A European Commission report threatens the continued existence of the Insurance Block Exemption Regulation. The regulation protects essential industry practices, namely the sharing of statistics and certain joint insurance arrangements (ie, line slips, pools, consortia and joint-binding authorities). The report invites the sector to explain why the regulation should not lapse, without renewal, when it ends in 2017. The report appears to be based on partial data and misunderstands key features of the industry. It will be important for stakeholders to respond robustly to the report's conclusions. The loss of this protection – especially on flawed grounds – creates potentially damaging legal uncertainty.
Joint compilations, tables and studies
The insurance industry is unusual. The sharing of data is essential to its operation. The better the information available on claims or risks, the more efficient and accurate the rating process. For 'longtail' claims (eg, asbestosis or environmental contamination), claims data can be scarce and the true costs revealed only decades after the insurance has been sold. The better the data, the easier it is for incumbents and new entrants to compete and so competition and consumers benefit. With the EU Solvency II Directive setting out stricter risk-capital requirements and requiring 'best-estimate' liability calculations, data sources are becoming more important than ever.(1)
The regulation exempts agreements on joint compilations, joint tables and studies for certain purposes, including the calculation of the average cost of covering a specified risk. The commission states that industry-specific protection is not required to protect these benefits. The Guidelines on Horizontal Cooperation already offer guidance on information sharing between competitors.Why would the insurance sector require greater protection? The report states that, at most, it might supplement general guidance on competitor information sharing with an insurance-specific communication.
Practitioners are likely to take issue with this appraisal. The general guidance starts from the unspoken premise of bricks-and-mortar industries where any price-related information sharing is regarded with suspicion as an adjunct to cartel-type activity. The exchange of risk premium information (the average cost of underwriting a risk based on historic claims), in particular, has no analogue in traditional industry information exchange. Under the general guidance, it would be quickly branded price-related information likely to have an adverse effect on competition.(2) Also, as a technical matter, the guidance has no binding effect on courts or national regulators, in the absence of specific national rules to that effect. In the litigious insurance sector, any legal edge is seen as fair game in a dispute. Abandoning the binding contours of a block exemption for ill-fitting and non-binding guidance leaves the industry little protection.
Co-insurance and co-reinsurance pools
The insurance industry relies on being able to pool capacity quickly and nimbly. This responds to customer demand for expanded capacity or new or improved insurance products (eg, bloodstock, cyberterrorism, data corruption, classic cars or other niche or evolving areas of risk). An entrepreneurial underwriter or broker who sees an opening in the market can quickly gather a group of likeminded insurance carriers to add capacity or offer a new or improved product. The names given to the arrangements (eg, joint binders, pools, consortia, line slips and panels) are diverse and differ from market to market. However, they all involve some form of capacity pooling. The pools tend to be led by a lead insurer (or insurers) or intermediary, which accepts business on behalf of the group based on a pre-determined pricing mechanic. They tend to coalesce and write business over a short term, and represent a means by which smaller or medium-sized insurers can gain exposure to lines of business that they might otherwise not write, or only to a limited degree. This arrangement allows brokers to place capacity behind their own design of wordings, proposition benefits and branding. The broker's scheme then competes with the insurer's own brand offering.
These should not – as the EU report mistakenly does – be confused with the kinds of large-scale institutional pools set up by the state or often with state backing to cover otherwise potentially uninsurable risks, such as terrorism or nuclear catastrophe.(3) While these certainly exist, they are exceptional, often state-backed arrangements and are unlikely to irk a competition authority. Consortia, line slips and joint binders, in contrast, are common features of many markets.
The regulation exempts co-insurance pools – and 'co-reinsurance' defined as a pool which mutually reinsures and incidentally collectively buys in reinsurance – if they cover a new risk or if the joint market share of the pool members does not exceed 20% for co-insurance pools (or 25% for co-reinsurance).
The report focuses on institutional pools, confirming that insurers often need to cooperate to cover certain large unconventional risks (eg, terrorism, nuclear power or environmental risks). In these sectors, the size and spread of risks render them difficult or impossible to insure individually.
The report concludes that regulatory protection is unnecessary. First, it says that institutional pools often fall outside regulation conditions because they tend to exceed the market-share threshold. This is difficult to understand in light of the regulation's recitals explaining that co-insurance of these large atypical risks commonly falls outside Article 101(1) of the Treaty on the Functioning of the European Union (TFEU) altogether. If the state-backed pool becomes the last-resort insurer, it demonstrates that there was never any individual insurer competition for these risks in the first place.(4)
Confusingly, it adds that other types of pool (eg, line slips or other broker-led arrangements) are alternatives to institutional pools. Hence, there is no need for protection.
However, this is a fundamental misunderstanding of the regulation's terms, which plainly cover any type of intermediary or insurer-led form of co-insurance group. Nowhere in the regulation are pools labelled 'institutional' or 'non-institutional'. Rather, the definition is broad – namely, a group of insurers underwriting risks directly, through a lead insurer or via intermediaries.(5) The only exception is subscription market business.(6) Thus, any type of joint arrangement falling short of subscription market placement (eg, consortia, joint binders, line slips and other intermediary-led arrangements) falls under the regulation. Further confusion is caused by the report's implication that the Dutch Competition Authority concluded that such arrangements fell outside the regulation. The Dutch authority has not suggested this; in fact, its only statements strongly imply the contrary.(7)
To let hard-pressed compliance departments quickly review line slips and consortia, the regulation represents important administrative simplicity. The alternative – a self-assessment entailing lengthy economic and legal review – creates harmful legal uncertainty. A risk-averse compliance team may simply put the pool on the 'too difficult' pile and so kill off a valuable and important entrepreneurial market practice.
The second reason given for dispensing with protection is still less credible. The report states that the insurance sector is not the only sector in which undertakings tend to cooperate on specific large projects to spread the costs and risks involved (eg, large construction projects). The report does not evidence why construction and insurance services may be lumped together. More fundamentally, it fails to understand why and how insurers engage in joint insurance arrangements hitherto protected by the regulation.
The third reason is also difficult to understand. Market definition in the insurance industry is said to be complex and national competition authorities find the rules difficult to apply. The report refers to the Federal Cartel Office's (Bundeskartellamt) well-publicised reversal before the appeal courts, when both the German Court of Appeal and the Supreme Court criticised the Bundeskartellamt's prohibition of a professional services insurance pool based on market definition. Market definition is central to competition law – this includes guidelines and block exemptions in every other field. The report is not necessarily saying that market definition poses an insuperable barrier for this one industry.
Can it be right that the regulation's survival should depend on making its application easier for competition authorities – with a staff of antitrust experts – rather than the already stretched, typically non-specialist compliance resources of insurers? It could be said that the regulation is there to make the regulated entities' compliance obligations lighter, not to stack the deck against them to assist expert regulators.
If the commission decides not to extend the regulation, protection will lapse for:
- joint compilations, tables and studies; and
- co-insurance and co-reinsurance pools.
This would not mean that they will be prohibited; rather, they would have to be assessed under the same general EU competition law principles as other sectors.
Insurers involved in co-insurance or statistical exchanges would have to assess whether the scheme is restrictive under Article 101(1) of the TFEU and, if so, assess whether it generates pro-competitive benefits of a scope and kind to satisfy Article 101(3) of the TFEU. This is no easy task. It calls for economic appraisals and proof of unknowables – for example, is there a less restrictive means of achieving the co-insurance arrangement (with fewer insurers or different insurers) or statistical exchange? And would other data sources or a less comprehensive data set be just as effective? The burden falls on the insurer to prove that benefits outweigh the restriction.
The insurance industry is unusual in antitrust terms. Cooperation is wide-ranging and essential to allow for efficient rating and underwriting of risks. It enables smaller players to compete effectively with larger rivals. Those facets of the industry are recognised in special legal regimes, not just in Europe – via the regulation – but also in the United States (through the McCarran-Ferguson Act) and elsewhere.
The regulation report is regrettable – it challenges legitimate practices based on flawed reasoning and partial data. Rather than asking whether the quarter-century application of the regulation (and its predecessors) has harmed competition,(8) the presumption is that any sector-specific block exemption should be removed. However, there is no suggestion that markets are uncompetitive due to the regulation. To take just one data point (which is not cited in the report), annual premiums and claims costs in the motor vehicle insurance sector in Germany decreased after the first (1992) and second (2003) adoption of the regulation.(9)
If the regulation lapses, companies could conclude that abandonment of co-insurance arrangements or statistical exchanges is the only way to de-risk these projects. Perhaps more practically, compliance departments will tell underwriters that a co-insurance or statistics arrangement can be approved only after an expensive and lengthy legal review, by which time the commercial opportunity will have passed.
Perversely, this would be to the detriment of competition. Insurers able to underwrite the entirety of a risk without co-insurers would no longer face competition from pools abandoned or delayed due to compliance risks. Insurers with a self-generated data set to inform their underwriting strategy will have an edge over those denied access to jointly compiled data because – post the demise of the regulation – it is considered legally too risky.
Policymakers often say stakeholders' pleas for legal certainty mask narrow sectoral interests and special pleading. In the insurance industry – which both needs competitor cooperation to function effectively and is intensely litigious – the need for legal certainty cannot be so easily dismissed. Without the regulation, the industry would face a materially less secure legal environment. As demonstrated in the past, in every cycle downturn there can be a strong incentive for counterparties to turn to competition law to avoid commercially unattractive agreements or claims.
The industry would be better served by a thorough review of the need to continue the exemption or, indeed, to conduct a full overhaul of the regulation if amendments are required. The report does not do this.
On April 26 2016 the commission will organise a meeting with stakeholders to provide an opportunity to discuss the report's findings. Registration for the event will open shortly on the commission's website. After the meeting, the commission will decide on the extension of the regulation, which is due to expire on March 31 2017.
For further information on this topic please contact Bill Batchelor or Jan Kresken at Baker & McKenzie by telephone (+32 2 639 36 11) or email (email@example.com or firstname.lastname@example.org). The Baker & McKenzie website can be accessed at www.bakermckenzie.com.
(1) November 25 2009, Directive 2009/138/EU of the European Parliament and European Council on the Taking-up and Pursuit of the Business of Insurance and Reinsurance, OJ L 335, page 1, December 17 2009.
(2) Horizontal Cooperation Guidelines, paragraph 76: requiring an assessment of how the information exchanged affects pricing strategy. Risk premium data collected and disseminated market-wide can have a direct impact on rating.
(3) See, for example, COMP/37.363, Svenska Atomförsikringspoolen; COMP/34.985, Pool Italiano Rischi Atomici; and COMP/34.558, Aseguradores Riesgos Nucleares (nuclear insurance pools cleared because there would be no supply of liability insurance with adequate coverage for the risks involved without the pooling agreements); Austrian Terrorpool and German Pool Extremus (2002) (pools outside Article 101(1) as the commission could not clearly demonstrate that that the pool was unnecessary or could be replaced by more than one pool); OFT, CA98/03/2004 Pool Reinsurance Company Limited (although restrictive of competition, the requirements for individual exemption were satisfied).
"Co-insurance pools means groups set up by insurance undertakings either directly or through brokers or authorized agents… which:
(a) agree to underwrite, in the name and for the account of all the participants, the insurance of a specified risk category; or
(b) entrust the underwriting and management of the insurance of a specified risk category, in their name and on their behalf, to one of the insurance undertakings, to a common broker or to a common body set up for this purpose"
(6) Only subscription market business is excluded from the regulation 'pool' definition. However, subscription business is fundamentally different from co-insurance arrangements via pools, consortia, line slips or joint binders. When a single risk is placed on a subscription market basis, it entails only bilateral negotiation between the broker or insured and each insurer. The insurer is free to refuse a quote or, as a follower, to decline a request to match a lead insurer's terms. There is no pre-existing arrangement which limits the insurer's pricing freedom by committing to underwrite business at pre-agreed rates or as approved by a lead, or co-lead, underwriter.
(7) The report states that the Dutch insurance industry did not consider co-insurance arrangements established by intermediaries to be pools within the meaning of the regulation (page 26); however, a press release states that the notaries pool under examination was covered by the regulation (see www.acm.nl/en/publications/publication/6289/NMa-more-competition-among-insurers).
(8) See EU Regulation 3932/92, December 21 1992, on the Application of Article 85(3) of the Treaty to Certain Categories of Agreements, Decisions and Concerted Practices in the Insurance Sector, OJ L 398, December 31 1992, pages 7 to 14; and EU Regulation 358/2003, February 27 2003, on the Application of Article 81(3) of the Treaty to Certain Categories of Agreements, Decisions and Concerted Practices in the Insurance Sector, OJ L 53, February 28 2003, pages 8 to 16.
This article was first published by the International Law Office, a premium online legal update service for major companies and law firms worldwide. Register for a free subscription.