RESEARCH DOSSIER
Insurance Sector Vertical Integration and Distribution Foreclosure
Abstract. Insurance distribution connects risk-bearing capacity to households and enterprises. Vertical integration combines control over successive functions, while exclusivity and preferential contracts can confer substantial control over distribution without common ownership. This monograph examines how bank-insurer integration, insurer-intermediary relationships, and digital distribution restrictions alter entry conditions, incentives, consumer choice, and competition. A comparative documentary method integrates industrial organization, insurance distribution scholarship, the European credit protection insurance review, primary legal materials from the European Union, the United Kingdom and the United States, and selected Asian distribution evidence. The analysis distinguishes the organization of insurance production from the ability of independent insurers to reach customers, and separates adverse market conduct from an established anticompetitive effect. Explicit hypothetical numerical exercises demonstrate how channel capacity, diversion, distribution margins and risk-adjusted underwriting costs interact. The principal conclusion is that foreclosure depends on the economic significance of inaccessible customer opportunities, credible alternatives and the combined firm's incentives, rather than a formal classification of the organizational arrangement. Integration may improve distribution efficiency and product access; effective competition requires inquiry into the quality of alternatives and actual market response. The study provides an evidence-based synthesis and a replicable research design; it does not estimate causal effects of particular transactions or introduce a new theory.
Keywords: Vertical integration; insurance distribution; bancassurance; customer foreclosure; input foreclosure; exclusive dealing; channel access; credit protection insurance; conflicts of interest; insurance competition; market conduct; industrial organization.
Introduction
Insurance is supplied through an institutional sequence. A risk carrier underwrites and capitalizes the contractual promise, distribution identifies and matches potential insureds, administration maintains the contract, and claims processes determine whether and how a realized contingency gives rise to payment. These functions can be housed within one group, divided among specialist firms, or coordinated through long-term commercial agreements. Their allocation affects both the resource cost of insurance and the alternatives available to a prospective policyholder.
Distribution is particularly important because the insurance contract is a contingent service. A purchaser frequently assesses price and observable contractual features before experiencing the quality of claims performance. An intermediary can reduce the cost of searching, interpreting exclusions, submitting underwriting information and negotiating terms. The same intermediary can determine which carriers reach a customer, which alternatives receive attention, and whether an independent producer can acquire sufficient business to maintain a competitive operation.
Vertical integration in this study means common control of activities performed at successive economic stages of providing insurance. Contractual integration means intensive coordination through agreements governing those stages, potentially including exclusive appointment, minimum production, renewal rights or specialized interfaces. A bank and insurer can also supply complementary products without performing successive stages in relation to a particular transaction; the relevant relationship must therefore be mapped at the level of the activity and the product.
Distribution foreclosure is an impairment of commercially significant access to distribution opportunities or risk-bearing inputs, arising from arrangements that alter the ability or incentive of actual or potential rivals to compete. Complete exclusion is one limiting case. A rival that remains nominally present may face higher acquisition costs, lower quote visibility, reduced conversion or slower access to data. The competition question concerns the consequences of those changes for the competitive process and end customers.
The approved scientific scope consists of three dimensions: bank-insurance integration, vertical conflicts and channel restrictions. The research questions are how these arrangements are documented, under which conditions they change competitive incentives, and how effects can be distinguished from selection, product differences and genuine economies of organization. The study combines institutional evidence and economic reasoning. It uses dated jurisdiction-specific materials and selected observations, not a worldwide census of insurer contracts. The comparison includes both supportive and adverse evidence for the proposed mechanisms.
Conceptual and Historical Foundations
The institutional sequence of insurance production
A carrier's underwriting contract creates a financial obligation under prescribed conditions. Capital, reinsurance and technical pricing enable the carrier to accept that obligation. Customer acquisition, advice, explanation, identity verification, risk information collection and enrollment make it possible for the contract to be formed. An agent may act on behalf of a carrier, a broker may arrange coverage across carriers, and a bank may provide distribution in the course of supplying a separate financial product. The contractual and regulatory capacities of these institutions vary by jurisdiction. A platform may compare quotes, generate leads or complete distribution, depending on its actual function.
The economic stage matters more than the institutional label. A bank selling its own group's insurance is downstream of the underwriting function with respect to distribution. As a lender, it also supplies a complementary product. A retailer offering embedded travel insurance controls the buying occasion but commonly does not bear the insured risk. A managing general agent may exercise delegated underwriting authority while a separate insurer remains contractually responsible for claims. A broker that acquires a carrier may combine access to customers with the ability to supply underwriting capacity. A reinsurer acquiring a distributor creates yet another relation, depending on contractual authority and actual control.
These differences determine the direction of a possible exclusionary mechanism. When an integrated carrier restricts independent distributors' access to attractive underwriting capacity, the disputed input is insurance supply. When an integrated distributor directs transactions away from independent carriers, the disputed customer base consists of the policy opportunities available to those carriers. Distribution as an input may itself be purchased by insurers; the label input foreclosure or customer foreclosure must be linked to an explicitly specified market and direction of supply. Confusion over terminology can obscure the same underlying constraint.
Organizational economics and channel diversity
Historic insurance distribution research provides an empirical reason to avoid a presumption that the lowest measured acquisition cost determines the optimal organizational form. Regan and Tennyson's (1996) study of agent discretion relates organizational choice to the role of intermediary information in matching risks to coverage. Berger, Cummins and Weiss (1997) compare direct writers with independent agencies in their distribution system study; their reported efficiency findings support a product-quality explanation for the persistence of independent channels. These findings are based here on the published abstracts, which establish the direction of the reported results rather than the detailed estimates.
A carrier that delegates distribution must manage the alignment of information, effort and compensation. A commissioned agent may devote insufficient attention to a carrier whose product pays less, while a carrier may change the terms applicable to investments made by its independent agent. Integration can resolve some contracting problems by consolidating control of customer data, marketing expenditures, technology and product design. It may also replace market-based discipline with internal administrative discretion. Whether the substitution is beneficial depends on the quality of the relevant decisions and the ability of customers or trading partners to change providers.
Specialized investments clarify this trade-off. Suppose a bank invests in training thousands of staff members to explain a complex insurance product and installs secure data-transfer software. An insurer benefits from those investments; the bank may fear that an insurer could withdraw support before the investment is recovered. A longer agreement can protect the joint investment. However, when the contract reserves the majority of the bank's valuable sales opportunities well beyond the recovery period, the marginal extension of exclusivity requires a separate economic explanation. Investments recoverable only under exclusivity should be separated from investments that a nonexclusive agreement could support through a service fee, minimum volume or defined intellectual-property arrangements.
Product differentiation and efficiency
The costs of selling a simple policy to a known bank customer differ from the costs of placing complex industrial risks. A broad intermediary panel can support market comparison, heterogeneous underwriting appetite and the negotiation of specialized clauses. A captive distribution network can be efficient for standardized coverage and rapid customer service. Comparing average commission rates between those two functions without controlling for product complexity and the value of service would obscure the economic question.
The relevant cost of distribution includes lead acquisition, risk screening, advice, conversion, administration, cancellation, claims assistance and compliance. A narrow comparison of marketing expense may place a channel that selects easy risks alongside a channel serving costly risks. Some activities generate higher initial cost and lower downstream cost by preventing unsuitable contracts and disputed claims. An efficiency comparison therefore needs comparable risk classes, service outputs and observation periods.
The merger of successive economic stages may eliminate a double markup when both stages independently choose positive margins over marginal cost. The combined entity internalizes the impact of a distribution price on underwriting demand, and may lower the retail price. This textbook mechanism requires the relevant margin to be avoidable and demand to respond. A fixed commission paid under a negotiated contract can already coordinate incentives. A reduction in internal transfer charges is an accounting change unless it corresponds to lower real costs or improved final terms. In insurance, the product premium also compensates for expected claims, operating costs and the cost of risk-bearing; reduced distribution expense may not be passed through when competition is weak.
Conditions for foreclosure
The economic assessment follows established industrial organization rather than treating vertical control as a sufficient criterion. The European Commission's 2008 non-horizontal merger guidelines distinguish limitations on upstream inputs and downstream customer access and analyze their ability, incentive and likely effect. The 2021 UK merger guidelines likewise examine the relative margins and substitution relevant to a customer-foreclosure scenario. The US agencies' 2023 Guideline 5 explicitly addresses access to commercially important routes to market.
Three connected inquiries follow. First, does the controlled resource matter, and can alternatives absorb diverted demand on comparable terms? Second, does restricting access increase the integrated firm's profit after allowing for foregone intermediary business, costs and customer loss? Third, is the predicted change large and durable enough to weaken competition, impair innovation or raise the quality-adjusted cost of protection? An affirmative answer to one inquiry alone is insufficient for a complete economic assessment.
The 2026 EU merger guidelines review reports draft revisions published in April and a completed consultation in June. The analysis uses the documented 2008 framework as a dated reference; it does not treat the consultation draft as an enacted replacement. Agency methodologies assist the research but statutory legal standards, procedure and evidentiary burdens depend on the applicable jurisdiction.
Primary Institutional Documents and Comparative Research Design
Evidence population and units of observation
The most useful empirical unit is a carrier-channel-product-geography-time combination. A corporate group can own several licensed carriers, distribute through both affiliated and independent outlets, and sell products with sharply different distribution requirements. A bank may represent one carrier for mortgage-related protection and several carriers for other products. Consolidated premium figures will not identify these relationships. A dataset must distinguish the legal risk bearer, ultimate controller, contractual intermediary, customer segment, coverage type, observation date and the market boundary used for calculations.
One transaction can produce several records. A master group-policy contract may cover many insured persons, each of whom may be enrolled through separate certificates. New premiums and renewal premiums represent different moments of competitive exposure. Written premium measures contracted business over a period; earned premium allocates it to the period in which risk is carried. An insured person's active certificate, a master contract, a new policy and gross written premium are distinct denominators. The unit must remain fixed across comparisons.
Contract documentation establishes exclusivity, duration, territory, minimum volume, financial incentives, exceptions and renewal provisions. Distribution data establish where opportunities occur, how many customers receive quotes, which carrier accepts each risk, and how many ultimately enroll. Claims and complaints data establish aspects of product outcome. Official decisions provide legal characterization and evidentiary status. No single source type establishes the whole causal sequence.
The international supervisory context is framed by the IAIS (2024) ICPs and ComFrame overview, which describes coordination and supervision of internationally active insurance groups. Group-level risk oversight and local conduct responsibilities need to be distinguished from market competition analysis. Implementation through national rules must be checked independently.
A documentary comparison across institutions
The European Insurance and Occupational Pensions Authority's (EIOPA, 2022) credit protection insurance review covers mortgage, consumer-credit and credit-card protection distributed by banks. Its dataset comprises 174 insurers and 145 banks. Participating insurers reported 2020 gross written premiums of approximately EUR 8.3 billion, EUR 5.5 billion and EUR 615 million respectively for those three product categories. The population is the participating undertakings, and the period and product definitions belong to the review. These figures are not the total value of bank-distributed insurance worldwide.
| Evidence category | Unit | Documented use | Inference boundary |
|---|---|---|---|
| EU credit protection review | Participating insurers and banks; product observations | Organization, distribution incentives, product outcomes | Observational sample; no causal merger estimate |
| UK market investigation and appeal | Defined affected markets, agency findings, tribunal review | Restriction mechanisms and evidentiary burden | Each proceeding has its own product scope and legal test |
| EU insurance distribution legislation | Applicable distributors and products | Conduct, customer needs and cross-selling duties | Domestic implementation and exceptions |
| US federal merger guidance and credit disclosure | Transaction or covered credit transaction | Access mechanisms and finance-charge treatment | Distinct legal regimes |
| Company filing | Consolidated company and disclosed business segments | Contractual access and channel coexistence | Self-reported reach, not market-wide foreclosure |
| Insurance scholarship | Study-specific populations and models | Organizational choice and efficiency hypotheses | Abstract-level review of selected studies |
The cases are selected for analytical contrast, not statistically representative of all legal systems. Official regulatory summaries are distinguished from judicial rulings and adopted legislation. A press release about a newly opened proceeding is evidence of the proceeding and the authority's described allegation, not of an ultimate violation. Company disclosure establishes a management-reported business arrangement, not independent proof of its competitive effect. The resulting evidentiary hierarchy allows substantial institutional conclusions while restricting quantitative claims to observable aggregates.
Measurement and identification
A reproducible study would first freeze the transaction boundary and the relevant product and geographical market for each hypothesis. Define whether a carrier seeks access to a particular customer segment, a specified distribution technology or a class of risk capacity. Construct an opportunity panel using denominator-consistent exposure quantities and identify the ownership and contract status on each observation date. Distinguish offers made, offers shown, accepted quotes and bound policies. Record rejected applications and reasons. Without rejected and unshown alternatives, estimates of consumer choice can be selected on realized sales.
Data should preserve versioned product terms, exclusions, geographic coverage, net and gross premiums, commission components, cancellation rules, policy duration, customer risk attributes, claims development and time at risk. Nominal currency values require price-level or exchange-rate adjustments appropriate to the question. New business may be compared using annualized premium or number of customers, but investments with large single premiums can dominate premium-weighted measures. Group membership needs to be coded at the applicable time, including divestitures and changed joint ventures.
The selection problem is fundamental. Banks and insurers choose a channel because of expected demand, product type, distribution resources and anticipated regulatory conditions. A high rate of policy conversion following a partnership may reflect a pre-existing growth forecast. Comparing countries that permit different structures risks conflating income, public protection, consumer preferences, tax treatment, credit cycles, licensing and product composition. A causal estimate requires variation that can distinguish the channel arrangement from those simultaneous influences.
A candidate longitudinal design follows product cohorts through a contract transition and compares them with suitable untreated observations. The analyst would evaluate pre-transition trajectories, anticipate contract announcements, examine concurrent regulatory changes, test alternative geographic boundaries and use uncertainty intervals that reflect the sampling level. Selection into treatment and spillovers to control channels may remain substantial. The present work specifies this design and checks illustrative arithmetic; it makes no assertion that a microdata regression, market test or event study has been executed.
Bank Insurance Integration
Organizational forms and control rights
Bancassurance is the distribution of insurance through banking relationships and infrastructure. It encompasses referral agreements, open product panels, preferential appointments, exclusive arrangements, joint ventures and common-control structures. Each form allocates control of the customer relationship, insurance risk, data and contract design differently. A bank with a minority investment and strong contractual veto rights may have more influence over an insurer's distribution than a nominal equity percentage suggests. A long-term alliance may produce functionally similar customer-access restrictions to common ownership while preserving distinct prudential balance sheets.
Integration should be recorded along separate dimensions: equity ownership and voting rights; governance and appointment rights; exclusivity in each sales channel; control of customer access and data; insurer capacity allocation; distribution fees and contingent bonuses; investment and termination obligations. Labeling all arrangements as bancassurance is appropriate for broad descriptive purposes. A competition analysis also requires these narrower contractual facts.
Banks can acquire customers through existing account and lending relationships. They can identify a protection gap at a financially important moment and reduce duplication in identity verification and payment administration. A carrier may gain exposure to risks that would be costly to reach individually. The integrated network may expand access to coverage, including outside major cities, when physical branches or mobile services are established before an insurance network could feasibly be built.
The value of access is not the same for every policy. A borrower negotiating time-sensitive financing may be less willing to interrupt the transaction to obtain alternative insurance quotes than a purchaser planning a long-term savings product. A product purchased during loan approval can combine administrative convenience with a heightened dependency on bank advice. Understanding the timing of choice is therefore necessary even when the coverage itself is optional in the contract documents.
Evidence from bank-distributed credit protection
EIOPA's (2022) review of bank relationships reports close bank relationships, encompassing common-group structures, alliances and joint ventures, for 63% of participating insurers. It also records that 115 of 174 insurers used CPI group policies with the bank as policyholder and the customer as insured person. The same review identifies variation in national markets and product design. These observations establish the economic salience of close relationships in the studied CPI population, but neither percentage measures the share of all EU insurance premiums or the proportion of potential rival sales excluded.
For group contracts, the bank may negotiate master terms and the consumer becomes covered through affiliation or enrollment. A consumer's practical ability to compare or replace coverage may therefore depend on whether an equivalent individual policy is permitted, how equivalence is assessed and whether leaving the group contract affects the loan terms or requires further administrative steps. The insurer's obligations and the bank's roles remain separately identifiable even when contractual form makes their economic relationship close.
A single premium financed through a loan has a different economic profile from a periodic premium. Its cost to a borrower may include interest on the financed premium; early loan repayment or refinancing introduces questions about termination and refund terms. A regulatory review of cost or claims ratios that ignores premium timing, loan duration or changing insured balances risks distorting comparison.
Consider an explicitly hypothetical single-premium policy costing 600 currency units and financed for 36 months at a nominal annual interest rate of 9%, with a monthly installment loan rate of 0.75%. The incremental monthly installment is calculated by multiplying 600 by 0.0075 and dividing by (1 - 1.0075 raised to the power -36). This is approximately 19.08 per month, or about 687 in total across 36 months before other charges. A nominal 600 price and the financing-inclusive cash cost therefore answer different questions. The exercise is a standard amortization illustration, not a price observed in the EIOPA sample.
International evidence and channel efficiency
Manulife Financial Corporation's (2026) Annual Information Form reports exclusive Asian bank partnerships providing access to more than 35 million bank customers, including its regional partnership with DBS across Singapore, Hong Kong, mainland China and Indonesia. It also describes agency, broker, financial-adviser and bank-based distribution in the region. The reported customer reach is an opportunity measure, not a count of policies sold or denied to rivals. A company can use exclusive partnerships while competing through several other channels.
The Taiwanese study by Chang, Peng and Fan (2011), comparing bank and traditional channels, reports a higher efficiency score for traditional distribution in its comparable-channel data envelopment analysis. The available abstract supplies a useful counterpoint to the presumption that banks always achieve superior measured efficiency. Its research design, coefficients and robustness cannot be independently reproduced from the abstract alone. The result belongs to its observed setting and measurement framework.
Cross-country efficiency rankings require information on underwriting functions, product complexity, banking-network scale, policy acquisition cost, persistence and claim service. A bank intermediary measured as a distribution business and a carrier measured as a balance-sheet institution have different output denominators. An efficient bank sales operation can distribute an expensive or poorly matched product. Conversely, relatively high distribution expense may support informed choice and better policy fit. The appropriate final performance measure is protection delivered relative to risk, cost and service quality.
Competition at contract award and renewal
An exclusive agreement can be awarded competitively even when customers cannot choose among insurers during its term. Competition for the distribution contract can induce carriers to invest in technology, improve contractual coverage or offer attractive compensation. Whether those gains reach policyholders depends on the bank's objective and the tender criteria. If the distributor primarily maximizes upfront consideration or commission, vigorous competition for the agreement could increase the cost of obtaining distribution without improving insurance outcomes.
Contract duration matters through the flow of opportunities. Short, staggered contracts may permit repeated entry attempts. Long, synchronized arrangements may reserve the most valuable bank networks for extended periods. A prospective entrant that needs several agreements to reach efficient scale may face a practical multi-year exclusion even if an occasional smaller contract is available. Assessment should use the expiry calendar and the timing of credible alternatives rather than nominal contract counts.
Renewal clauses, matching rights, access to customer histories and operational portability determine whether a new insurer can replace an incumbent effectively. Formal termination rights are of little competitive significance if the bank cannot migrate administration or transfer consented information on commercially reasonable terms. Equally, migration costs may represent genuine service continuity requirements. Distinguishing necessary integration from avoidable barriers requires documentation of actual implementation costs and alternatives.
Joint control through distribution: Fortis and BCP
The European Commission's 2005 Fortis/BCP decision provides a direct institutional illustration of why an ownership database alone cannot measure control. The new insurance joint venture combined four former BCP insurers, with Fortis holding 51 percent of capital and BCP 49 percent. Decisive commercial rights nevertheless supported joint control. The venture depended on BCP for approximately 94 percent of its turnover, and distributing through another bank required BCP's consent. The Commission cleared the transaction after considering the particular market circumstances. The decision establishes control and a prospective competitive assessment, not subsequently observed consumer outcomes.
For the Portuguese market in 2003, the decision reports that the bank channel supplied 81 percent of life and 12 percent of non-life gross written premiums. BCP accounted for 20 percent of the life bank channel and 29 percent of the non-life bank channel. Multiplication of the nested shares yields an approximate 16.2 percent of total life premium and 3.48 percent of total non-life premium distributed through BCP, assuming the reported denominators and rounding are consistent. Neither derived figure measures the share of policies actually excluded from rivals. The difference between them reflects both channel prevalence and BCP's channel position.
This comparison distinguishes three quantities: the venture's dependence on its parent distributor, the parent's share within that distribution channel, and the distribution channel's share of the wider product market. Dependence can be extremely high for an individual undertaking even when the contested route constitutes a minority of all market opportunities. Conversely, a moderate national share can mask dependence within a narrow customer group or lending product. Any inference about exclusion requires a denominator corresponding to the economically affected market and a counterfactual describing the access that would otherwise have existed.
Empirical evidence on channel coexistence
Rubio-Misas (2022), in Bancassurance and channel coexistence, applies data envelopment analysis and a two-stage Heckman panel specification to Spanish life insurers. The banking channel is associated with improved cost efficiency, while the estimated distribution-system variables do not significantly explain profit inefficiency. The paper's robustness notes also report an insignificant interaction between bank-subsidiary ownership and use of the banking channel. These results are consistent with the possibility that higher-service channels earn revenues compensating for higher costs.
The interpretation is limited by the nature of the measures. A frontier efficiency score compares an insurer with a constructed benchmark conditional on its measured inputs, outputs and assumptions. It does not directly observe whether a particular policyholder understood exclusions, received appropriate advice, or could obtain a rival quote. The selection correction responds to nonrandom channel choice through a specified statistical structure; it does not transform observational data into randomized assignment. A channel-related improvement in cost efficiency can therefore coexist with a competitive concern over particular restrictions. The two inquiries assess different outcomes.
Taken together with the Taiwanese and US evidence discussed above, the Spanish findings motivate a research distinction among resource efficiency, remuneration, customer service, and accessibility. A claim that an integrated network produces insurance more cheaply requires a comparable resource-cost measure. A claim that it improves welfare requires evidence about price, coverage, quality and availability. A claim that it forecloses competitors requires evidence about their lost access and the consequences of that loss. These questions can be examined jointly without treating evidence for one as an adequate substitute for the others.
Vertical Conflicts
The economics of remuneration
A distributor may be paid a fixed fee, percentage commission, per-policy allowance, volume-based bonus, profit share or a combination. The carrier will evaluate the distributor's contribution to underwriting value after acquisition and servicing costs. The distributor may rank products differently, depending on immediate income and sales effort. The customer values effective coverage and the full cost over the contract's life. These three objective functions can diverge even when the insurer and intermediary are independent.
Common ownership removes some conflicts between the carrier and distributor because profits can be consolidated. The conflict with the policyholder may become stronger if the consolidated group gains more from directing customers to affiliated products. Internal accounting transfers disappear at group consolidation, but staff incentives, product allocation and the opportunity cost of a customer relationship continue to affect behavior.
A simple illustrative calculation clarifies why commissions alone are insufficient. Suppose a premium of 100 funds expected claims of 45, non-distribution operating costs of 15, a distribution commission of 25 and a residual 15 before required capital return and other unallocated costs. If affiliation reduces the real acquisition cost to 12 instead of 25, total resource saving may be 13. It becomes a consumer benefit if competition, contractual arrangements or governance induce lower premiums or better coverage. If the price and coverage remain constant, group surplus can rise while the customer's contract is unchanged. The numerical values are assumed solely to isolate accounting and incentive relationships.
The EIOPA (2022) review of claims ratios also reports average claims ratios for 2018–2020 of roughly 26%, 18% and 8% of gross written premium for mortgage, consumer-credit and credit-card CPI respectively. These ratios provide a supervisory signal about the studied product categories. They do not decompose expected risk cost, acquisition cost, capital, administrative expenses and final profit. Low claims ratios may arise from low covered risk, immature cohorts, exclusions, claim friction, pricing or combinations of these factors. The proper inference is that further outcome and cost investigation is warranted, not that every contract has a corresponding excessive margin.
Conduct duties and the competition boundary
The European Union's Insurance Distribution Directive (2016, consolidated 2024) supplies a regulatory basis for customer interests, demands and needs, cross-selling arrangements and product oversight. Its additional investment-product provisions address specified conflict management obligations. Its exceptions and local implementing measures matter to particular products and sellers. These requirements structure conduct independently of whether a seller has market power.
In the United States, the CFPB's Regulation Z finance-charge rules specify conditions for excluding voluntary credit-insurance charges from the finance charge; actual optionality is a factual matter, and the property-insurance exclusion includes choice and disclosure conditions. The instrument addresses credit-cost treatment. Its application cannot be treated as a universal prohibition on exclusive insurance distribution or as a determination of competition liability.
An unfair sales practice may arise even within a fragmented market when an individual borrower is misled or pressured. A competition restriction can operate without misleading disclosures, for example through a contract that reserves sufficient distribution capacity to undermine rival entry. An exclusionary arrangement may combine both phenomena. A robust empirical record therefore maintains separate outcome columns for actual consent, product suitability, exclusion of suppliers, market concentration and competitive effect.
The distinction has implications for remedies. Disclosure can improve understanding without opening a reserved channel; opening a channel may expand the supplier set without correcting inappropriate coverage. Product governance can revise terms but cannot by itself secure timely access for a competitor. A competition remedy must address the causal bottleneck, while conduct intervention addresses the customer-facing mechanism. Coordinated oversight is useful because the incidents may share data even when the legal tests differ.
Monitoring design and governance
A vertically organized business can make conflicts visible by recording the full set of eligible carriers and reasons for absence from the presented offer, including underwriting appetite, system constraints and contractual exclusivity. Distribution scorecards can compare risk-adjusted price, cancellation, claim outcomes, service quality and complaints alongside production. A governance reviewer would distinguish a legitimate decision to decline an unsuitable product from a decision to exclude an eligible carrier because of affiliation.
One must also monitor human and digital sales systems. A bank's internal staff targets may reward attached products; a platform may apply ranking rules that prioritize a sponsor; an automated user journey may default to the affiliated insurer. These mechanisms require event-level evidence: what options existed, which were shown, what the customer selected, whether ranking varied by seller ownership, and which rules generated each display. Privacy-protective data handling requires lawful collection and minimization of sensitive applicant attributes.
At group level, internal transfers should be analyzed together with external cash flows. For example, the carrier may pay a commission to a group bank while the bank incurs real operating expenses, and a holding company may recognize capital and liquidity impacts across subsidiaries. Competition analysis requires incremental group contribution and alternatives; prudential analysis additionally requires separate legal-entity solvency and contractual accountability. A consolidated positive margin does not establish that each entity's obligations are sustainable.
Channel Restrictions
Access, concentration and effective foreclosure
Exclusive supply and purchase clauses, conditional rebates, parity clauses, tied offers, preferential ranking, technological compatibility rules and contractual rights over data may restrict access to distribution. A restriction can be evaluated by distinguishing its coverage, strength, duration and enforceability. A clause covering a narrow product segment for a short interval is economically different from a network-wide obligation combining long terms, costly exit and access to a strategically important customer segment.
A first descriptive measure is the share of defined customer opportunities covered by binding restrictions. Suppose a relevant annual opportunity population comprises 100,000 eligible purchase occasions. Distributor A handles 40,000 opportunities, of which 75% are covered by exclusivity; distributor B handles 30,000, of which 50% are covered; the remaining 30,000 are open. Assuming mutually exclusive channels and equally measured occasions, the restricted count is 40,000 times 0.75 plus 30,000 times 0.50, or 45,000. The descriptive restricted share is 45%. It is not a loss estimate or a legal foreclosure threshold.
Customers may use several channels, openings may remain in direct sales and channel capacity may adjust. If a quarter of the affected opportunities could migrate to an equally effective alternative at unchanged cost, the upper-bound displaced opportunity count under that simplified assumption would be 33,750, or 33.75% of the original total. If three quarters could migrate, the count would instead be 11,250. These scenarios are arithmetic illustrations. Their significance depends on observed substitution, duplication of customers and the ability of other channels to handle business. A channel with high traffic, low conversion and redundant customer reach may exert less control than one with lower traffic but exclusive access to high-value sales occasions.
Concentration measures support but cannot replace this exercise. A market with insurer shares of 40%, 30%, 20% and 10% has a Herfindahl-Hirschman Index of 3,000 on the percentage-squared scale. If the same four risk carriers face a single bank gateway for a specific customer segment, carrier concentration does not measure the distribution constraint. Separate carrier and gateway measures may be necessary, while actual competitive effects depend on substitution and the relative importance of the affected segment.
Profit incentive and sensitivity
In a stylized incremental analysis, suppose an integrated distributor foregoes net revenue of 8 per previously placed independent policy when it prevents an independent placement. A fraction d of the affected customers switches to the group's carrier. The incremental consolidated contribution per switched policy, after real distribution costs, expected claims, expenses and risk-capital charges, is assumed to be 20. The net change per affected opportunity is 20d minus 8, before any fixed costs, reputational loss, sanctions or changes in equilibrium prices.
| Assumed diversion to affiliate | Incremental gain | Foregone contribution | Net change per affected opportunity |
|---|---|---|---|
| 20% | 4 | 8 | -4 |
| 40% | 8 | 8 | 0 |
| 60% | 12 | 8 | 4 |
| 80% | 16 | 8 | 8 |
Under those assumptions the break-even diversion fraction is 40%. An increase of real incremental costs by 5 reduces the affiliate contribution to 15, increasing the break-even threshold to approximately 53.33%. An alternative distribution fee that raises the foregone contribution to 12 raises the original threshold to 60%. None of these values is an estimate of any named firm's margins.
The same calculation changes when a customer diversion affects the bank's lending relationship. Losing a borrower may sacrifice future interest and cross-selling value, exceeding the gain on affiliated insurance. When that risk is material, it belongs in the opportunity cost. A complete marginal incentive calculation also distinguishes redirected existing customers from newly created insurance sales and incorporates the cost of additional insurance exposure. Insurance portfolios may have binding underwriting and capital constraints. Recaptured risk of poor quality can reduce rather than improve the integrated group's expected contribution.
Effect depends on the response of independent carriers. A rival may lower commissions, deploy direct sales or specialize, while new entry may require sufficient premium scale to support licensing, underwriting infrastructure and fixed operating costs. Hypothetically, a carrier with incremental contribution of 40 per policy and annual avoidable fixed costs of 2 million needs 50,000 comparable policies merely to cover those expenses. If it previously obtained 60,000, loss of 15,000 with no substitute would move the remaining volume below that illustrative threshold. A competitor with alternative portfolios or lower fixed costs may respond differently. The exercise demonstrates the link between opportunity access and minimum viable scale; real thresholds require proper insurance-cost data.
UK evidence on purchase occasions and platforms
The UK's Payment Protection Insurance Order (2011) arose from a competition market investigation. Its official description identifies restrictions on selling PPI before or at the credit purchase and on single-premium payment. This is an institutional example of intervention directed at timing and payment structure. The order's detailed scope and exceptions require consultation of the operative text in individual applications.
The FCA's (2014) general insurance add-ons study examined cover attached to financial and nonfinancial primary purchases using data and materials from carriers and intermediaries. It illustrates how insurance can be sold at purchase occasions controlled by non-insurance businesses. Its price comparison website review in the same year estimated that about a third of 26.6 million motor policies written in 2013 were sold through comparison sites and raised concerns about how characteristics beyond headline price were conveyed. Comparison infrastructure can enhance search and access while also making presentation and platform incentives relevant.
The 2015 motor insurance order constitutes a distinct market-investigation remedy under the Enterprise Act 2002 and imposes continuing compliance obligations described on the official publication page. Its existence does not establish that a superficially similar clause has the same legal status in another insurance market.
A wide most-favored-nation provision in a comparison-platform contract can restrict a carrier from offering a lower quote for the same policy through alternative platforms or direct channels. A narrow parity clause is conceptually limited to the direct channel. Wide parity can reduce platforms' ability to attract customers through lower commissions passed on in insurance quotes; a defense may point to investments in customer acquisition and the risk that users obtain information on one platform and transact elsewhere. The economic magnitude depends on actual differences in distribution costs, the fraction of contracts effectively covered, platform switching and how consumers use the comparison service.
The CMA's Compare the Market case chronology records a 2020 infringement decision concerning wide parity terms in home insurance. In BGL v CMA (2022), the Competition Appeal Tribunal held that the authority had failed to establish its asserted anticompetitive effects and ruled for BGL. The case shows why evidence of contractual restriction and a plausible economic mechanism must be distinguished from proven market effects. The source examined here is the official judgment summary and chronology; the study does not offer a paragraph-by-paragraph interpretation of the full judgment.
Comparative cases outside the UK platform setting
The Italian authority's Compass non-compliance announcement of May 2020 describes an earlier finding about compulsory bundling of personal credit with insurance that was unnecessary for obtaining the loan and reports the opening of a subsequent compliance proceeding. The evidence supports scrutiny of real purchase freedom and timing. The announcement is not a final decision on that compliance proceeding and does not establish exclusion of other insurers under a defined competition-market test.
The US health-sector transaction involving a pharmacy and insurance group illustrates the importance of identifying each competitive relationship separately. The Justice Department's 2018 CVS-Aetna divestiture announcement described a proposed remedy concerning the individual Medicare Part D prescription-drug-plan overlap. The horizontal overlap was the issue specified in the announcement, despite the broader combination of different health-sector functions. That remedy cannot be used as proof that vertical pharmacy or benefit-management foreclosure was separately established.
The comparison reveals different institutional instruments. A sector investigation may prohibit a practice within an identified category; a credit regulation may condition disclosure treatment upon voluntary purchase; a competition action may require proof under the applicable legal test; merger guidelines assess prospective effects. These instruments can coexist. Their findings cannot be pooled as if they estimated one universal foreclosure rate.
A two-direction foreclosure assessment: UK motorcycle insurance
The CMA's 2024 Venus/Atlanta merger decision supplies an especially useful contrast because the investigation expressly tested foreclosure in both directions. In UK motorcycle insurance, the upstream activity involved underwriting and the downstream activity involved brokerage. The authority examined whether the merged business could impair rival brokers' access to underwriting and whether the acquired broker could restrict rival underwriters' access to policies. It concluded that the transaction did not warrant referral.
For input foreclosure, the authority found insufficient incentive in light of the underwriting benefits of diversified distribution. Maintaining brokers serving different customer groups helped an underwriter reach a broader portfolio. The decision also considered diversion and the possibility that lost supply would shift to alternative underwriters. For customer foreclosure, the authority concluded that there was insufficient ability: Atlanta's approximately 20–30 percent distribution share had to be considered alongside alternative routes and existing broker-panel relationships. The remaining stages of the respective tests were unnecessary once one cumulative condition failed.
Insurance adds a distinctive dimension to these incentives. A distribution outlet delivers a portfolio of contracted contingencies as well as premium income. Concentrating the portfolio within one outlet can alter the dependence structure of risk, the expected claims mix, and access to information used in future underwriting. A strategy that concentrates sales could increase marketing margins while worsening risk-adjusted underwriting contribution. The appropriate measure of the integrated entity's incentive therefore incorporates underwriting profit and risk as well as the retail commission.
In a purely illustrative portfolio of 1,000 policies, each generating an expected net underwriting contribution of 40 before fixed costs and capital costs, a change of customer mix that raises expected claims by 12 per policy reduces aggregate expected contribution from 40,000 to 28,000. A distribution margin improvement of 8 per policy would restore only 8,000 of the lost contribution. The combined change remains negative by 4,000 before considering the cost of additional capital or a change in claim dependence. All inputs are hypothetical. The example demonstrates why a margin calculation based exclusively on the broker can misstate economic incentives.
A parallel analysis concerns the rival deprived of business. Losing a distribution agreement may reduce premium income, but the change in expected claims depends on the risk characteristics of the departing policyholders. If losses are concentrated among similar risks, less exposure could partly offset lost revenue. If the affected portfolio supported diversification or statistical precision, the loss could instead undermine pricing and increase the cost of capital. A foreclosure assessment should distinguish accounting volume, risk-adjusted contribution and competitive viability.
Comparing the decisions
The Fortis/BCP and Venus/Atlanta records illustrate the importance of separating control from foreclosure. In Portugal, the joint venture's dependence on one parent distribution network helped establish joint control despite unequal equity stakes. In the UK motorcycle case, the authority examined the ability and incentive to damage outside rivals and concluded that the particular transaction did not meet the applicable prospective competition test. Neither decision establishes a universal safe share or an automatic prohibition based on ownership.
A useful comparative case record therefore stores the legal standard, relevant activity, product segment, geography, estimated share denominator, counterfactual, observed alternatives, incentive evidence and the authority's procedural conclusion. The event date and status are essential: a clearance records an assessment of a transaction as notified, whereas a conduct infringement decision concerns a different type of evidence and legal test. Outcomes after implementation would require a separate empirical inquiry. The comparison establishes the analytical value of the two-direction inquiry without imputing unobserved post-merger effects.
Integrated Scientific Synthesis and Future Research
The central analytical result
Vertical integration alters who controls the opportunity to make an insurance offer and who captures the returns from a completed sale. The control of a valuable customer gateway can matter even when underwriting concentration is unchanged. Competitive assessment should therefore describe the network of risk carriers, distribution institutions, product types and customer purchase occasions before applying a market-level conclusion.
The institutional evidence supports a conditional interpretation. Close bank-insurer ties and group products in the European credit protection review show why ownership and contract structure deserve study. The Asian company filing shows that exclusive alliances can coexist with several distribution channels. Insurance distribution scholarship identifies possible service and matching advantages of alternative channels. The UK appeal demonstrates the consequence of insufficient proof of competitive effects. These observations are consistent with established economics: integration may be an efficient way of organizing a transaction, and under other circumstances it may reserve access or reduce the degree of competition.
A key distinction is that channel diversity must be assessed in effective rather than nominal terms. The availability of a website does not show that a borrower can complete the same protection purchase while preserving the loan conditions. The existence of a broker does not show that it can reach the relevant population at comparable customer-acquisition cost. Conversely, the existence of a dominant bank partner does not show that rival carriers lack accessible customers or alternatives. The empirical question is which alternatives are realistic for the relevant seller or customer over the time horizon studied.
Designing a proportionate intervention
Economic assessment can organize potential responses according to the actual obstacle. If the problem is insufficient price and coverage information, comparable disclosure and better presentation may improve informed choice. If a customer is pressured at the point of financing, independent choice procedures, separation of decisions and documented alternatives may help address that timing problem. If a contract reserves capacity or customer access on excessive terms, analysis can examine whether shorter terms, nonexclusive rights or objectively specified access would protect the relevant investment while preserving competitive opportunities.
These examples are conditional policy implications, not proposed universal legal requirements. An access obligation may create new operational costs, data risks and service failures because insurers differ in underwriting appetite, capital and product expertise. A nondiscrimination commitment requires measurable comparable transactions, transparent review, justified exceptions and a mechanism to detect circumvention. A divestiture can remove shared control but may leave contractual dependence intact. Monitoring should examine both implementation and outcomes.
Where feasible, credible efficiency claims should identify incremental real benefits, their beneficiaries, the alternative organization under which they might be obtained, required investment and how reductions in costs affect policyholder outcomes. The comparison can include expected risk-adjusted claims cost, coverage, underwriting responsiveness, premium persistence and service quality. A remedy that opens access but destabilizes existing contracts or jeopardizes claims continuity needs careful transition safeguards.
Empirical research agenda
A contract-level research program would gather anonymized distribution agreements, carrier ownership records and regulatory product classifications, and join them to event-level offers and enrollments. The most informative outcomes would include quality-adjusted premium, quote success, conversion, acquisition expense, product cancellation, complaints, claim acceptance and time to settlement. Data access and privacy constraints must be handled before statistical design.
A credible analysis of exclusivity expiry would track commercially meaningful openings and compare entry attempts across unaffected product segments. A design examining a remedy could exploit phased implementation or clearly dated contractual changes where plausible comparable untreated observations exist. Estimation should test pre-existing trends, anticipate policy announcements, account for spillovers, and evaluate composition changes. Claims measures need sufficient development periods so that a short-run improvement in price is not conflated with delayed deterioration of claims service. A loan-cohort design should align insurance coverage duration with the relevant loan exposure and distinguish full repayment from risk events.
Market boundaries should be varied using reasonable substitutes. Sensitivity analyses can evaluate product-specific and combined channels, direct and intermediary placements, policy counts and premium measures, gross and net margins, and alternative estimates of diversion. The identification strategy, data cleaning, cohort exclusions and standard-error treatment require reproducible documentation. Researchers should archive data versions and code where confidentiality permits and report null findings and inconsistent outcomes as carefully as positive findings.
The open questions concern the strength and persistence of alternative distribution routes, the conditions under which a competitive tender transfers surplus to customers, and how digital presentation affects effective supplier access. A separate question concerns the extent to which customer-experience and conduct indicators predict market structure effects. Answering these questions requires firm-level contract and event data unavailable in this documentary study. The authorial contribution is the integration of institutional findings, source-status distinctions, economic incentive calculations and a transparent research design within the approved subject boundaries. No priority or novel theoretical system is claimed.
Conclusion
Insurance sector vertical integration coordinates the production and distribution of contingent financial protection. Efficiency gains can arise from joint investment, lower search and acquisition costs, coordinated service and reduced successive markups. Competitive risks arise where control of significant routes to market gives an integrated business the ability and incentive to diminish alternatives and weaken the capacity of rivals to compete.
Bancassurance provides a particularly important setting because customer relationships originate in banking services and the timing of credit decisions can influence insurance purchases. The documented European supervisory sample demonstrates that close relationships and group contracts are prevalent within the studied credit protection population. Company disclosure and insurance distribution research show that institutional forms and measured efficiency vary. These findings justify investigation of control rights, remuneration, risk-adjusted cost and effective customer choice.
The international comparison also establishes an evidentiary discipline: conduct concerns, market-investigation remedies, regulatory obligations and adjudicated competition effects must be interpreted according to their legal scope and procedural status. A restriction alone does not establish harm; a formal choice does not establish commercially realistic alternatives. The analysis identifies the measures and observations necessary to evaluate foreclosure while keeping hypothetical calculations separate from observed institutional evidence.
The resulting conclusion is conditional and testable. An arrangement's economic consequences depend on the value of the controlled distribution opportunity, available substitution, consolidated incentives, the competing firms' ability to adapt and the quality of protection delivered. Documented efficiencies and plausible counterfactuals form part of the same analysis. The empirical agenda specifies how a subsequent contract-level study could estimate the magnitude and distribution of effects. The present monograph establishes a comparative institutional synthesis rather than a global causal estimate.
References
Berger, A. N., Cummins, J. D., & Weiss, M. A. (1997). The coexistence of multiple distribution systems for financial services: The case of property-liability insurance. Journal of Business, 70(4), 515–546. Source
Chang, P.-R., Peng, J.-L., & Fan, C. K. (2011). A comparison of bancassurance and traditional insurer sales channels. Geneva Papers on Risk and Insurance—Issues and Practice, 36(1), 76–93. Source
Competition and Markets Authority (CMA). (2015). Private Motor Insurance Market Investigation Order 2015. March 18; publication page updated December 24, 2024. Source
Competition and Markets Authority (CMA). (2021). Merger Assessment Guidelines (CMA129). March 18. Source
Competition and Markets Authority (CMA). (2022). Price comparison website: use of most favoured nation clauses. Case 50505, case chronology, including infringement decision of November 19, 2020 and appeal update of August 8, 2022. Source
Competition and Markets Authority (CMA). (2024). Anticipated acquisition by Venus Topco Limited of Atlanta Investment Holdings 3 Limited, and acquisition by Tara TopCo Limited of an interest in Venus Topco Limited and Saturn Holdings Limited. ME/7071/23. Decision March 26, published May 14. Source
Competition Appeal Tribunal. (2022). BGL (Holdings) Limited & Others v Competition and Markets Authority, [2022] CAT 36, Case 1380/1/12/21, August 8. Official judgment summary. Source
Competition Commission. (2011). Payment Protection Insurance Market Investigation Order 2011. Official publication page maintained by the Competition and Markets Authority, including the 2018 variation. Source
Consumer Financial Protection Bureau (CFPB). (n.d.). Regulation Z, 12 CFR § 1026.4, Finance charge, including official interpretations. Electronic regulation. Source
European Commission. (2005). Case COMP/M.3556 — Fortis/BCP. Decision under Article 6(1)(b) of Regulation (EC) No 139/2004, January 19. Source
European Commission. (2008). Guidelines on the assessment of non-horizontal mergers under the Council Regulation on the control of concentrations between undertakings. Official Journal of the European Union, C 265, 6–25, October 18. Source
European Commission. (2026). Review of the Merger Guidelines. Directorate-General for Competition. Source
European Insurance and Occupational Pensions Authority (EIOPA). (2022). Credit Protection Insurance (CPI) Sold via Banks: Thematic Review. Publications Office of the European Union. Source
European Parliament and Council. (2016). Directive (EU) 2016/97 of 20 January 2016 on insurance distribution (recast). Official Journal L 26, 19–59. Consolidated text dated October 9, 2024. Source
Financial Conduct Authority (FCA). (2014). General insurance add-ons market study: Final findings. July 24; publication page updated April 1, 2016. Source
Financial Conduct Authority (FCA). (2014). TR14/11: Price comparison websites in the general insurance sector. July. Source
International Association of Insurance Supervisors (IAIS). (2024). Insurance Core Principles and Common Framework for the Supervision of Internationally Active Insurance Groups. Standard-setting overview, including December 2024 update. Source
Italian Competition Authority (AGCM). (2020). IP333—ICA: opening of a non-compliance proceedings against Compass. May 4. Source
Manulife Financial Corporation. (2026). Annual Information Form, February 11, 2026, Exhibit 99.3, U.S. Securities and Exchange Commission filing. Source
Regan, L., & Tennyson, S. (1996). Agent discretion and the choice of insurance marketing system. Journal of Law and Economics, 39(2), 637–666. Source
Rubio-Misas, M. (2022). Bancassurance and the coexistence of multiple insurance distribution channels. International Journal of Bank Marketing, 40(4), 724–745. DOI: 10.1108/IJBM-04-2021-0129. Source
U.S. Department of Justice and Federal Trade Commission. (2023). Merger Guidelines, Guideline 5: Mergers Can Violate the Law When They Create a Firm That May Limit Access to Products or Services That Its Rivals Use to Compete. December 18. Source
U.S. Department of Justice. (2018). Justice Department Requires CVS and Aetna to Divest Aetna’s Medicare Individual Part D Prescription Drug Plan Business to Proceed with Merger. October 10. Source