed unreported premium reserve is $420,000. This is an arithmetic scenario, not a loss estimate or statistically calibrated buffer. A conservative design would account for variable volumes, exceptional batches, endorsements and cancellations, and decide how uncertainty affects the authority available.
Event concentration and dependence
Three policies have policy limits of $1.5 million each. The binder permits $2 million per policy and all three pass that individual test. They insure buildings in the same flood-prone district. A hypothetical accumulation schedule limits full gross exposure in that district to $4 million. If the metric is the simple sum of these policy limits, the combined exposure is $4.5 million and the third transaction takes the portfolio $0.5 million above the cap.
Now suppose each policy includes a 40 percent line for the carrier and assume the limit is otherwise shared proportionally. The carrier's gross signed exposure is $0.6 million per policy and $1.8 million for all three. Which result governs depends on whether the accumulation schedule actually measures total policy limits, carrier share or another quantity. A test of a total-market limit cannot be satisfied by substituting the smaller carrier-share figure.
A modeled expected or percentile loss is another measurement entirely. An aggregate may depend on limits, deductibles, locations, event definitions, business interruption and uncertain dependence. A full-limit sum is convenient as a conservative exposure indicator in some situations but cannot be described as a modeled probable maximum loss without defined assumptions and evidence. Reinsurance may reduce retained loss, yet recoveries depend on treaty attachment, exclusions, exhaustion, reinstatement and counterparties. The authority schedule must identify whether gross, ceded or net exposure is measured and how treaty eligibility is established.
There is a useful operational separation between the individual threshold used by the intermediary and the portfolio allocation retained by the carrier. A carrier with several binders cannot infer the total geographic exposure from the fact that each individual delegate stays within its own premium authorization. It must allocate portfolio headroom or approve transactions based on a shared, reconciled exposure measure.
Concurrent commitments and currency
Assume a portfolio has $300,000 of remaining gross exposure capacity and two intermediaries independently request $200,000 each. Each request passes an isolated snapshot test; together they exceed the remaining capacity by $100,000. One way to avoid the inconsistency is to reserve capacity when a binding operation begins and commit or release the reservation atomically. A second is to allocate nonoverlapping quotas, each tracked independently. The former requires timely coordinated infrastructure; the latter can strand available capacity in an unused quota.
Currency rules introduce a similar problem. A policy of €900,000 at an assumed conversion rate of $1.10 per euro represents $990,000. At $1.15 it represents $1,035,000. A $1 million authority ceiling therefore has different results under the two rates. The agreement should specify the conversion source and time, the treatment of later exchange movements and the distinction between a historical binding test and ongoing monitoring of current exposure. The example is deliberately hypothetical and does not report actual exchange rates.
An operational control consumes a defined amount of capacity and preserves the ledger that explains why the commitment passed. A change in an exchange rate or exposure estimate can generate a subsequent portfolio breach without demonstrating that the original intermediary acted outside its authority. The monitoring response and contractual breach assessment should reflect that difference.
Threshold architecture
Hard restrictions, warning triggers and referrals perform distinct functions. A hard restriction prevents the intermediary from independently binding a transaction. A warning activates review before the hard threshold is reached. A referral transfers the decision to a person with the required authority. A system should distinguish these outcomes explicitly, because exception approval and expansion of delegated power are not equivalent.
A warning must have an owner, a response time and a consequence. If crossing it merely generates a message no one reads, it has no demonstrable control effect. A referral should present the exact exposure, the rule that triggered the referral and any risk characteristics needed for the decision. The approval should be linked to that transaction and checked again if its inputs change.
The choice of warning level is a question of operational delay, uncertainty, potential severity and appetite. A universal percentage would conceal the different economics of low-frequency catastrophe risks, rapidly written homogeneous products and long-tail liability. Threshold selection is therefore contingent on the exposure definition and the observed speed at which a portfolio changes.
Duties of agent and carrier
Delegation and retained oversight
Lloyd's delegated underwriting guidance treats underwriting accountability as continuing with the managing agent and emphasizes proportionate authority, data and alignment with business appetite. Its guidance is expressly nonprescriptive. It supports a functional reading of governance: dividing work among underwriting, claims, compliance and operations does not remove the need for an accountable carrier decision.
The delegate's capabilities should be assessed against the tasks assigned. Evidence can include staff qualifications, sample underwriting files, documented workflows, data quality, handling of exceptions, cash controls, incident response and service continuity. A specialized intermediary may be suitable for a narrow business class without being suitable for claims authority or product development. Due diligence should match the actual allocation of duties.
The carrier's own resources matter just as much. An insurer that grants a sophisticated pricing mandate but lacks the information or expertise to evaluate performance has retained a responsibility that it cannot adequately discharge. Contracts should allocate the evidence that the carrier needs to understand its exposure and establish escalation routes for underwriting, operations and conduct issues.
Lloyd's oversight and monitoring guidance specifies ongoing monitoring, documented action and registration of applicable delegation contracts, with its stated dispensation qualification. The study's inference is that oversight must be evidenced both in the period when the mandate is established and when it operates: a signed agreement is the starting point for subsequent observations, not a substitute for them.
Reporting as a measurement system
A bordereau is a structured report of transactions or claims associated with a delegated arrangement. Lloyd's reporting standards establish core data requirements for relevant contract types. The scientific design problem beyond the market's specific schema is how to make observations complete, interpretable and reconcilable with the mandate.
Data should allow the carrier to distinguish quotations from binding events, renewals from new business, premium movements from payments and claim movements from cumulative balances. A policy reference by itself may not uniquely identify a transaction, because an endorsement can change limits or premium after inception. A stable risk identifier and transaction sequence permit reliable aggregation and correction.
A sample dataset might contain the binder version, transaction timestamp, policy identifier, risk location, insured characteristic relevant to the class, coverage and wording version, exposure measure, currency, signed share, premium movement, referral decision and cancellation state. Claims reporting should distinguish paid loss, outstanding case reserves, expenses, recoveries and the date to which they relate. The agreement must also address confidentiality, lawful data access and retention.
Reconciliation involves at least three perspectives: the delegate's policy administration records, the carrier's underwriting or finance ledger and the reported bordereaux. Differences may be due to timing, duplication, cancellation, misclassification or a missing transaction. A completeness control checks the source transaction population before relying on summary totals. A quality control records the nature, direction and resolution of each discrepancy.
The monitoring clock should begin at the event that creates the commitment. Measuring delays from the time a report arrives obscures delays between binding and reporting. Median latency alone may overlook a small number of severely late transactions; a distribution of delays and an explicit maximum for relevant incidents are more informative.
Compensation and development uncertainty
The contract may combine base commission, fees for administrative work and a profit-related component. Defining that component requires a coherent financial basis. In a simplified illustration, earned premium of $1 million, incurred losses of $580,000 and allocated costs of $320,000 leave a provisional margin of $100,000. A profit share of 20 percent would be $20,000 on that assumed base. If ultimate losses later rise by $150,000, the margin becomes negative $50,000. Without an adjustment mechanism, the early payment would have rewarded a result that did not survive development.
Loss development, allocation of expenses, reinsurance cost, treatment of recoveries, prior deficits and the timing of settlement are substantive parts of a contingent compensation clause. A short period for settlement may suit one product and distort another. Deferral and true-up mechanisms can reduce timing mismatch, but their cost includes liquidity, complexity and disputes over estimates. An accounting formula needs a defined data source, review process and an allocation of responsibility for corrections.
A performance measure based only on loss ratios can have unintended incentives. Delaying claim recognition or avoiding appropriate settlements can temporarily improve a reported result. Measures of service, claim timeliness, complaints, claim denial patterns and data accuracy help identify possible problems; they do not individually prove misconduct.
The carrier can also create agency problems through its own conduct. Abruptly changing pricing expectations, withholding a decision on a referral or making an ambiguous withdrawal of authority imposes uncertainty on the delegate and on customers. A balanced agreement sets responsibilities for both institutions, including response times and a documented amendment process.
The LMA's August 5, 2025 announcement of model profit commission clauses identifies two structures: a single later payment with overpayment recovery provisions, and staged payments with subsequent adjustments for developing claims and profitability. The announcement supports the distinction between delayed and provisional compensation, although the underlying clause texts were not independently retrieved. The economic choice depends on claim development, liquidity, recovery enforceability and the cost of repeated reconciliation. A short-tail binder may support earlier settlement than a long-tail liability portfolio; contractual cash timing cannot eliminate uncertainty in ultimate loss.
Conduct and observed supervisory evidence
The FCA's historical TR15/7 thematic review examined outsourced general insurance functions and reported weaknesses in contract specification, oversight and customer-focused information. Its 2025 historical designation prevents the report from being interpreted as a statement of current rules. The accompanying FCA findings release identifies 12 insurers and 19 intermediaries and administrators in the review. These are observed supervisory findings from a selected historical sample, not an estimate of the current frequency of failures.
The useful inference is that a financially profitable binder may still produce inadequate service or unsuitable coverage. A carrier needs evidence of whether the product reaches its intended market, policy terms are explained correctly, covered claims are handled fairly and complaints are resolved through the responsible entity. An intermediary with authority over product design or claims may have functions that require separate oversight; the customer's experience spans these contractual boundaries.
A review of outcome measures should consider product mix, eligibility rules, loss experience, exposure differences and time. A rise in claim denials could indicate changes in risk selection, clearer event definitions, inappropriate sales, inconsistent claims handling or some combination. Investigation requires file-level evidence rather than a single KPI threshold.
Subdelegation and audit rights
Chains of authority and capacity
Subdelegation introduces an additional decision-maker between the original capacity provider and the person committing a risk. A chain may also contain brokers, service companies, data providers and administrators who perform supporting functions without independently binding insurance. Contract drafting must classify each relationship by its actual permissions and responsibilities.
Lloyd's master coverholder arrangements guidance distinguishes primary and secondary contracts from master broker services. It describes the relationship between the agreements, underwriter signatures, separate references, capacity monitoring and coordinated runoff. Its qualified treatment of higher secondary authority demonstrates that a layered arrangement cannot be reduced to a universal hierarchy of numerical limits without examining the actual approvals.
In a simplified arrangement, a principal grants a master delegate $5 million in aggregate allocation for its own business and two secondary portfolios. If the master reserves $2 million for its own writing and allocates $2 million to each secondary portfolio, total allocations are $6 million. Each allocation may look plausible in isolation, but the allocations exceed the primary capacity by $1 million if they are intended as concurrent, exclusive amounts on the same measure. The solution is either reallocation or a separately authorized shared-capacity rule with effective coordination. It is not enough to observe that each secondary amount is below the primary ceiling.
A contract should identify the permitted chain, responsibilities for due diligence, the treatment of further appointments, reporting ownership and access to individual risks. A secondary agreement should preserve the evidence needed by the original carrier to reconcile commitments and evaluate the mandate. If an independent carrier directly appoints an additional entity, the actual legal authority must be traced to that appointment rather than assumed to arise from a master intermediary.
Audit scope and reconstructability
An audit examines whether actual conduct conforms to the agreed responsibilities. Lloyd's coordinated audits describes standardized scope and risk-based examination of contracts and claims. Coordination reduces duplicated effort; it does not eliminate a carrier's need to understand material differences among its own mandates.
An effective contract permits access to the records required to reconstruct decisions. Such access includes relevant documents and data, approved schedules, changes, correspondence, payment records and exceptions, subject to appropriate legal safeguards. Audit rights need an identified custodian, timescale, method of access and mechanism for urgent investigations. Access should remain practicable when the relationship is strained or ending.
A file-level audit may take a risk-stratified sample: routine transactions, high-limit submissions, unusual territories, referred applications, amended policies and claims close to settlement thresholds. Each sampled transaction should be matched to the authority version, original inputs, decision and resulting document. An auditor can then distinguish documentation absence, data inconsistencies and a confirmed excess of authority.
Statistical extrapolation requires a probability sample with a known design. Suppose a fictional targeted review selects 30 unusual policies and finds three deviations. The observed rate in that sample is 10 percent, but it cannot be generalized to the entire portfolio because the selection intentionally overrepresents risk. A separate random sample could estimate a portfolio frequency under its sampling assumptions. A targeted sample is valuable because it discovers failure mechanisms, not because it estimates prevalence.
Audit findings need an accountable owner, severity assessment, corrective action and evidence of closure. If an error affected an insured or a financial statement, correction should consider those consequences as well as the underlying control failure. A new system rule may prevent recurrence while leaving historic errors unresolved; both dimensions need tracking.
Disputes and evidence
Dispute-resolution clauses should specify applicable law, forum, notice obligations and the process for interpreting competing schedules or endorsements. A dispute can concern premium remittance, the validity of an approval, the allocation of a claim expense, access to records, profit commission or authority exercised after termination.
A workable dispute process preserves evidence before allocating blame. It establishes the policyholder communication, transaction timestamp, relevant authority, any referral or ratification, and the amount and basis of the contested financial commitment. The decision should distinguish undisputed customer obligations from the parties' internal allocation of responsibility.
Contractual indemnities can allocate some financial consequences between institutions, subject to law and enforceability. They cannot substitute for available insurance capital or automatically determine the insurer's liability to the policyholder. The governance objective is to prevent ambiguous commitments and secure timely servicing of commitments that remain legally effective.
Termination and claims runoff
Ending authority and preserving obligations
A binder's termination concerns the mandate to make or administer commitments on behalf of an insurer. Individual policies can continue under their own terms. A mandate can expire at the scheduled date, be terminated on notice, be suspended during investigation or end upon a defined event, subject to applicable law. Each state should have a defined effect on new binding, renewal, endorsements, documentation, claims and record access.
The historical LMA 3113 wording illustrates the distinction through sections 36 and 37. It addresses termination notices and the continuing performance of servicing obligations subject to its conditions and regulatory prohibitions. The significance of this structure is that withdrawal of future underwriting power and management of past obligations require separate contractual provisions.
Suppose a binder expires on December 31, while policies bound during December have twelve-month terms. On January 1, its right to bind new business may have ended, yet most December policies remain within their stated periods. A claim occurring in the subsequent year might be covered according to the underlying policy. A run-off arrangement must identify which party receives and investigates that claim, who can settle it and how the carrier gains access to the corresponding policy records.
A termination instruction should therefore address more than disabling an underwriting user account. It should identify permitted residual operations, outstanding referrals, open quotations, renewal notices, unfinished endorsements, premium adjustments, cash balances, claim files and ongoing regulatory reporting. The exact treatment depends on the contract and law.
Continuous delegation and annual capacity
Lloyd's continuous contract wording checklist identifies mandatory and advisory amendments to LMA3113A. The mandatory provisions address an agreement continuing until termination, annual premium and aggregate exposure limits measured from January 1, post-termination authority and transfer between consecutive underwriting years of account. The checklist separates continuing delegation from the annual capacity-accounting process; its model transfer clause distinguishes incoming and outgoing underwriters. These are specific Lloyd's arrangements rather than universal features of continuous contracts.
Continuous contractual duration therefore does not imply unlimited aggregate exposure or perpetual liability for a single capital provider. Consider a hypothetical continuous binder with an annual premium ceiling of $12 million beginning July 1. If the contractual convention is half-year pro-rating, initial authority is $6 million. A $5.8 million ledger leaves $0.2 million headroom. At January 1 the next year's premium ceiling is a separate authorization, while policies already issued and related claims remain assigned to their applicable underwriting periods. The $6 million cap and calendar convention are illustrative assumptions; the governing instrument determines the actual method.
Runoff control should record the original period, the party responsible for the existing commitment, any subsequent authorized amendment and the applicable limit ledger. A system which resets every historical exposure record at the annual transfer can lose visibility of outstanding liability. Conversely, allocating every historic policy to incoming underwriting capacity can misstate the obligations assigned to the outgoing year. The continuing binder is therefore a contract-administration relationship over multiple periods, with period-specific accounting and servicing responsibilities.
Triggers and orderly transition
A carrier may wish to intervene because of loss of licensing, insolvency, fraud, failure to report, deterioration in underwriting performance, exhaustion of capacity or breach of material conditions. Different situations call for different responses. A temporary data failure could require a stop on new binding until data are reconciled; suspected fraud or withdrawal of legal permission may require urgent protection of records and customer servicing.
The contract should define notice channels, the party empowered to issue instructions, effective times and processes required by mandatory law. Remediation periods may be appropriate for some failures; serious threats to customer protection or binding capacity may require immediate restriction where lawful. A notice that is operationally ineffective because it reaches an unmonitored address creates a further risk distinct from its legal validity.
A run-off plan should be developed before termination and identify responsibility for residual operations. It should cover transfer of necessary data, unearned premium reconciliations, claims authority changes, ongoing reporting and assistance with disputes. The plan should be tested under conditions where the intermediary's normal staff or infrastructure may be unavailable.
Claims authority and the long tail
Claims authority needs its own schedule, specifying functions, limits, referral circumstances and responsibilities for reserving, payment, litigation and recoveries. A limit of $100,000 for settlement can be formulated on a per-claim basis, per claimant or total related claims, with or without expenses. A contract that does not define its basis makes apparently clear authority ambiguous.
Assume two connected liability claims each have an estimated value of $70,000. If the relevant threshold is per claim, each may qualify for delegated settlement, subject to other restrictions. If the threshold applies to related claims collectively, $140,000 exceeds the limit. A reserve increase may require referral before payment is considered. The design of aggregation terms matters because the decision to treat related claims separately can change the result.
Long-tail claims may continue long after a binder termination. Ownership or control of data, access rights, preservation of coverage documents and funding for servicing become especially consequential. A carrier replacing the delegate must maintain a record of the former authority and the policies issued under it. Transfer should also preserve correspondence and key dates to avoid loss of defenses, recoveries or customer entitlements.
The intended objective of runoff is continuity of lawful and accurate service. The insurer should maintain capability to assume or transfer essential operations, with due regard to data protection and regulatory authorization. An intermediary that loses its permission to act may be unable to perform residual tasks; the alternative servicing arrangements must accommodate that possibility.
Comparative synthesis, limitations and reproducibility
Comparative institutional interpretation
The institutional comparison uses functional equivalence rather than terminological equivalence. In the US, the NAIC's (2002) Managing General Agents Act supplies a model for states, including written underwriting guidelines, records and claims reporting. The model does not become binding law solely through publication. The North Carolina General Statutes provide one concrete state implementation example: section 58-34-2(d) addresses written MGA contractual responsibilities, underwriting guidelines and claims reporting. Its statutory MGA definition and conditions govern its own application; the label MGA in a commercial description is not sufficient to establish statutory coverage.
In the EU prudential context, Article 274 on outsourcing of Delegated Regulation 2015/35 specifies governance and contractual safeguards for outsourcing within its scope, including the treatment of sub-outsourcing and supervisory information access. It should be read in its applicable legislative framework, including Solvency II Article 49 and relevant subsequent changes. The present inquiry uses the EIOPA text of Article 274, without asserting a complete current consolidation or advising on classification of a particular transaction.
The IAIS Insurance Core Principles address retained governance responsibility for outsourced activities and consideration of provider concentrations. These international supervisory materials offer a comparison at the level of objectives and supervisory expectations; domestic implementation and enforcement remain matters of the relevant legal system.
| Framework | Legal or institutional status | Contribution to agreement analysis |
|---|---|---|
| Lloyd's requirements | Market requirements under the relevant byelaw | Contract particulars and delegated decision boundaries |
| Lloyd's guidance | Market guidance and identified expectations | Oversight and implementation |
| NAIC Model 225 | Model legislation | Written MGA mandate and safeguards |
| North Carolina § 58-34-2 | Enacted state statute within scope | Specific written contract duties |
| EU Article 274 | Delegated regulation within applicable scope | Outsourcing governance and contractual access |
| IAIS ICP 8 | International supervisory framework | Retained responsibility and material outsourcing |
| FCA TR15/7 | Historical supervisory thematic evidence | Observed problems in oversight and customer outcomes |
The similarities concern the need to know what has been delegated, how performance is observed and what happens when it deteriorates. Differences concern statutory definitions, the scope of regulated activities, mandatory provisions, institutional architecture and legal remedies. A comparative synthesis becomes unreliable when a market-specific requirement is translated directly into a universal legal duty.
Competing explanations and scope conditions
One explanation of restrictive contracts emphasizes incentives: restrictions reduce opportunities for the delegate to obtain private benefits at the expense of the carrier. A second emphasizes uncertainty: detailed limits are necessary because the carrier cannot immediately observe the full financial consequences of each transaction. A third emphasizes administrative capability: restrictions may reflect the ability of a particular institution to collect data and process exceptions. These explanations are compatible in many situations but suggest different empirical predictions.
If incentive misalignment dominates, remuneration and exception patterns should be closely associated with the distribution of unauthorized or unprofitable transactions after controlling for mix and market conditions. If delayed observation dominates, reductions in reporting latency should improve timely detection of approaching thresholds, even without changing commission. If capability constraints dominate, staff and system improvements may permit a wider mandate without increasing deviations. These are testable propositions for future empirical research, not observed causal results.
A restrictive mandate can also create undesirable effects. If a locally knowledgeable delegate cannot respond in time to changing conditions within a reasonable pricing range, customers may experience delay or reduced access. If hard thresholds are set using incomplete exposure measures, compliant transactions can still add to an undesirable concentration. The relative benefits of delegated discretion depend on the quality of classification, observability of commitments and the capacity to intervene.
The illustrative numerical tests establish logical relationships: multiple constraints can disagree, delayed information can overstate headroom, concurrent commitments can exceed a shared limit and early profit estimates are sensitive to loss development. They do not show the size or frequency of those effects in market practice. Numerical authority limits therefore require calibration against the relevant portfolio, legal framework and operational system.
Reproducible evaluation design
An empirical investigation could define a binder-year panel, with stable binder and carrier identifiers, class, territory, policy and transaction records, authority versions, time-stamped referral decisions, premium movements, exposure measures, incurred and ultimate loss estimates and service outcomes. Contracts should be classified at the level of each function rather than by name alone. Changes in limits, product mix and reporting arrangements should be recorded with effective times.
One descriptive measure would be the proportion of binding transactions found outside their contemporaneous authority, calculated separately for confirmed breach, ambiguous documentation and missing evidence. Another measure would be reporting latency, measured from the actual binding event to successful receipt and reconciliation. Accumulation exposure would require a prespecified metric, location aggregation and treatment of shared limits and reinsurance.
For a causal study, changes in the design of authority schedules cannot be treated as randomly assigned. More restrictive mandates may be chosen precisely for riskier portfolios. Comparisons would need to consider selection, carrier characteristics, market cycles, business mix, loss development and measurement error. A before-and-after comparison without a credible counterfactual would support descriptive monitoring, not a causal claim. Appropriate anonymization, secure handling and permissions would be necessary for any actual policy or claims data.
The arithmetic scenarios in this monograph are individually reproducible from the given inputs; the primary source selections are documented in the evidence register. Contract version dates and access restrictions matter for replication. The historical market wording and FCA review are explicitly dated or identified as historical. Selected original-text sections of Jensen and Meckling (1976) and Regan (1997) were inspected; the Regan and Tennyson (1996) claim remains confined to the accessible abstract. Later Lloyd's checklist and LMA bulletins supplement the historical base form. The full LMA3113A document and underlying 2025 clause files were not retrieved; no unverified text is ascribed to them. The current consolidated law of all jurisdictions, proprietary binder experience and independent external peer review are outside the observed evidence population. The study therefore establishes a comparative and analytical treatment, not a comprehensive jurisdictional compliance manual or an empirically validated optimization rule.
Conclusion
The binding authority agreement establishes a bounded transfer of underwriting decision rights. The carrier retains a substantive interest in every commitment made through that authority, and the effectiveness of delegation depends on the relationship between contract terms, operational capability and portfolio evidence.
The first research question concerns the definition of permitted risk. An intelligible mandate identifies authorized parties, classes and coverage forms, territory dimensions, pricing discretion, per-risk ceilings, premium basis, policy terms, endorsement powers and referral rules. Its meaning must remain traceable through amendments and transaction records. A binder does not by itself determine the outcome of a third-party coverage dispute or confer missing regulatory permissions.
The second concerns the connection between subdelegation, aggregate limits and reporting. Individual permissions must be reconciled with total available capacity. A premium limit is not an event concentration limit; a gross policy limit is not the carrier's signed or net retained exposure; claims authority is its own decision boundary. The worked cases demonstrate that separate transactions can be compliant individually while exceeding a shared constraint collectively. A delegate's reporting duty and the carrier's reconciliation and intervention capability are therefore integral to the effective limit.
The third question concerns audit and dispute resolution. Reconstructability requires authoritative agreement versions, decision evidence, usable records and access rights that survive disruption where law permits. Termination must distinguish new commitments from the residual obligations of policies already issued, including claims and customer servicing. Audit findings should be resolved against actual transactions and their consequences rather than closed through documentation alone.
The original contribution is an integrated interpretation of the contractual, financial and operational dimensions of binding authority. It explains why appropriate measurement, timing and allocation of responsibility are necessary for risk limits to perform their intended function. It derives no new theory and claims no empirical estimate of the superiority of a specific contract design.
References
Lloyd's. (n.d.). Contracts of delegated authority. Requirements made pursuant to the Intermediaries Byelaw, paragraphs 10, 12A and 13. https://www.lloyds.com/market-resources/requirements-and-standards/lloyds-requirements-of-the-council/operating-at-lloyds/intermediaries/contracts
Lloyd's. (n.d.). Wording expectations for contracts of delegation. https://www.lloyds.com/market-resources/delegated-authorities/market-knowledge/Wording-expectations
Lloyd's. (n.d.). Delegated underwriting. https://www.lloyds.com/market-resources/delegated-authorities/market-knowledge/delegated-underwriting-guidance/
Lloyd's. (n.d.). Oversight and monitoring of contracts of delegation of underwriting authority. https://www.lloyds.com/market-resources/delegated-authorities/market-knowledge/oversight-and-monitoring-of-contracts
Lloyd's. (n.d.). Reporting standards. https://www.lloyds.com/market-resources/delegated-authorities/market-knowledge/reporting-standards
Lloyd's. (n.d.). Coordinated audits. https://www.lloyds.com/market-resources/delegated-authorities/market-knowledge/audit
Lloyd's. (n.d.). Master coverholder arrangements and master broker arrangements. https://www.lloyds.com/market-resources/delegated-authorities/market-knowledge/masters-and-subs
Lloyd's Europe. (n.d.). Binding authority wordings. https://lloydseurope.com/market-resources/delegated-authorities/compliance-operations/binding-authority-wordings
Lloyd's Market Association. (n.d.). LMA 3113: Model Binding Authority Agreement (worldwide excluding USA and Canada-domiciled coverholders). Historical model wording. https://assets.lloyds.com/media/a7f8df06-b7d1-42b4-a385-327333b550f4/pdf-lma-3113-model-binding-authority-agreement-excluding-us-canada-domiciled-coverholders.pdf
National Association of Insurance Commissioners. (2002). Managing General Agents Act (Model 225), October 2002 compilation, section 4. https://content.naic.org/sites/default/files/model-law-225.pdf
North Carolina General Assembly. (n.d.). North Carolina General Statutes, Chapter 58, Article 34, section 58-34-2 (Managing general agents). https://www.ncleg.gov/EnactedLegislation/Statutes/HTML/ByArticle/Chapter_58/Article_34.html
European Commission. (2014). Commission Delegated Regulation (EU) 2015/35 of 10 October 2014, Article 274 (Outsourcing). EIOPA Single Rulebook. https://www.eiopa.europa.eu/rulebook/solvency-ii-single-rulebook/article-5937_en
International Association of Insurance Supervisors. (2024). Insurance Core Principles and Common Framework for the Supervision of Internationally Active Insurance Groups. ICP 8, outsourcing (online text). https://www.iais.org/icp-online-tool/
Financial Conduct Authority. (2015). Delegated authority: Outsourcing in the general insurance market. Thematic Review TR15/7. June 2015; marked historical August 2025. https://www.fca.org.uk/publication/thematic-reviews/tr15-07.pdf
Financial Conduct Authority. (2015). FCA says firms not treating delegated authority arrangements as outsourcing in the general insurance market. June 2, 2015. https://www.fca.org.uk/news/press-releases/fca-says-firms-not-treating-delegated-authority-arrangements-outsourcing-general
Jensen, M. C., & Meckling, W. H. (1976). Theory of the firm: Managerial behavior, agency costs and ownership structure. Journal of Financial Economics, 3(4), 305–360. DOI: 10.1016/0304-405X(76)90026-X. Original article, accessible author-independent PDF copy. https://www.liuyanecon.com/wp-content/uploads/JensenMeckling-1976.pdf
Regan, L., & Tennyson, S. (1996). Agent discretion and the choice of insurance marketing system. Journal of Law and Economics, 39(2), 637–666. DOI: 10.1086/467361. IDEAS abstract record. https://ideas.repec.org/a/ucp/jlawec/v39y1996i2p637-66.html
Regan, L. (1997). Vertical integration in the property-liability insurance industry: A transaction cost approach. Journal of Risk and Insurance, 64(1), 41–62. Original article, accessible academic PDF copy. https://firn.garven.com/readings/distribution/Regan_1997JRI.pdf
Lloyd's Market Association. (n.d.). LMA model agreements / GDPR and Criminal Finance Act 2017 amendments. https://lmalloyds.com/specialist-areas/legal/legal-and-compliance-useful-documents/lma-model-agreements-gdpr-and-criminal-finance-act-2017-amendments/
Lloyd's Market Association. (2025, April 1). Mandate of Lloyd's Lead Security – Delegated Authority Agreement Endorsements. Bulletin LMA25-008-DG. https://lmalloyds.com/mandate-of-lloyds-lead-security-delegated-authority-agreement-endorsements/
Lloyd's Market Association. (2025, August 5). Lloyd's Market Association publishes new model wording for binding authority profit commission clauses. https://lmalloyds.com/lloyds-market-association-publishes-new-model-wording-for-binding-authority-profit-commission-clauses/
Lloyd's. (n.d.). Continuous contracts wordings: Checklist for contract changes to LMA3113A for continuous contracts. Updates February 22, 2023. https://www.lloyds.com/market-resources/delegated-authorities/market-knowledge/continuous-contracts/wording